Early Algebraization

Advances in Mathematics Education
Editors:
Prof. Dr. Gabriele Kaiser, University of Hamburg
Prof. Dr. Bharath Sriraman, The University of Montana
For other titles published in this series, go to

www.springer.com/series/8392
Jinfa Cai Eric Knuth
Editors
Early
Algebraization
A Global Dialogue
from Multiple Perspectives
Prof. Jinfa Cai Prof. Eric Knuth
Dept. Mathematical Sciences University of Wisconsin-Madison
University of Delaware Teacher Education Building,
Ewing Hall 501 225 N. Mills Street
Newark, DE 19716 Madison, WI 53706
USA USA
[email protected] [email protected]
ISSN 1869-4918 e-ISSN 1869-4926

ISBN 978-3-642-17734-7 e-ISBN 978-3-642-17735-4
DOI 10.1007/978-3-642-17735-4
Springer Heidelberg Dordrecht London New York
Library of Congress Control Number: 2011922614
© Springer-Verlag Berlin Heidelberg 2011

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9,
1965, in its current version, and permission for use must always be obtained from Springer. Violations
are liable to prosecution under the German Copyright Law.
The use of general descriptive names, registered names, trademarks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective
laws and regulations and therefore free for general use.
Cover design: deblik
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Foreword
Early Algebraization: A Global Dialogue from Multiple Perspectives is the second

monograph in the Advances in Mathematics Education (AiME) series launched by
Springer in 2009. The book follows in the tradition of Theories of Mathematics
Education (Sriraman and English, monograph 1), stemming from a previous ZDM
issues on early algebraic thinking (vol. 37, no. 1, 2005 and vol. 40, no. 1, 2008).
That is, although it uses the previous issues as a basis for the current monograph, the
monograph itself goes beyond simply revisiting the past. It conveys the present state
of the art on existing research on early algebraization since 2005. The eight previous
articles (five from vol. 37 and 3 from vol. 40) have been reworked and updated
in addition to 18 new chapters from researchers involved in early algebraization
research projects in different parts of the world, which include 4 commentaries on
the scope of the research.
The book editors Jinfa Cai and Eric Knuth have compiled the book in three
substantial parts between the bookends of a general introduction and an over-
all commentary addressing perspectives for research and teaching in this do-
main of inquiry. These three parts of the book examine curricular, cognitive
and instructional components of early algebraization. Unlike the ZDM issue
which was predominantly articles from researchers based in North America, this
book contains ongoing research from different parts of the world, and initi-
ates a global conversation on where the community stands in its research find-
ings.
AiME is distinct from other mathematics education series because it attempts
to draw the reader into a conversation, and be dialogic in its presentation. This is
the purpose of soliciting commentaries from those that are able to synthesize ideas,
expose them in a larger light of what is known, and directions in which they can
be further pushed. This book continues in this tradition and attempts to draw us
into the issues of understanding, implementing and assessing early algebraization in
projects and curricula in different parts of the world. We hope this monograph is of
value to the research community of mathematics educators interested in the role and
significance of early algebraic thinking within the current research architecture. We
v
vi Foreword
appreciate the efforts of the authors that are a part of this book and thank the book
editors (Cai & Knuth) for this superb book.
Gabriele Kaiser
Bharath Sriraman
Introduction
A Global Dialogue About Early Algebraization from
Multiple Perspectives
Kilpatrick and Izsák (2008) quoted an anonymous editorial writer to start their
chapter in the National Council of Teachers of Mathematics’ 70th Yearbook:
“If there is a heaven for school subjects, algebra will never go there. It is the
one subject in the curriculum that has kept children from finishing high school,
from developing their special interests and from enjoying much of their home
study work. It has caused more family rows, more tears, more heartaches, and
more sleepless nights than any other school subject.” (p. 3) Even though there
has been a dramatic change for the world 70 years ago when the editorial was
written to nowadays, the status of algebra as a school subject has not changed
much—algebra is important but many students experience difficulties (Kieran 2007;
Loveless 2008; National Mathematics Advisory Panel [NMAP] 2008). In fact, alge-
bra has been characterized as the most important “gatekeeper” in school mathemat-
ics.
Given its gatekeeper role as well as growing concern about students’ inade-
quate understandings and preparation in algebra, algebra curricula and instruc-
tion have become focal points for policy makers and mathematics education re-
searchers around the world (e.g., Bednarz et al. 1996; Lacampagne et al. 1995;
RAND Mathematics Study Panel 2003; Stacey et al. 2004). An important emphasis,
common around the globe, is the development of students’ algebraic thinking in ear-
lier grades. The development of students’ algebraic thinking in earlier grades is not
a new idea; in China and Russia, for example, algebraic concepts were introduced
to elementary school students in the 50s and 60s. In other countries (e.g., Europe,
North America), the discussion of integrating algebraic ideas into mathematics cur-
ricula in the earlier grades started in the 70s. In the past decade, however, there has
been an increased emphasis on and wider acceptance for developing students’ alge-
braic ideas and thinking in earlier grades, reflected in a number of influential policy
documents. For example, in the United States, the NCTM proposed algebra as a
content strand for all grade levels (NCTM 2000). In fact, it is widely accepted that
to achieve the goal of “algebra for all”, students in elementary and middle school
must have experiences that better prepare them for more formal study of algebra in
vii
viii Introduction
the later grades. Yet, only recently have researchers started to explore issues related
to early algebraization.
Although a chapter on research on school algebra appeared in the Handbook of
the Research on Mathematics Teaching and Learning (Grouws 1992), its focus was
primarily on algebra at the secondary school level. In the Second Handbook of the
Research on Mathematics Teaching and Learning (Lester 2007), there is again a
chapter on algebra at the secondary school level, however, the volume now also in-
cludes a chapter on early algebra learning. In fact, this is the only chapter with such
a focus in mathematics education research handbooks published in the past two
decades. A similar trend can also be seen in publications directed toward teachers:
In 1993, NCTM published a volume focused on research ideas for the elementary
school classroom that did not include any chapters focused on early algebra learn-
ing; in contrast, NCTM recently published a similar volume (Lambdin and Lester
2010) that does include a chapter on early algebra learning. On one hand, such
changes suggest that the field has known enough about early algebraization to syn-
thesize research findings in the area. On the other hand, as Carraher and Schliemann
(2007) recently pointed out: “Although there is some agreement that algebra has a
place in the elementary school curriculum, the research basis needed for integrat-
ing algebra into the early mathematics curriculum is still emerging, little known,
and far from consolidated.” (p. 671) In fact, curriculum developers, educational re-
searchers, teachers, and policy makers are just beginning to think about and explore
the kinds of mathematical experiences and knowledge students in early grades need
to be successfully prepared for the formal study of algebra in the later grades. This
monograph is part of such an effort.
Early Algebraization
Traditionally, most school mathematics curricula separate the study of arithmetic
and algebra—arithmetic being the primary focus of elementary school mathemat-
ics and algebra the primary focus of middle and high school mathematics. There is
a growing consensus, however, that this separation makes it more difficult for stu-
dents to learn algebra in the later grades (Kieran 2007). Moreover, based on recent
research on learning, there are many obvious and widely accepted reasons for de-
veloping algebraic ideas in the earlier grades (Cai and Knuth 2005). The field has
gradually reached consensus that students can learn and should be exposed to alge-
braic ideas as they develop the computational proficiency emphasized in arithmetic.
In addition, it is agreed that the means for developing algebraic ideas in earlier
grades is not to simply push the traditional secondary school algebra curriculum
down into the elementary school mathematics curriculum. Rather, developing alge-
braic ideas in the earlier grades requires fundamentally reforming how arithmetic
should be viewed and taught as well as a better understanding of the various factors
that make the transition from arithmetic to algebra difficult for students.
The transition from arithmetic to algebra is difficult for many students, even for
those students who are quite proficient in arithmetic, as it often requires them to
Introduction ix
think in very different ways (Kieran 2007; Kilpatrick et al. 2001). Kieran, for ex-
ample, suggested the following shifts from thinking arithmetically to thinking alge-
braically: (1) A focus on relations and not merely on the calculation of a numerical
answer; (2) A focus on operations as well as their inverses, and on the related idea of
doing/undoing; (3) A focus on both representing and solving a problem rather than
on merely solving it; (4) A focus on both numbers and letters, rather than on num-
bers alone; and (5) A refocusing of the meaning of the equal sign from a signifier
to calculate to a symbol that denotes an equivalence relationship between quanti-
ties. These five shifts certainly fall within the domain of arithmetic, yet, they also
represent a movement toward developing ideas fundamental to the study of algebra.
Thus, in this view, the boundary between arithmetic and algebra is not as distinct as
often is believed to be the case.
What is algebraic thinking in earlier grades then? Algebraic thinking in earlier
grades should go beyond mastery of arithmetic and computational fluency to at-
tend to the deeper underlying structure of mathematics. The development of alge-
braic thinking in the earlier grades requires the development of particular ways of
thinking, including analyzing relationships between quantities, noticing structure,
studying change, generalizing, problem solving, modeling, justifying, proving, and
predicting. That is, early algebra learning develops not only new tools to understand
mathematical relationships, but also new habits of mind. In this volume, we focus
on the development of algebraic ideas in both elementary and middle schools.
In this volume, the authors address the issues of early algebraization from curricular,
cognitive, and instructional perspectives. The inclusion of middle grades is desirable
because of the critical transition from elementary to the middle grades, particularly
related to algebra learning. The inclusion of issues related to curriculum, cognition,
and instruction is based on the consideration that they are the three most fundamen-
tal perspectives for mathematics education. Curricula have a significant influence
on what students learn (NCTM 2000) and have been found to contribute to mathe-
matical performance differences in cross-national studies (Schmidt et al. 1996). Ac-
cordingly, the examination of curricula from various nations can provide a broader
point of view regarding curricular approaches to integrating algebraic ideas into
earlier grades as well as providing insights regarding the development of students’
algebraic thinking.
Although curricula can provide elementary and middle school students with op-
portunities to develop their algebraic thinking, teachers are arguably the most impor-
tant influence on what students actually learn. Thus, the success of efforts to develop
students’ algebraic thinking rests largely with the ability of teachers to foster such
thinking.
The design of curricula and professional development programs as well as the en-
actment of instructional practices intended to support the development of students’
x Introduction
algebraic thinking are all dependent, to a great extent, on what we know about stu-
dents’ algebraic thinking and its development. Thus it is critical to examine issues
related to students’ cognition in algebra learning.
As we look across this set of articles in this volume, with their variety of foci
and perspectives, two cross-cutting themes surfaced. First is the importance of bet-
ter integrating into current school mathematics practices opportunities for students
to develop their algebraic thinking. These opportunities include both the design of
curricula, at the elementary school level in particular, that pays explicit attention to
making connections between arithmetic and algebra, and the recognition of oppor-
tunities to strengthen these connections as students progress through middle school.
The second theme to emerge is the importance of supporting teachers’ efforts to
implement practices that foster the development of students’ algebraic thinking. If
future generations of students are to become better prepared for more formal study
of algebra in the later grades, then likewise teachers must also be better prepared.
The articles in this volume provide guidance and suggestions for continued work
in the area of early algebra research regarding teachers’ instructional practices and
professional development.
One of the important features of this volume is its international in nature, which
promotes a global dialogue on the topic. Research is presented from many parts
of the world, including Australia, Canada, China, France, India, Italy, Japan, New
Zealand, Russia, Singapore, South Korea, the United Kingdom, and the United
States of America. Such a global dialogue will help us address issues related to
early algebra learning and, ultimately, better prepare greater numbers of students
for success in algebra.
Acknowledgements We gratefully acknowledge the assistance of the National Science Founda-

tion for its support of our research during the past decade. We also thank the Springer editorial
team and the series’ editorial board for their support, including the sustained support of Clemens
Heine, Gabriele Kaiser, and Bharath Sriraman in the process of preparing this volume during the
past three years. We particularly acknowledge the helpful reviews for some of the chapters by
Miriam Amit, Rongjin Huang, Jane Jane Lo, and Manya Sundstrom. Finally, we thank all of the
contributing authors for their quality work and patience.
Jinfa Cai
Eric Knuth
References
Bednarz, N., Kieran, C., & Lee, L. (Eds.) (1996). Approaches to Algebra: Perspectives for Research
and Teaching. Dordrecht: Kluwer Academic Publishers.
Cai, J., & Knuth, E. (Eds.) (2005). Developing algebraic thinking: Multiple perspectives. Zentral-
blatt fuer Didaktik der Mathematik (International Journal on Mathematics Education), 37(1).
Special Issue.
Carraher, D. W., & Schliemann, A. D. (2007). Early algebra and algebraic reasoning. In F. K. Lester
Jr. (Ed.), Second Handbook of Research on Mathematics Teaching and Learning: A Project of
the National Council of Teachers of Mathematics (pp. 669–706). Charlotte, NC: Information
Age Publishing; Reston, VA: National Council of Teachers of Mathematics.
Introduction xi
Chick, E., Stacey, K., Vincent, J., & Vincent, J. (Eds.) (2001). The Future of the Teaching and
Learning of Algebra: Proceedings of 12th ICMI Study. University Melbourne, Australia.
Grouws, D. A. (Ed.) (1992). Handbook of Research on Mathematics Teaching and Learning. Re-
ston, VA: National Council of Teachers of Mathematics.
Kieran, C. (2007). Learning and teaching algebra at the middle school through college levels. In
F.K. Lester Jr. (Ed.), Second Handbook of Research on Mathematics Teaching and Learning.
Charlotte, NC: New Age Publishing; Reston, VA: National Council of Teachers of Mathemat-
ics.
Kilpatrick, J., & Izsák, A. (2008). A history of algebra in the school curriculum. In C. E. Greenes &
R. Rubenstein (Eds.), Algebra and Algebraic Thinking in School Mathematics. 70th Yearbook
of the National Council of Teachers of Mathematics (pp. 3–18). Reston, VA: National Council
of Teachers of Mathematics.
Kilpatrick, J., Swafford, J., & Findell, B. (Eds.) (2001). Adding it up. Washington, DC: National
Academy Press.
Lacampagne, C., Blair, W., & Kaput, J. (1995). The Algebra Initiative Colloquium. Washington,
DC: U.S. Department of Education & Office of Educational Research and Improvement.
Lambdin, D. V., & Lester, F. K. Jr. (Eds.) (2010). Teaching and Learning Mathematics: Translat-
ing Research for Elementary School Teachers. Reston, VA: National Council of Teachers of
Mathematics.
Lester, F. K. Jr. (Ed.) (2007). Second Handbook of Research on Mathematics Teaching and Learn-
ing. Charlotte, NC: Information Age Publishing; Reston, VA: National Council of Teachers of
Mathematics.
Loveless, T. (2008). The misplaced math student: lost in eight-grade algebra. The Brown Center
Report on American Education.
National Council of Teachers of Mathematics (2000). Principles and Standards for School Mathe-
matics. Reston, VA: Author.
National Mathematics Advisory Panel (2008). Foundations for Success: The Final Report of the
National Mathematics Advisory Panel. Washington, DC: U.S. Department of Education.
RAND Mathematics Study Panel (2003). Mathematical proficiency for all students: Towards a
strategic development program in mathematics education. Santa Monica, CA: RAND Corpo-
ration MR-1643-OERI.
Schmidt, W. H., McKnight, C. C., & Raizen, S. A. (1996). A Splintered Vision: An Investigation of
U.S. Science and Mathematics Education. Boston: Kluwer.
Stacey, K., Chick, H., & Kendal, M. (Eds.) (2004). The Future of the Teaching and Learning of
Algebra: The 12th ICMI Study. Kluwer: Dordrecht.
Contents
Part I: Curricular Perspective
Preface to Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Jinfa Cai and Eric Knuth
Functional Thinking as a Route Into Algebra in the Elementary Grades . 5

Maria L. Blanton and James J. Kaput
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
The Challenge of Curriculum and Instruction . . . . . . . . . 6
Functional Thinking as a Route to Algebraic Thinking . . . . 7
Functional Thinking in the Elementary Grades . . . . . . . . . . . 8
Children’s Capacity for Functional Thinking . . . . . . . . . 9
Integrating Functional Thinking into Curriculum and Instruction . . 16
Transforming Teachers’ Resource Base to Support Students’
Functional Thinking . . . . . . . . . . . . . . . . . . 17
Using Children’s Functional Thinking to Leverage Teacher
Learning . . . . . . . . . . . . . . . . . . . . . . . . 19
Creating Classroom Culture and Practice to Support
Functional Thinking . . . . . . . . . . . . . . . . . . 20
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Developing Students’ Algebraic Thinking in Earlier Grades: Lessons

from China and Singapore . . . . . . . . . . . . . . . . . . . . . . . . 25
Jinfa Cai, Swee Fong Ng, and John C. Moyer
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Features of the Chinese and Singaporean Curricula . . . . . . . . . 27
Algebra Emphases in the Chinese and Singaporean
Curricula . . . . . . . . . . . . . . . . . . . . . . . . 27
The Chinese Curriculum . . . . . . . . . . . . . . . . . . . . 28
The Singaporean Curriculum . . . . . . . . . . . . . . . . . . 32
xiii
xiv Contents
Lessons from Chinese and Singaporean School Mathematics . . . . 34

Why Should Curricula Expect Students in Early Grades
to Think Algebraically? . . . . . . . . . . . . . . . . 35
Are Young Children Capable of Thinking Algebraically? . . . 36
How Can We Help Students to Think Arithmetically
and Algebraically? . . . . . . . . . . . . . . . . . . . 37
Are Authentic Applications Necessary for Students in Early
Grades? . . . . . . . . . . . . . . . . . . . . . . . . . 38
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Developing Algebraic Thinking in the Context of Arithmetic . . . . . . . 43

Susan Jo Russell, Deborah Schifter, and Virginia Bastable
Understanding the Behavior of the Operations . . . . . . . . . . . . 45
Generalizing and Justifying . . . . . . . . . . . . . . . . . . . . . 51
1. Articulating General Claims . . . . . . . . . . . . . . . . . 51
2. Developing a Mathematical Argument to Justify a General
Claim . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3. Representation-Based Proof: Tools for Proving
in the Elementary Grades . . . . . . . . . . . . . . . 56
Extending the Number System . . . . . . . . . . . . . . . . . . . . 59
Using Notation with Meaning . . . . . . . . . . . . . . . . . . . . 63
Connecting Arithmetic and Algebra . . . . . . . . . . . . . . . . . 67
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
The Role of Theoretical Analysis in Developing Algebraic Thinking:

A Vygotskian Perspective . . . . . . . . . . . . . . . . . . . . . . . . 71
Jean Schmittau
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Orienting Children to Theoretical Concepts . . . . . . . . . . . . . 74
Role of Psychological Tools . . . . . . . . . . . . . . . . . . . . . 76
The Part-Whole Relation . . . . . . . . . . . . . . . . . . . . 76
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . 84
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
The Arithmetic-Algebra Connection: A Historical-Pedagogical Perspective 87

K. Subramaniam and Rakhi Banerjee
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Arithmetic and Algebra in the Indian Mathematical Tradition . . . . 91
Building on Students’ Understanding of Arithmetic . . . . . . . . . 95
The Arithmetic Algebra Connection—A Framework . . . . . . . . 98
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Shiki: A Critical Foundation for School Algebra in Japanese Elementary

School Mathematics . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Tad Watanabe
School Algebra and Algebra in Early Grades . . . . . . . . . . . . 110
Contents xv
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Algebra in Japanese Curriculum . . . . . . . . . . . . . . . . . . . 112
Mathematical Expressions in Japanese Curriculum . . . . . . . . . 114
Mathematical Expressions in Japanese Textbooks . . . . . . . . . . 114
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Commentary on Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Jeremy Kilpatrick
Algebra First . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
A Curriculum Topic . . . . . . . . . . . . . . . . . . . . . . . . . 127
Numerical Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Word Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Multiple Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . 129
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Part II: Cognitive Perspective
Preface to Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Eric Knuth and Jinfa Cai
Algebraic Thinking with and without Algebraic Representation:

A Pathway for Learning . . . . . . . . . . . . . . . . . . . . . . . . . 137
Murray S. Britt and Kathryn C. Irwin
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Children’s Understanding of Generalities for Operations
Before Schooling . . . . . . . . . . . . . . . . . . . . . . . . 139
Algebraic Thinking and the New Zealand Numeracy Project . . . . 140
Students’ Algebraic Thinking in the Last Year of Intermediate
School (Age 11–12) . . . . . . . . . . . . . . . . . . . . . . 146
The Growth of Algebraic Thinking from Numbers to Symbols:
A Longitudinal Study . . . . . . . . . . . . . . . . . . . . . 147
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
A Pathway for Algebraic Thinking . . . . . . . . . . . . . . . 153
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Examining Students’ Algebraic Thinking in a Curricular Context:

A Longitudinal Study . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Jinfa Cai, John C. Moyer, Ning Wang, and Bikai Nie
Standards-Based and Traditional Curricula in the United States . . 162
LieCal Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Highlights of the Differences between CMP and Non-CMP
Curricula . . . . . . . . . . . . . . . . . . . . . . . . . 164
Defining Variables . . . . . . . . . . . . . . . . . . . . . . . 165
Defining Equations . . . . . . . . . . . . . . . . . . . . . . . 165
Introducing Equation Solving . . . . . . . . . . . . . . . . . 166
Using Mathematical Problems . . . . . . . . . . . . . . . . . 168
xvi Contents

Classroom Instruction . . . . . . . . . . . . . . . . . . . . . 169
Conceptual and Procedural Emphases . . . . . . . . . . . . . 170
Instructional Tasks . . . . . . . . . . . . . . . . . . . . . . . 171
Students’ Development of Algebraic Thinking: Methodological
Considerations . . . . . . . . . . . . . . . . . . . . . . . . . 172
The Focus of Algebraic Thinking . . . . . . . . . . . . . . . 173
Tasks and Data Analysis . . . . . . . . . . . . . . . . . . . . 174
Findings about the Development of Students’ Algebraic Thinking . 174
Representing Situations . . . . . . . . . . . . . . . . . . . . 175
Solving Equations . . . . . . . . . . . . . . . . . . . . . . . 177
Making Generalizations . . . . . . . . . . . . . . . . . . . . 178
Conclusions and Instructional Implications . . . . . . . . . . . . . 180
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Years 2 to 6 Students’ Ability to Generalise: Models, Representations

and Theory for Teaching and Learning . . . . . . . . . . . . . . . . . 187
Tom J. Cooper and Elizabeth Warren
Perspectives on the Mathematics of Early Algebra . . . . . . . . . 188
Representation and Generalisation . . . . . . . . . . . . . . . . . . 190
Models and Representations . . . . . . . . . . . . . . . . . . 191
Generalisation . . . . . . . . . . . . . . . . . . . . . . . . . 191
Focus of EATP . . . . . . . . . . . . . . . . . . . . . . . . . 193
Focus of Chapter . . . . . . . . . . . . . . . . . . . . . . . . 194
Design of EATP . . . . . . . . . . . . . . . . . . . . . . . . 194
Findings and Discussion . . . . . . . . . . . . . . . . . . . . . . . 196
Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Change and Functions . . . . . . . . . . . . . . . . . . . . . 198
Equations and Equivalence . . . . . . . . . . . . . . . . . . . 201
Generalising Principles and Abstract Representations . . . . . 204
Conclusions and Implications . . . . . . . . . . . . . . . . . . . . 206
Models and Representations . . . . . . . . . . . . . . . . . . 206
Generalisation . . . . . . . . . . . . . . . . . . . . . . . . . 207
Theoretical Framework . . . . . . . . . . . . . . . . . . . . . 209
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Algebra in the Middle School: Developing Functional Relationships

Through Quantitative Reasoning . . . . . . . . . . . . . . . . . . . . 215
Amy B. Ellis
What Is Quantitative Reasoning? . . . . . . . . . . . . . . . . . . . 216
The Importance of (and Difficulties with) Functional Thinking . . . 218
An Alternative Approach to Function: Quantities and Covariation . 222
A Flexible Understanding of Functions . . . . . . . . . . . . . . . 226
Coordinating Covariation and Correspondence Approaches . . 226
Flexibility Across Forms . . . . . . . . . . . . . . . . . . . . 230
Contents xvii
Fostering a Focus on Quantities . . . . . . . . . . . . . . . . . . . 234

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Representational Competence and Algebraic Modeling . . . . . . . . . . 239

Andrew Izsák
Early Results on Students’ Understandings of Standard
Representations in Algebra . . . . . . . . . . . . . . . . . . . 241
Theoretical Accounts of Reasoning with External Representations . 241
Students’ Capacities to Reason with External Representations . . . 243
First Result: Criteria for Evaluating External Representations . . . . 244
Second Result: Adaptive Interpretation . . . . . . . . . . . . . . . 249
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Middle School Students’ Understanding of Core Algebraic Concepts:

Equivalence & Variable . . . . . . . . . . . . . . . . . . . . . . . . . 259
Eric J. Knuth, Martha W. Alibali, Nicole M. McNeil, Aaron Weinberg,
and Ana C. Stephens
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Student Understanding of Equivalence & Variable . . . . . . . . . 261
Equivalence . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . 263
Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Interpretation of the Equal Sign . . . . . . . . . . . . . . . . 266
Performance on the Equivalent Equations Problem . . . . . . 267
Interpretation of a Literal Symbol . . . . . . . . . . . . . . . 270
Performance on the which Is Larger Problem . . . . . . . . . 271
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Equivalence Results . . . . . . . . . . . . . . . . . . . . . . 273
Variable Results . . . . . . . . . . . . . . . . . . . . . . . . 274
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
An Approach to Geometric and Numeric Patterning that Fosters Second

Grade Students’ Reasoning and Generalizing about Functions
and Co-variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Joan Moss and Susan London McNab
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Our Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Our Approach: Theoretical . . . . . . . . . . . . . . . . . . . 279
Instructional Sequence . . . . . . . . . . . . . . . . . . . . . . . . 281
Visual Representation: Geometric Growing Patterns . . . . . 281
xviii Contents
Numeric Representations: Function Machine . . . . . . . . . 282

Integration Activities: Pattern Sidewalk . . . . . . . . . . . . 283
Role of the Teacher . . . . . . . . . . . . . . . . . . . . . . . 284
Procedures and Measures: Grade 2 Interventions . . . . . . . . . . 285
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Finding Rules for Patterns and Generating Patterns Based
on Given Rules . . . . . . . . . . . . . . . . . . . . . 286
Constructing a Pattern from a Rule: “A ‘number times two,
plus one’ pattern?” . . . . . . . . . . . . . . . . . . . 286
Finding a Rule for a Given Pattern: “Position number times
three, plus one” . . . . . . . . . . . . . . . . . . . . . 287
Students’ Invention of Multiplication . . . . . . . . . . . . . 288
Deconstructing Multiplication: “Double the position, plus
the position” . . . . . . . . . . . . . . . . . . . . . . 289
Using a Structural Understanding of Multiplication to Predict
Far Positions: “It’s 40 up, and 3 to the side” . . . . . . 289
The Discovery of Zero . . . . . . . . . . . . . . . . . . . . . 291
Zero as a Coefficient: “Zero groups of 4 million is zero” . . . 291
Zero as a Position Number: “the zero-th position” . . . . . . . 292
Transfer of Structure . . . . . . . . . . . . . . . . . . . . . . 293
Circumventing Whole Object Reasoning . . . . . . . . . . . 293
Informal Algebraic Expressions of Rules in the Sparky
Problem . . . . . . . . . . . . . . . . . . . . . . . . . 294
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
The Curriculum with Its Focus on Integration . . . . . . . . . 296
Prioritizing Visual Representations of Pattern . . . . . . . . . 297
Pedagogy and Student Inventions . . . . . . . . . . . . . . . 297
Concluding Thoughts . . . . . . . . . . . . . . . . . . . . . . . . . 298
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
Grade 2 Students’ Non-Symbolic Algebraic Thinking . . . . . . . . . . . 303

Luis Radford
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Extending Sequences . . . . . . . . . . . . . . . . . . . . . . . . . 305
Abstraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
The Boundaries of Arithmetic and Algebraic Thinking . . . . . . . 308
Layers of Generality . . . . . . . . . . . . . . . . . . . . . . . . . 311
Beyond Intuited Indeterminacy . . . . . . . . . . . . . . . . . . . . 312
A General Overview . . . . . . . . . . . . . . . . . . . . . . . . . 316
Synthesis and Concluding Remarks . . . . . . . . . . . . . . . . . 317
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
Formation of Pattern Generalization Involving Linear Figural Patterns

Among Middle School Students: Results of a Three-Year Study . . . 323
F.D. Rivera and Joanne Rossi Becker
Anticipating What Is to Come: Initial Reflections on Our
Three-Year Data from the Clinical Interviews . . . . . . . . . 327
Contents xix
Cognitive Issues Surrounding Pattern Generalization: What We

Know from Various Theoretical Perspectives and Empirical
Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Clarifying the Definition of Pattern Generalization . . . . . . 329
Types of Algebraic Generalization Involving Figural Patterns 330
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
Classroom Contexts from Years 1 to 3 of the Study . . . . . . 331
Nature and Content of Classroom Teaching Experiments
in Years 1 and 2 . . . . . . . . . . . . . . . . . . . . 332
Nature and Content of Classroom Teaching Experiments
in Year 3 . . . . . . . . . . . . . . . . . . . . . . . . 334
Nature and Content of Clinical Interview Tasks from Years 1
to 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
Data Collection and Analysis and Relevant Study Protocols . 335
Findings and Discussion Part 1: Accounting for Constructive
and Deconstructive Generalizations . . . . . . . . . . . . . . 338
Findings and Discussion Part 2: Understanding the Operations
Needed in Developing a Pattern Generalization . . . . . . . . 342
Findings and Discussion Part 3: Factors Affecting Students’
Ability to Develop CGs . . . . . . . . . . . . . . . . . . . . 344
Findings and Discussion Part 4: A Three-Year Account
of Classroom Mathematical Practices that Encouraged
the Formation of Generalization Among Our Middle School
Students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Year 1 Classroom Practices: From Figurally- to
Numerically-Driven CSGs . . . . . . . . . . . . . . . 348
Year 2 Practice: Continued Use of Numerically-Driven CSGs
and a Refinement in the Case of Decreasing Linear
Patterns . . . . . . . . . . . . . . . . . . . . . . . . . 351
Year 3 Practices: A Third Shift Back to Figural-based
Generalization and the Consequent Occurrence
of CSGs, CNGs, and DGs . . . . . . . . . . . . . . . 352
Findings and Discussion Part 5: Middle School Students’
Capability in Justifying CSGs . . . . . . . . . . . . . . . . . 354
Capability in Constructing and Justifying CNGs and DGs . . 357
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
Commentary on Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

Bharath Sriraman and Kyeong-Hwa Lee
Introductory Remarks . . . . . . . . . . . . . . . . . . . . . . . . 367
Early Algebraization Versus Meaningful Arithmetic . . . . . . . . 368
Generalized Arithmetic, Generalizing, Generalization . . . . . . . . 369
From Haeckel to Lamarck to Early Algebraization . . . . . . . . . 370
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
xx Contents
Part III: Instructional Perspective
Preface to Part III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

Prospective Middle-School Mathematics Teachers’ Knowledge

of Equations and Inequalities . . . . . . . . . . . . . . . . . . . . . . 379
Nerida F. Ellerton and M.A. (Ken) Clements
The Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
Mathematical Considerations Relating to the Teaching
and Learning of Equations and Inequalities . . . . . . . . . . 380
Student Misconceptions in Regard to Quadratic Equations . . 383
Student Misconceptions with Regard to Linear Inequalities . . 384
The Pre-Service Teachers Involved, and Tasks Used, in the Present
Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
“Clever” Tasks . . . . . . . . . . . . . . . . . . . . . . . . . 387
Developing the Pencil-and-Paper Instruments . . . . . . . . . . . . 389
The Eight Equation/Algebraic Inequality Pairs . . . . . . . . 389
Study Design, and Results . . . . . . . . . . . . . . . . . . . . . . 395
Population and Sample Considerations . . . . . . . . . . . . 395
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Conclusions in Relation to the Prospective Teachers’ Knowledge
of Algebraic Inequalities . . . . . . . . . . . . . . . . . . . . 399
Prospective Teachers’ Knowledge in Relation to Quadratic
Equations . . . . . . . . . . . . . . . . . . . . . . . . 401
Bad News, Good News and Some Concluding Comments . . . . . 402
Bad News . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Good News . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
Student Confidence Considerations . . . . . . . . . . . . . . 406
Concluding Comments . . . . . . . . . . . . . . . . . . . . . 406
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
The Algebraic Nature of Fractions: Developing Relational Thinking

in Elementary School . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Susan B. Empson, Linda Levi, and Thomas P. Carpenter
What Is Relational Thinking? . . . . . . . . . . . . . . . . . . . . 411
Use of Relational Thinking in Learning Fractions . . . . . . . . . . 413
Understanding Fractional Quantities Through Relational
Thinking . . . . . . . . . . . . . . . . . . . . . . . . 413
Use of Relational Thinking to Make Sense of Operations
Involving Fractions . . . . . . . . . . . . . . . . . . . 416
Discussion of Cases . . . . . . . . . . . . . . . . . . . . . . 422
A Conjecture Concerning Relational Thinking as a Tool in
Learning New Number Content . . . . . . . . . . . . . . . . 423
Contents xxi
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
Professional Development to Support Students’ Algebraic Reasoning:

An Example from the Problem-Solving Cycle Model . . . . . . . . . 429
Karen Koellner, Jennifer Jacobs, Hilda Borko, Sarah Roberts, and Craig
Schneider
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
The Problem-Solving Cycle Model of Professional Development . . 431
The PSC as Implemented in the STAAR Project . . . . . . . 432
Prior Research on the Development and Impact of the PSC . . 435
Impact of the PSC on Instructional Practice: A Case Study
Analysis . . . . . . . . . . . . . . . . . . . . . . . . 436
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Ken Bryant . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . 437
Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . 440
Patterns Drawn from QMI Coding and Analysis . . . . . . . . 440
Vignette Analysis: Ken’s Skyscraper Windows Lesson . . . . 447
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Using Habermas’ Theory of Rationality to Gain Insight into Students’

Understanding of Algebraic Language . . . . . . . . . . . . . . . . . 453
Francesca Morselli and Paolo Boero
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Habermas’ Construct of Rational Behaviour . . . . . . . . . . . . . 454
Adaptation of Habermas’ Construct of Rational Behavior
to the Case of the Use of Algebraic Language . . . . . . . . . 455
Epistemic Rationality . . . . . . . . . . . . . . . . . . . . . . 455
Teleological Rationality . . . . . . . . . . . . . . . . . . . . 456
Communicative Rationality . . . . . . . . . . . . . . . . . . 456
Relationships with Other Studies on Proving and Modeling
and on the Teaching and Learning of Algebra . . . . . . . . . 457
Proving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Teaching and Learning of Algebra . . . . . . . . . . . . . . . 459
Description and Interpretation of Student Behavior . . . . . . . . . 462
Habermas’ Analytical Tool: Examples of Analysis of Student
Behavior at Different School Levels . . . . . . . . . . 462
Habermas Analytical Tool: Analysis of a Teaching Experiment . . . 468
The Context of the Study: Description of the Research Project 468
First Task: Choose a Number. . . . . . . . . . . . . . . . . . . 469
Second Task: Representing the Game . . . . . . . . . . . . . 470
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
xxii Contents
Research Advances . . . . . . . . . . . . . . . . . . . . . . . 477

Educational Implications . . . . . . . . . . . . . . . . . . . . 478
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
Theoretical Issues and Educational Strategies for Encouraging Teachers

to Promote a Linguistic and Metacognitive Approach to Early
Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Annalisa Cusi, Nicolina A. Malara, and Giancarlo Navarra
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
In Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
From Traditional Algebra to Early Algebra . . . . . . . . . . 485
Early Algebra as a Meta-Subject and the ArAl Project . . . . . . . . 486
Socio-Constructive Teaching and Teacher Training . . . . . . . . . 487
The Role of the Teacher’s Reflection . . . . . . . . . . . . . . 488
The Role of the ArAl Glossary in Teacher Training . . . . . . . . . 490
Algebraic Babbling . . . . . . . . . . . . . . . . . . . . . . . 492
Algebraic Babbling → Algebra as a Language . . . . . . . . 493
Algebraic Babbling → Syntax, Semantics → Brioshi . . . . . 494
Brioshi → Canonical/Non Canonical form of a
Number → ‘=’ . . . . . . . . . . . . . . . . . . . . . 495
The Multi-Commented Transcripts Methodology (MCTM) . . . . . 496
From the Comments to a Classification of Attitudes . . . . . . . . . 499
Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
A Procedural Focus and a Relationship Focus to Algebra: How U.S.

Teachers and Japanese Teachers Treat Systems of Equations . . . . . 511
Margaret Smith
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
Algebraic Reasoning . . . . . . . . . . . . . . . . . . . . . . 512
TIMSS Video Studies . . . . . . . . . . . . . . . . . . . . . 514
Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Two Teachers’ Lessons . . . . . . . . . . . . . . . . . . . . . . . . 516
Discussion of Key Differences . . . . . . . . . . . . . . . . . . . . 516
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
Teaching Algebraic Equations with Variation in Chinese Classroom . . . 529

Jing Li, Aihui Peng, and Naiqing Song
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
The Source of the Data . . . . . . . . . . . . . . . . . . . . . . . . 531
Theoretical Framework . . . . . . . . . . . . . . . . . . . . . . . . 531
The Method of Research . . . . . . . . . . . . . . . . . . . . . . . 533
Analysis of Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
The Introduction of the Concept of Equation . . . . . . . . . 533
Contents xxiii
The Improvement of Understanding of Equation . . . . . . . 535

Equations Solving . . . . . . . . . . . . . . . . . . . . . . . 539
The Application of Equations . . . . . . . . . . . . . . . . . 541
Discussion and Conclusion . . . . . . . . . . . . . . . . . . . . . . 545
Process of Teaching Algebra with Variation . . . . . . . . . . 545
Operation of Teaching Algebra with Variation . . . . . . . . . 546
Final Comments . . . . . . . . . . . . . . . . . . . . . . . . 548
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
Commentary on Part III . . . . . . . . . . . . . . . . . . . . . . . . . . . 557

John Mason
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557
Systematics: Structure of Activity . . . . . . . . . . . . . . . . . . 558
What Is Algebra? . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
What Is and What Could Be: Teaching Algebra as an Activity . . . 560
Traditional Algebra Teaching . . . . . . . . . . . . . . . . . 561
Envisioned Algebra Teaching . . . . . . . . . . . . . . . . . 563
What Makes ‘Algebra’ Early? . . . . . . . . . . . . . . . . . 566
Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . 568
Transforming Algebra Teaching and Learning as an Activity . . . . 568
How Can Locally Successful Teaching Be Engineered for All? 569
What Is and Could Be Researched? . . . . . . . . . . . . . . . . . 570
What Is Really Researched? . . . . . . . . . . . . . . . . . . 571
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
Overall Commentary on Early Algebraization: Perspectives for

Research and Teaching . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Carolyn Kieran
Shaping the Notion of Algebraic Thinking within Early Algebra . . 580
Thinking about the General in the Particular . . . . . . . . . . . . . 581
Thinking Rule-Wise about Patterns . . . . . . . . . . . . . . . . . 582
Thinking Relationally about Quantity, Number, and Numerical
Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
Thinking Representationally about the Relations in Problem
Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
Thinking Conceptually about the Procedural . . . . . . . . . . . . 586
Anticipating, Conjecturing, and Justifying . . . . . . . . . . . . . . 588
Gesturing, Visualizing, and Languaging . . . . . . . . . . . . . . . 590
The View of Algebraic Thinking that Emerges from this Volume . . 591
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
Editors and Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . 615

Part I
Curricular Perspective
Preface to Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Jinfa Cai
Dept. Mathematical Sciences, University of Delaware, Ewing Hall 501,
Newark, DE, 19716 USA
Eric Knuth
University of Wisconsin-Madison, Teacher Education Building, 225 N.
Mills street, Madison, WI 53706, USA
Functional Thinking as a Route Into Algebra in the Elementary Grades . 5
Maria L. Blanton
STEM Department, University of Massachussetts at Dartmouth,
Dartmouth, USA
James J. Kaput
Developing Students’ Algebraic Thinking in Earlier Grades: Lessons
from China and Singapore . . . . . . . . . . . . . . . . . . . . . . . . 25
Jinfa Cai
Department of Mathematical Sciences, University of Delaware,
Newark, USA
Swee Fong Ng
National Institute of Education, Nanyang Technological University,
Singapore, Singapore
John C. Moyer
Department of Mathematics, Statistics and Computer Science,
Marquette University, Milwaukee, USA
Developing Algebraic Thinking in the Context of Arithmetic . . . . . . . 43
Susan Jo Russell
Education Research Collaborative, TERC, Cambridge, USA
Deborah Schifter
Education Development Center, Newton, USA
Virginia Bastable
SummerMath for Teachers, Mount Holyoke College, South Hadley, USA
2 Curricular Perspective
The Role of Theoretical Analysis in Developing Algebraic Thinking:

A Vygotskian Perspective . . . . . . . . . . . . . . . . . . . . . . . . 71
Jean Schmittau
School of Education, State University of New York at Binghamton,
Binghamton, USA

K. Subramaniam
Homi Bhabha Centre for Science Education, Tata Institute of
Fundamental Research, Mumbai, India
Rakhi Banerjee
School of Social Science, Tata Institute of Social Sciences, Mumbai, India
Shiki: A Critical Foundation for School Algebra in Japanese Elementary

School Mathematics . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Tad Watanabe
Department of Mathematics & Statistics, Kennesaw State University,
Kennesaw, GA, USA
Commentary on Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Jeremy Kilpatrick
Department of Mathematics and Science Education, University of
Georgia, Athens, USA
Preface to Part I
Jinfa Cai and Eric Knuth
Although it is widely accepted that we should expect students in early grades to

think algebraically, the real question is how can we prepare students in earlier grades
to think algebraically? As we discussed earlier in this volume, developing algebraic
ideas in earlier grades is not simply a matter of moving aspects of the traditional
secondary school algebra curriculum down into the elementary school mathematics
curriculum. Rather, developing algebraic ideas in the earlier grades requires funda-
mentally reforming how arithmetic should be taught as well as a identifying better
ways to develop algebraic thinking from traditional arithmetic topics. In the chapters
that follow, the authors focus on various aspects of curricula from different countries
(including China, India, Japan, Russia, Singapore, and the United States) in order to
examine how curriculum might be designed and delivered to help students develop
algebraic habits of mind.
The chapters in this part provide illustrative examples of successful efforts to help
students see algebra in the context of arithmetic. The chapters by Blanton and Rus-
sell et al. examine how instructional materials and school activities can be extended
to support students’ algebraic thinking. In Russell et al.’s chapter, they highlight
four mathematical activities that underlie arithmetic and algebra, and discuss how
these activities serve as an important bridge between the two domains. In Blanton’s
chapter, she proposes that elementary school mathematics should include curricu-
lum and instruction that deliberately attends to how two or more quantities vary in
relation to each other.
In the chapter by Cai et al., the authors analyzed Chinese and Singaporean cur-
ricula. They found that the Chinese and Singaporean curricula could be useful ref-
J. Cai ()
Dept. Mathematical Sciences, University of Delaware, Ewing Hall 501, Newark, DE, 19716 USA
e-mail: [email protected]
E. Knuth
University of Wisconsin-Madison, Teacher Education Building, 225 N. Mills street, Madison,
WI 53706, USA
J. Cai, E. Knuth (eds.), Early Algebraization, Advances in Mathematics Education, 3

DOI 10.1007/978-3-642-17735-4_1, © Springer-Verlag Berlin Heidelberg 2011
4 J. Cai and E. Knuth
erences for those wishing to help elementary students develop a stronger sense of
the connections between arithmetic and algebra. Specifically, the Chinese and Sin-
gaporean curricula provide concrete examples of promising ways to integrate arith-
metic and algebraic ideas in the earlier grades.
In the mathematics education research community, we know little about mathe-
matics education in India. Thus, we are especially pleased to have a chapter from In-
dia by Subramaniam and Banerjee. The authors shared India’s version of addressing
the connection between arithmetic and algebra. In India, algebra is seen as founda-
tional to arithmetic rather than as a generalization of arithmetic. Subramaniam and
Banerjee present a framework that illuminates the arithmetic-algebra connection
from an Indian perspective.
Schmittau’s chapter presents a description of a curricular approach to elementary
school mathematics based on the work of Russian psychologists Lev Vygotsky and
V.V. Davydov. In contrast to curricular approaches that view number as foundational
in children’s early mathematical development, the curricular approach Schmittau
describes views algebraic structure as foundational. In this approach, “traditional”
arithmetic understanding is developed as students apply their algebraic understand-
ing to concrete numerical instances. As Schmittau points out, this approach is very
different from recent reform approaches that seek to introduce elements of algebra
into the study of arithmetic.
In Watanabe’s chapter, he presents an analysis of the Japanese elementary school
(Grades 1 through 6) mathematics curriculum materials and notes that the study
of functional relationships (patterns) is a major emphasis in Japanese elementary
schools. However, the Japanese curriculum considers the ideas related to mathemat-
ical expressions, called “shiki” in Japanese, as a pillar of elementary school algebra.
In summary, the chapters that comprise this part highlight various curricular ap-
proaches, experiences, and practices that illustrate ways in which students in earlier
grades can be better prepared to think algebraically.
Functional Thinking as a Route Into Algebra
in the Elementary Grades
Maria L. Blanton and James J. Kaput
Abstract This chapter explores how elementary teachers can use functional think-
ing to build algebraic reasoning into curriculum and instruction. In particular, we
examine how children think about functions and how instructional materials and
school activities can be extended to support students’ functional thinking. Data are
taken from a five-year research and professional development project conducted in
an urban school district and from a graduate course for elementary teachers taught
by the first author. We propose that elementary grades mathematics should, from
the start of formal schooling, extend beyond the fairly common focus on recursive
patterning to include curriculum and instruction that deliberately attends to how two
or more quantities vary in relation to each other. We discuss how teachers can trans-
form and extend their current resources so that arithmetic content can provide oppor-
tunities for pattern building, conjecturing, generalizing, and justifying mathematical
relationships between quantities, and we examine how teachers might embed this
mathematics within the kinds of socio-mathematical norms that help children build
mathematical generality.
The research reported here was supported in part by a grant from the U.S. Department of
Education, Office of Educational Research and Improvement to the National Center for
Improving Student Learning and Achievement in Mathematics and Science (R305A600007-98).
The opinions expressed herein do not necessarily reflect the position, policy, or endorsement of
supporting agencies. This chapter is a revised version of an article published in
ZDM—International Reviews on Mathematical Education, 37(1), 34–42. DOI
10.1007/BF02655895.
J.J. Kaput is deceased.
M.L. Blanton ()
STEM Department, University of Massachussetts at Dartmouth, Dartmouth, USA

6 M.L. Blanton and J.J. Kaput
Introduction
Current research is redefining what we understand about the kind of mathematics
that young children can and should learn (National Research Council [NRC] 2001).
Consider the following Towers of Cubes problem (see Fig. 1) taken from the Na-
tional Council of Teachers of Mathematics [NCTM] Principles and Standards for
School Mathematics (2000, p. 160):
What is the surface area of each tower of cubes (include the bottom)? As the tower gets taller, how
does the surface area change? What is the surface area of a tower with fifty cubes?
Fig. 1 Towers of cubes
In the not so distant past, such a problem was mostly absent from typical ele-
mentary school1 curricula and instruction in the United States. While it might have
appeared as an enrichment task, it was likely marginalized by the press towards
computational skills (Thompson et al. 1994) and procedures that children were (and
are) compelled to memorize as a signal of their readiness for higher mathematical
thinking. Or, it might have appeared in an abbreviated, arithmetic form as “What is
the surface area of a tower built of 3 cubes?” However, algebraic reasoning as an
activity of generalizing mathematical ideas, using literal symbolic representations,
and representing functional relationships, all implicit in this task, is no longer re-
served for secondary grades and beyond, but is an increasingly common thread in
the fabric of ideas that constitute mathematical thinking at the elementary grades.
The Challenge of Curriculum and Instruction
Simply put, young children today need to learn a different kind of mathematics than
their parents learned. Some argue that they need to be “algebra ready” (e.g., Na-
tional Mathematics Advisory Panel 2008). But what experiences make them ready
for algebra, and for what kind of algebra are they being made ready? Romberg and
Kaput (1999) maintain that understanding the increasingly complex mathematics of
the 21st century will require children to have a type of elementary school experi-
ence that goes beyond arithmetic and computational fluency to attend to the deeper
underlying structure of mathematics. It will require experiences that help children
learn to recognize and articulate mathematical structure and relationships and to
1 Elementary school refers here to grades PreK-5.

Functional Thinking as a Route Into Algebra in the Elementary Grades 7
use these insights of mathematical reasoning as objects for mathematical reasoning.

This type of elementary school experience has come to be embodied in what many
refer to as early algebra,2 and because its underlying purpose is to deepen chil-
dren’s understanding of the structural form and generality of mathematics and not
just provide isolated experiences in computation, scholars increasingly agree that it
is the avenue through which young children can become mathematically successful
in later grades. Thus, our perspective on “algebra readiness” is that experiences in
building, expressing, and justifying mathematical generalizations—for us, the heart
of algebra and algebraic thinking—should be a seamless process that begins at the
start of formal schooling, not content for later grades for which elementary school
children are “made ready” through a singular, myopic focus on arithmetic.
But changing the mathematics elementary school children learn—their daily cur-
riculum—is only part of the solution. As Blanton and Kaput note, “most elementary
teachers have little experience with the kinds of algebraic thinking that need to be-
come the norm in schools and, instead, are often products of the type of school
mathematics instruction that we need to replace” (2005). However, these very teach-
ers are central to reforms in children’s school mathematical experiences. Moreover,
the instructional materials in most elementary schools today are basal texts, and
even newer, standards-based materials are just beginning to incorporate systematic
approaches to the development of algebraic reasoning (Kaput and Blanton 2005).
These constraints represent the challenge of building algebraic thinking into cur-
riculum and instruction.
There are two issues implicit in the above discussion that this article aims to
address: (1) how opportunities for algebraic thinking can be integrated into the el-
ementary grades to prepare students for more powerful mathematics in later years,
and (2) how elementary teachers can transform their own resources and instruction
in ways that effect (1).
Functional Thinking as a Route to Algebraic Thinking
Early algebra can occur in several interrelated forms in the classroom.3 We focus
here on functional thinking as a strand by which teachers can build generality into
their curriculum and instruction. We broadly conceptualize functional thinking to
2 While there are multiple perspectives on early algebra, Lins and Kaput (2004) describe a gen-
eral agreement among scholars that it involves “acts of deliberate generalization and expression of
generality. . . [and] reasoning based on the forms of syntactically guided actions on those expres-
sions”.
3 Kaput (2008) characterizes algebraic thinking as consisting of two core aspects: (1) making and
expressing generalizations in increasingly formal and conventional symbol systems, and (2) rea-
soning with symbolic forms, including the syntactically guided manipulations of those symbolic
forms. In turn, he argues that these core aspects cut across three longitudinal strands of school
algebra: (1) Algebra as the study of structures and systems abstracted from computations and rela-
tions (e.g., algebra as generalized arithmetic); (2) Algebra as the study of functions, relations, and
joint variation; and (3) Algebra as the application of a cluster of modeling languages to express
and support reasoning about situations being modeled.
incorporate building and generalizing patterns and relationships using diverse lin-
guistic and representational tools and treating generalized relationships, or func-
tions, that result as mathematical objects useful in their own right. As the NCTM
Principles and Standards (2000, p. 37) argues, children in the elementary grades
should be able to
(1) Understand patterns, relations, and functions;
(2) Represent and analyze mathematical situations and structures using algebraic
symbols;
(3) Use mathematical models to represent and understand quantitative relation-
ships; and
(4) Analyze change in various contexts.
In addition, we use here three modes of analyzing patterns and relationships, out-
lined by Smith (2008), as a framework to discuss the kinds of functional thinking
found in classroom data: (1) recursive patterning involves finding variation within a
sequence of values; (2) covariational thinking is based on analyzing how two quan-
tities vary simultaneously and keeping that change as an explicit, dynamic part of a
function’s description (e.g., “as x increases by one, y increases by three”) (Confrey
and Smith 1991); and (3) a correspondence relationship is based on identifying a
correlation between variables (e.g., “y is 3 times x plus 2”).
In what follows, we draw on data from a five-year research and professional de-
velopment project in an urban school district (Kaput and Blanton 2005) and a subse-
quent graduate course for elementary teachers, taught by the first author, to examine
how children think about functional relationships, its mathematical implications for
later grades, and how instructional materials and school activities can be deepened
and extended to support the development of functional thinking in the elementary
grades.
Functional Thinking in the Elementary Grades

The idea of function has, for over a century, been regarded by mathematicians as a
powerful, unifying idea in mathematics that merits a central place in the curriculum
(Freudenthal 1982; Hamley 1934; Schwartz 1990). Indeed, the idea can be traced
back to Leibniz (Boyer 1946). However, until very recently, the study of functions
has been treated largely in the US as something to be learned in high school algebra,
or even middle school mathematics. The perspective taken here is that the study of
functions should be treated longitudinally and in its full richness beginning in early
elementary school (NCTM 2000; Smith 2003).
But what capacity do young children have for functional thinking? Even though
elementary school mathematics has more recently included recursive patterning, it
has not attended pervasively to covariation or correspondence in functional think-
ing, especially in grades PreK-2. For instance, even NCTM (2000) suggests that, as
late as fourth-grade, students might find a recursive pattern in the Towers of Cubes
problem (see Fig. 1), and not until fifth-grade would they develop a correspondence
relationship. Can elementary students, in fact, make the conceptual shift from simple
recursive patterning to account for simultaneous changes in two or more variables?
Moreover, at what grades can they do this? And can they, or in what ways can they,
symbolize and operate on covariational or correspondence relationships in data?
Children’s Capacity for Functional Thinking
Current research challenges the developmental constraints traditionally placed on

young learners and their capacity for functional thinking. For example, researchers
have found that elementary school children can develop and use a variety of rep-
resentational tools to reason about functions, they can describe in words and sym-
bols recursive, covarying, and correspondence relationships in data, and they can
use symbolic language to model and solve equations with unknown quantities (e.g.,
Blanton 2008; Brizuela and Schliemann 2003; Brizuela et al. 2000; Carraher et al.
2008; Kaput and Blanton 2005; Moss et al. 2008; Schliemann and Carraher 2002;
Schliemann et al. 2001).
While much of this research focuses on functional thinking in grades 3–5, we
have found that students are not only capable of deeper functional analysis than pre-
viously thought, but that the genesis of these ideas appear at grades earlier than typi-
cally expected. In particular, we have found that the types of representations students
use, the progression of mathematical language in their descriptions of functional re-
lationships, the ways students track and organize data, the mathematical operations
they use to interpret functional relationships, and how they express covariation and
correspondence among quantities, can be scaffolded in instruction beginning with
the very earliest grades, at the start of formal schooling (Blanton and Kaput 2004).
The following discussion draws on our research data to elaborate these capacities
in children’s functional thinking across elementary grades. We note that the data in-
cluded here are intended to convey existence proofs of what is possible in children’s
thinking; our goal is not to examine the regularity with which functional thinking
occurred in instruction.
The Development of Representational Infrastructure: Children’s Use

of Function Tables
Research, including early algebra research, suggests that students’ flexibility with
multiple representations both reflects and promotes deeper mathematical insights
(Behr et al. 1983; Brizuela and Earnest 2008; Goldin and Shteingold 2001). Brizuela
and Earnest note that “the connections between different representations help to
resolve some of the ambiguity of isolated representations, [so] in order for concepts
to be fully developed, children will need to represent them in various ways”.
We found that teachers across the elementary grades were able to scaffold chil-
dren’s use of tables, graphs, pictures, words and symbols in gradually more sophis-
ticated ways in order for them to make sense of data and interpret functional rela-
tionships (Blanton and Kaput 2004). For example, while students in grades PreK-1
Fig. 2 Kindergarten
students’ representation for
the numbers of eyes and eyes
and tails for two dogs
Fig. 3 A first-grader’s t-chart

for the Handshake Problem5
relied on counting visible objects or hand-written marks and registering their counts
through inscriptions in t-charts4 or through dots and hatch marks for eyes and tails
(see Figs. 2 and 3), by second and third grade, students could routinely operate on
data that had no iconic or tangible counterpart (e.g., tracking the number of eyes
on ten dogs without pictures or physical objects). Moreover, while grades PreK-1
teachers typically led students in developing t-charts to organize their data, the re-
sponsibility for this began to shift to students during first grade. Figure 3 shows a
t-chart, constructed by a first-grader, which records the total number of handshakes
in a group of varying size (Blanton 2008).
We have found that the t-chart, or function table, becomes an important structure
in children’s mathematical reasoning. In the earlier grades (PreK-1), it provided a
context to re-represent marks with numerals as children worked on the arithmetic of
correspondence between quantity and numeral. But its introduction in these grades
4 T-charts are teacher-termed function tables with a column of data for the independent variable
followed by a column of data for the dependent variable.

5 The Handshake Problem can be stated as follows: If 3 people are in a group, how many total
handshakes would there be if every person shook hands with all people in the group once? How
many handshakes would there be if there were 4 people in the group? Five people? Six people?
Twenty people? Can you find a relationship between the number of people in the group and the
total number of handshakes?
Fig. 4 The growing snake
as a tool for organizing covarying data also initiated its transformation from opaque
to transparent object (Kaput 1995) in children’s functional thinking as a represen-
tation that one could “look through” to “see” new relationships. The first grader’s
analysis of differences in the numbers of handshakes in Fig. 3 illustrates that as early
as first grade, students can begin to transition beyond an understanding of t-charts
as opaque objects—a place to record numbers—to a transparent object that can be
used to determine relationships in data. We maintain that introducing such repre-
sentational tools from the start of formal schooling can help spread the cognitive
load across grades in a way that allows students in second and third grades (and
beyond) to focus on more difficult tasks such as symbolizing correspondence and
covariational relationships.
By second and third grades, we have found that students are able to use this
tool transparently, as a mathematical object, in thinking about data. The following
teacher narrative illustrates this algebraic reasoning with third-graders. The third-
grade teacher who authored the narrative, Mrs. Gardiner, had designed a task in
which students were to find the number of body parts a growing snake would have
on day 10 and on day n, where each triangle equaled a body part. She drew the
growing snake on the board for Days 1, 2, and 3 (see Fig. 4).
The class worked on this problem for approximately 10 minutes. All organized their data
with a t-chart. When I pulled the group together to discuss the problem, it was Karlie6 who
had her hand waving hard. . . . Karlie usually just sits and listens during math time, so her
enthusiasm was very special. I called on her right away. ‘I know that on day 10 the snake
will have 101 body parts and I know that on day n the snake will have n × n + 1. I know
this because I used my t-chart and I looked for the relationship between n and body parts.
This is the first time I saw the pattern, so please tell me I’m right!’ she said excitedly. . . .
The class had all come to pretty much the same answer.
This suggests to us that the t-chart helped structure Karlie’s thinking about re-
lationships between quantities. Unlike in the earlier elementary grades, where stu-
dents were more likely to use t-charts opaquely as a storage system for numbers
and were not yet able to attend to the meanings embedded in how data were po-
sitioned in the chart, the t-chart became the object, or tool, by which Karlie could
compare data and find relationships. She was able to attend to how numbers were
located in the chart and see through it to the relationships it made available to her.
In this sense, we maintain that the t-chart had become transparent in how she used
it to think about functions. Our point is that critical instruction in the earlier grades
(PreK-1) can initiate the transition of representational tools from opaque to trans-
parent objects in children’s thinking so that children are able to shift their attention
to more complex tasks in later elementary grades and beyond. This is exactly how
6 All student names are pseudonyms.

mathematics has grown in power historically, as new representation systems were

developed (including that of algebra itself) to increase the power of human thinking.
The Development of Students’ Symbol Sense
One particularly vital aspect of early algebra is the transition from natural language
to symbolic notational systems. If one’s perspective is that development precedes
learning, then the use of symbols as variables in elementary grades is, perhaps, not
without controversy. However, we take the view here that learning promotes de-
velopment and that it entails a pseudo-conceptual stage of concept formation in
students’ development of symbol sense. In describing the development of higher
mental functioning in children, Vygotsky (1962) identified the notion of a pseudo-
concept as an essential bridge in children’s thinking to the final stage of concept
formation. While the pseudo-concept a child possesses is phenotypically equivalent
to that of an adult, it is psychologically different. As a result, the child is able to
“operate with [the concept], to practice conceptual thinking, before he is clearly
aware of the nature of these operations” (p. 69). This suggests that learning to think
mathematically involves the acquisition of notational tools that are within the child’s
zone of proximal development, but not entirely owned by the child. In essence, it in-
volves students’ transition from an opaque to transparent use of symbols. Moreover,
the dialectic between thought and language in learning (Vygotsky 1962) implies
that symbolic notational systems are more fully conceptually formed in children’s
thinking as a result of children’s interaction with them in meaningful contexts. In
short, children can develop symbol sense as they have opportunity to use symbolic
notation in meaningful ways (see also Brizuela et al. 2000).
We have found that, when curriculum and instruction provide opportunity for
thinking about functional relationships, students can transition linguistically from
iconic and natural language registers at grades PreK-1 to symbolic notational sys-
tems by grade 3 (Blanton and Kaput 2004). A first grade teacher described how one
of her students made this transition while thinking about the Handshake Problem:
I asked, ‘Can I label one side [of the t-chart] ‘people’ and the other side ‘handshake’?’ One
little boy said, ‘Just write ‘p’ and ‘h’.’ I immediately stopped what I was doing. I asked,
‘What did you say?’ He continued to repeat what I heard him say. ‘Awesome, how did you
come up with that?’ I probed. He continued, ‘Well, ‘people’ begins with p and ‘handshakes’
begin with h.’ (Blanton 2008, p. 43)
While this student’s understanding of variable is certainly in its early stages (for
example, care must be taken to ensure that the student does not confuse the variable
as representing the object and not the quantity), the point here is that giving children
opportunities to begin using symbolic representations can occur as early as first
grade, and acquiring these more basic ideas in the early grades allows them greater
cognitive room to explore more complex ideas in later elementary grades.
By third-grade, students can move beyond this more primitive act of symbolizing
to describe and discuss functional relationships. We include the following teacher
narrative to illustrate third-grade students using symbolic notation to think about the
number of circle-shaped body parts on a growing caterpillar. In this vignette, Mrs.

Gardiner has just described the Growing Caterpillar task to her students.7
I showed my students my caterpillar example and all I wanted them to see was how I devel-
oped the problem. I had no idea that they would begin to solve the problem. I couldn’t stop
them. There were hands going up all over the place. Everyone wanted to tell me the pattern
they saw when they looked at the growth of the caterpillar. I said, ‘Guys, I haven’t even
asked you the question yet’. ‘But I see the pattern, Mrs. Gardiner’, yelled Jak. ‘Okay, what
do you think the pattern is?’ I asked. ‘I think it is x times 2 plus one’, he said. ‘How many
of you agree with Jak?’ I questioned. ‘I don’t know. I have to do a t-chart’, explained Meg.
‘Well, then let’s do that together on the board’, I said. With the students’ help, we drew the
following t-chart on the board (see Fig. 5):
Fig. 5 T-chart for growing

caterpillar
‘Now that we have that on the board, I don’t agree with Jak’, said Meg. ‘Why is that Meg?’
I asked. ‘Because if it was x times 2 plus 1, then x would be one and y would be three. And,
it’s not. It’s x = 1 and y = 2’, she explained. . . . [If x equaled 1, then by Jak’s formula, y
would be 2(1) + 1, or 3, not 2, as the t-chart indicated.] The class struggled with the pattern
for a long time. Then Shane saw a pattern that I had not seen. He came up to the t-chart on
the board and with a red marker highlighted the pattern. It looked like this (see Fig. 6):
Fig. 6 Shawn’s Pattern for

Growing Caterpillar task
So, what Shane was saying is that if you add 1 + 2 + 2, it equals 5. If you then add 2 + 5 + 3,
it equals 10. This . . . didn’t help him find the formula, but it did help Joe! ‘I see it, I know
7 Growing Caterpillar was similar to Growing Snake except for the shape of the body parts. The
growth rates for the snake and caterpillar were the same. Mrs. Gardiner had given Growing Cater-
pillar to students two weeks prior to Growing Snake.
the formula!’ Joe cried out. ‘Well, what is it?’ I prodded. ‘It’s x × x + 1 = y’, he said. At
that moment, a loud group of ‘Oh yeah’s’ could be heard in the room. . . . I asked everyone
why this was algebra. I think Jak put it best. He said, ‘Because we have people looking for
patterns and relationships and we have them developing a formula’.
There are several points with respect to students’ use of symbols that bear men-
tioning here. First, before any data were publicly recorded and without any prompt-
ing from the teacher, Jak proposed a symbolic relationship between an arbitrary day,
x, and the number of caterpillar body parts. His spontaneous use of symbols conveys
the generality with which he was beginning to think about functional relationships.
Second, Meg was beginning to reason transparently with the t-chart and the sym-
bolic relationship conjectured by Jak in order to refute his idea (“Now that we have
that on the board, I don’t agree with Jak”). That is, implicit in her refutation was her
reasoning with both the meaning embedded in the structure of the t-chart, including
the unique roles of dependent and independent variables, as well as the symbolic
notation (“Because if it was x times 2 plus 1, then x would be one and y would
be three. And, it’s not. It’s x = 1 and y = 2”). Meg’s emerging transparent use of
symbolic notation (as well as her evident understanding of equality, another critical
issue in the development of children’s algebraic thinking) is further indicated by
her treatment of the expression ‘x times 2 plus 1’ and the dependent variable, y, as
equivalent quantities.
Because the elementary grades often incorporate meaningful imagery and con-
crete experiences to support conceptual development, they, more so than secondary
grades, can provide a rich, inquiry-based atmosphere for introducing symbolic nota-
tion. Thus, as with the development of representational infrastructure, we maintain
that instruction should begin to scaffold students’ thinking toward symbolic notation
from the start of formal schooling so that students can transition from an opaque to
transparent use of symbols as they progress through the elementary grades. Ulti-
mately, elementary students who have learned to reason symbolically in meaningful
ways will be much better prepared for the abstractions of more advanced mathemat-
ical thinking in later grades.
The Emergence of Thinking About Covariational and Correspondence

Relationships
We have found it particularly compelling that, even as early as kindergarten, chil-

dren can think about how quantities co-vary and, as early as first grade, can describe
how quantities correspond (Blanton and Kaput 2004). For instance, in the task Cut-
ting String (Blanton 2008; see also Cramer 2001), children are asked to look for a
relationship between the number of cuts on a piece of string and the resulting num-
ber of pieces of string when the string is folded in a single loop (see Fig. 7). First
graders were able to describe the relationship not only in recursive terms (“It gets
two more each time”), but also in terms of a co-varying relationship “Every time
you make one more snip it’s two more” (Blanton 2008).
Fig. 7 Folded piece of string
In a task in which students were asked to describe the total number of eyes or
the total number of eyes and tails for an increasing number of dogs, one kinder-
garten class described an additive covariational relationship between the numbers
of eyes and dogs as “every time we add one more dog we get two more eyes”. In
first and second grade, students identified a multiplicative relationship of “doubles”
and “triples” between the number of eyes and the number of eyes and tails, respec-
tively, for an increasing number of dogs. The observation that the pattern “doubles”
or “triples” suggests that students could attend to how quantities corresponded. For
example, some quantity (in particular, the independent variable) needed to be dou-
bled to get the total amount of eyes. Since data representing the total number of
eyes (i.e., 2, 4, 6, 8. . .) were not doubled to get subsequent quantities of dog eyes (4
doubled does not yield the next value of 6; 6 doubled does not yield the next value
of 8), this suggests that students were not looking for a recursive pattern such “add
2 every time” or “count by 2’s”, but a relationship between two quantities.
We recognize that some children might be responding to a known relationship
without fully understanding its functional aspect. “Doubles”, for example, is not
uncommon in the vocabulary of early grades mathematics, and to say “it doubles”
does not necessarily indicate a full conceptual understanding of correspondence or
covariation, including an explicit understanding that the value of the independent
variable is being doubled to obtain the value of the dependent variable. For some
children, “doubles” could be code for a pattern recognized as adding by two’s. How-
ever, these situations can prompt discussions that scaffold students’ thinking about
relationships between data, not just recursive patterning.
As the Growing Snake and Growing Caterpillar excerpts suggest, by third grade
students can attend to how quantities co-vary and, moreover, symbolize relation-
ships as a functional correspondence (e.g., “It’s x × x + 1 = y”). Thus, although
the data on cutting string and dog eyes and tails illustrate a simple mathematical
relationship for which some children used only natural language to describe co-
variational and correspondence relationships, we think this represents the critical
kinds of experiences that children need in the earlier elementary grades in order to
leverage deeper, more complex functional thinking in later elementary grades and
beyond.
Implications of Children’s Functional Thinking for Later Grades
The preceding discussion underscores how early algebra, and functional thinking in
particular, can nurture the development of students’ mathematical thinking in later
grades. To begin with, it can help children build critical representational and lin-
guistic tools for analyzing, describing and symbolizing patterns and relationships.
Moreover, if teachers scaffold these ideas from the start of formal schooling, these
experiences can provide a continuum of mathematical development whereby opaque

symbols and tools can be transformed into transparent objects of functional thinking.
T-charts and graphs become not just visual configurations, but structures embedded
with meaning about relationships; symbols are no longer meaningless abstract in-
scriptions, but tools by which broader ideas can be mediated and communicated.
Moreover, the elementary grades, because of its inclination towards concrete, tac-
tile, and visual experiences in learning, can bridge the expression of mathematical
ideas from natural, everyday language to symbolic notational systems in meaningful
ways. For example, students in secondary grades are often given, a priori, a sym-
bolic generalization about the commutative property of addition for real numbers
a and b (a + b = b + a). In contrast, early algebra entails exploring this property
through operations on particular numbers, then generalizing the property using ev-
eryday or symbolic language systems, where the symbolizing develops as a valid
linguistic form of expression through children’s interactions with number and oper-
ation (Carpenter et al. 2003).
All of these experiences—the development of representational and linguistic
tools, the transformation of mathematical structures and symbols from opaque to
transparent objects, and the integration of concrete, tactile, and visual experiences
to support the development of mathematics with understanding—coalesce to build
mathematical thinkers for whom abstract ideas are rooted in meaningful, concrete
events. As a result, we argue that children for whom functional thinking is a routine
part of mathematics in the elementary grades are better prepared than those who
spend the first six or seven years of formal schooling fine-tuning arithmetic skills,
procedures and facts.
Integrating Functional Thinking into Curriculum and

Instruction
While much more could be said about children’s capacity for functional thinking,
our point thus far is that young children can identify and express functional rela-
tionships in progressively more symbolic ways and that instruction in the elemen-
tary grades that nurtures this kind of thinking can support students’ mathematical
thinking in later grades. Although this suggests a mandate for change in elemen-
tary school curricula, our reality is often working with teachers who have limited
resources that, more often than not, focus on the development of children’s arith-
metic thinking. Moreover, curricular innovations alone, without the development
of teachers’ instructional and mathematical knowledge on how to build children’s
functional thinking, are not sufficient to produce real change in children’s mathe-
matical thinking. Smith (2003) notes that “elementary school teachers may create
rich classroom experiences around patterns, yet not have a sense of how this topic
ties into the ongoing mathematical development of their students, much less into
the topic of functions” (p. 136). To address this, our early algebra work with teach-
ers has involved three connected dimensions of change: (1) transforming teachers’
instructional resource base, (2) using children’s thinking to leverage teacher learn-
ing, and (3) creating classroom culture and practice to support algebraic thinking.
In what follows, we address each of these and how they support the integration of
functional thinking into curriculum and instruction.
Transforming Teachers’ Resource Base to Support Students’

Functional Thinking
In spite of limited resources or the lack of materials that integrate functional think-
ing in viable ways, elementary teachers can transform their existing instructional
resource base to include the exploration of covariational and correspondence rela-
tionships. Our approach with teachers is to help them deliberately transform single-
numerical-answer arithmetic problems to opportunities for pattern building, conjec-
turing, generalizing, and justifying mathematical relationships by varying the given
parameters of a problem (Blanton and Kaput 2003). This is easily done with tasks
such as the Telephone Problem, which might typically be posed as an arithmetic
task with a single numerical answer:
How many telephone calls could be made among 5 friends if each person spoke
with each friend exactly once on the telephone?
Stated this way, students simply need to compute a sum, although they might first
draw a picture or diagram to keep track of the phone calls. Functional thinking can
be introduced into the task by varying the number of friends in the group:
How many telephone calls would there be if there were 6 friends? Seven friends?
Eight friends? Twenty friends? One hundred friends? Organize your data in a ta-
ble. Describe any relationship you see between the number of phone calls and the
number of friends in the group. How many phone calls would there be for n friends?
The tasks included here (e.g., Growing Snake, Growing Caterpillar, Towers of
Cubes) are examples of this type of transformation; all are derived from single-
numerical-answer tasks. For example, Towers of Cubes can be seen as an extension
of the arithmetic problem “What is the surface area of a tower built of 3 one-inch
cubes?” Similarly, Growing Snake can be seen as an extension of an arithmetic task
in which students count the total number of body parts for a particular snake.
Varying Task Parameters Introduces Algebraic Thinking into the Curriculum
But how does this transformation lead to algebraic thinking or, specifically, func-
tional thinking? First, varying a problem parameter enables students to generate a
set of data that has a mathematical relationship, and using sufficiently large quan-
tities for that parameter leads to the algebraic use of number. For example, in the
Telephone Problem, finding the number of phone calls for a group whose size is
large enough so that children cannot (or would not want to) model the problem and
write down a corresponding sum to compute requires children to think about the
structure in the numbers and how the numbers of phone calls for the various groups
are related to the number of people in the group. From their analysis, children can
identify a recursive pattern or conjecture a covariational or correspondence relation-
ship between the total numbers of phone calls and the variations in the parameter
that produces them. Moreover, depending on the grade and skill of the student, the
teacher can scaffold students in describing their conjectures with symbolic nota-
tion. Children can then develop justifications for whether or not their conjectured
relationships and patterns hold true. Finally, the mathematical generalizations that
result, while important results in and of themselves, can become objects of mathe-
matical reasoning as students become more sophisticated algebraic thinkers (Blan-
ton 2008; Blanton and Kaput 2000). None of these processes occur if tasks remain
purely arithmetic in scope.
As we describe elsewhere, our approach “recasts elementary mathematics in a
profound way, not by ignoring its computational agenda, but by enlarging the agenda
in ways that include the old in new forms that deliberately contextualize, deepen,
and leverage the learning of basic skills and number sense by integrating them into
the formulation of deeper mathematical understandings” (Kaput and Blanton 2005).
In essence, a powerful result of transforming arithmetic tasks in this way is that
children are doing many important things all at once, including building number
sense, practicing number facts, building and recognizing patterns to model situa-
tions, and so forth (Blanton and Kaput 2003). In fact, this genre of tasks can provide
large amounts of computational practice in a context that intrigues students and that
avoids the mindlessness of numerical worksheets.
Transforming the Curriculum Empowers Teachers
Moreover, we have found that when teachers transform their own instructional re-
source base so that arithmetic tasks are extended to include opportunities for es-
tablishing and expressing mathematical generalizations, they are able to transcend
constraints imposed by their existing school culture such as limited or inadequate
resources, or even their own lack of experience with teaching algebraic thinking.
Instead, they are able to see algebraic thinking as a fluid domain of thinking which
permeates all of mathematics, not as a set of tasks or a prescribed curriculum. Thus,
what we advocate, more so than an “early algebra curriculum” per se, involves the
development of a habit of mind that transcends the particular resource being used
and allows elementary teachers to see opportunities for algebraic thinking, and func-
tional thinking in particular, in the mathematics they already teach, using the cur-
riculum they have in place. After using only two functional thinking tasks with her
students, one third-grade teacher wrote
I had a new outlook on math. I knew I wanted to integrate algebraic thinking into every
topic I did. The truth was that our curriculum was wonderful. It allowed plenty of ways to
integrate this way of thinking. I just hadn’t noticed up to this point (Blanton 2008, p. xii).
This sense of empowerment, as well as the development of an algebraic habit of

mind, was later echoed by a first-grade teacher:
[Functional thinking] activities at the beginning seemed like they were going to be hard
to do, never mind creating my own. I’ve realized that they are a lot simpler to create and
implement than I thought. I am really impressed with how these activities have shaped my
way and my students’ way of thinking algebraically. They have really opened my mind up
about algebra and how, if we put it into a simple form, our students can do it! (Blanton
2008, p. 147)
Using Children’s Functional Thinking to Leverage Teacher

Learning
Integrating functional thinking into instruction does not rest solely on the particular
materials the teacher chooses or develops. It requires an “algebra sense” by which
teachers can identify occasions in children’s thinking to extend conversations about
arithmetic to those that explore mathematical generality. While the task one chooses
can certainly support this, teachers also need the skills to interpret what children are
writing about and talking about. In turn, a written or verbal record of student think-
ing can serve as a tool to engage teachers in thinking about content and practice. As
teachers think collectively about how children make sense of data, whether and how
they attend to how quantities relate, the kinds of meaning they derive from tables
and graphs, and how they use symbols in describing and reasoning with mathemat-
ical ideas, they have the potential to build functional thinking into instruction in
deeper and more compelling ways (Kaput and Blanton 2005).
The work of Cognitively-Guided Instruction (Carpenter and Fennema 1999) has
been significant in bringing student thinking to the fore in how people conceptu-
alize and engage in teacher professional development. More recently, researchers
have extended this approach as a tool in the development of teachers’ early alge-
braic thinking (Franke et al. 2001; Kaput and Blanton 2005). The assumption is
that focusing on children’s (algebraic) thinking in professional development builds
teachers’ capacity to identify classroom opportunities for generalization and to un-
derstand the representational, linguistic and symbolic tools that support this and the
particular ways students use these to reason algebraically. Thus, if teachers are to
build algebraic thinking into their instruction, they must become engaged in and
by what students are saying, doing and writing as a catalyst for building their own
classroom algebraic discourse. Moreover, they must be given occasions to use these
classroom artifacts to negotiate mathematical and instructional knowledge within
teacher communities of practice as a way to develop their own knowledge of alge-
bra and teaching algebra. One fourth-grade teacher described her early experience
in leading this work with her teacher peers:
At our last professional development day, I told the other teachers that I have been doing
algebra problems during my math workshop time. I told them the types of problems we did
and how I have been implementing the problems in class. I told them it was a great way
to get kids to look at numbers in different ways. I explained how it was more than algebra;
it also helps kids practice basic arithmetic. I showed them samples of students’ work. I
even explained the importance of organizing data, finding a recursive pattern and finding a
function. I talked so confidently about algebra that the teachers were intrigued. For the first
time in my life, I was a math teacher! (Blanton 2008, p. 147)
Creating Classroom Culture and Practice to Support Functional

Thinking
Building on children’s functional thinking in instruction requires that a culture of

practice that promotes this type of thinking exists. Classrooms in which children’s
functional thinking can thrive are those in which the teacher has established socio-
mathematical norms of conjecturing, arguing, and generalizing in purposeful ways,
where the arguments are taken seriously by students as ways of building reliable
knowledge. Robust functional thinking requires children to interact with complex
mathematical ideas, to negotiate new notational systems and to understand and use
representational tools as objects for mathematical reasoning. It requires that the
teacher respect and encourage these processes as standard practice on a daily ba-
sis, not as occasional enrichment treated as separate from the “regular” work of
learning and practicing arithmetic.
The teacher narratives included here illustrate the kinds of classroom practice
and culture that can support the development of children’s functional thinking. For
example, the Growing Caterpillar narrative depicts ways of doing mathematics in
which the teacher (1) followed students’ thinking in shaping a lesson’s agenda (“I
showed my students my caterpillar example and all I wanted them to see was how
I developed the problem. I had no idea that they would begin to solve the problem.
I couldn’t stop them”.), (2) placed the responsibility for conjecture, argumentation
and justification with students (“Okay, what do you think the pattern is?”, “How
many of you agree with Jak?”, “Why is that Meg?”), (3) cultivated children’s use of
representational structures as tools for reasoning (Meg: “I don’t know. I have to do
a t-chart”), (4) encouraged the use of symbolic notational systems as valid forms of
mathematical expression (Jak: “I think it is x times 2 plus one”; Meg: “Because if it
was x times 2 plus 1, then x would be one and y would be three. And, it’s not. It’s
x = 1 and y = 2”; Joe: “I see it, I know the formula!. . . . It’s x × x + 1 = y”), and
(5) used children’s utterances to craft an idea-building, dialogic discourse that led
to symbolizing a functional relationship. In short, these aspects of practice allowed
children to construct a mathematical generalization about the caterpillar’s growth.
Children’s role in this process is critical; we are not advocating a form of practice
in which children do not actively participate in the development of conjectures, the
construction of arguments, the establishment of generalizations, or the use of nota-
tion, language, and tools for reasoning about functions. All of these experiences are
critical components of the kind of classroom culture that makes functional thinking
viable when it does occur.
Conclusion
This chapter elaborates the position that elementary school children are capable of
functional thinking and that its study in the elementary grades can affect their suc-
cess in mathematics in later grades. We propose that elementary grades mathematics
extend beyond the fairly common, initial focus on recursive patterning to include
curriculum and instruction that deliberately attends to how two or more quantities
vary in relation to each other and that begins to scaffold these notions from the start
of formal schooling. Because there is a fundamental conceptual shift that must oc-
cur in how teachers and students attend to data in recursive patterning as opposed
to covariational or correspondence relationships, we speculate that the emphasis on
recursive patterning that does occur in the early elementary grades curricula, could,
if taught in isolation, impede the development of covariational and correspondence
thinking about functions in later grades.
Children’s capacity for functional thinking raises the issue of how it might be
nurtured by curriculum and instruction in the elementary grades. We advocate here
a habit of mind, not just curricular materials, whereby teachers understand both how
to transform and extend their current resources so that the mostly arithmetic content
of the elementary grades can be extended to opportunities for pattern building, con-
jecturing, generalizing, and justifying mathematical relationships and how to embed
this mathematics within the kinds of socio-mathematical norms that allow children
to build mathematical generality. Generalizing is a human activity and an innate,
natural capacity that young children bring to the classroom (Mason 2008). Curricu-
lum and instruction should build on these natural abilities to provide a deeper, more
compelling mathematical experience for young children.
References
Behr, M. J., Lesh, R., Post, T. R., & Silver, E. A. (1983). Rational-number concepts. In R. Lesh
& M. Landau (Eds.), Acquisition of Mathematics Concepts and Processes (pp. 91–126). New
York: Academic Press.
Blanton, M. (2008). Algebra and the Elementary Classroom: Transforming Thinking, Transform-
ing Practice. Portsmouth, NA: Heinemann.
Blanton, M., & Kaput, J. (2000). Generalizing and progressively formalizing in a third grade math-
ematics classroom: Conversations about even and odd numbers. In M. Fernández (Ed.), Pro-
ceedings of the Twenty-Second Annual Meeting of the North American Chapter of the Interna-
tional Group for the Psychology of Mathematics Education, Columbus, OH, ERIC Clearing-
house (pp. 115–119).
Blanton, M., & Kaput, J. (2003). Developing elementary teachers’ algebra eyes and ears. Teaching
Children Mathematics, 10(2), 70–77.
Blanton, M., & Kaput, J. (2004). Elementary grades students’ capacity for functional thinking. In
Proceedings of the International Group for the Psychology of Mathematics Education (Vol. 2,
pp. 135–142). Bergen, Norway: Bergen University College.
Blanton, M., & Kaput, J. (2005). Characterizing a classroom practice that promotes algebraic rea-
soning. Journal for Research in Mathematics Education, 36(5), 412–446.
Boyer, C. (1946). Proportion, equation, function: Three steps in the development of a concept.
Scripta Mathematica, 12, 5–13.
Brizuela, B., & Earnest, D. (2008). Multiple notational systems and algebraic understandings: The
case of the “Best Deal” problem. In J. Kaput, D. Carraher, & M. Blanton (Eds.), Algebra in
the Early Grades. Mahwah, NJ: Lawrence Erlbaum Associates/Taylor & Francis Group and
National Council of Teachers of Mathematics.
Brizuela, B., & Schliemann, A. (2003). Fourth-graders solving equations. In N. Pateman, B.
Dougherty, & J. Zilliox (Eds.), Proceedings of the 2003 Joint Meeting of PME and PME-NA
(Vol. 2, pp. 137–143). Honolulu, HI: College of Education.
Brizuela, B., Carraher, D. W., & Schliemann, A. D. (2000). Mathematical notation to support and
further reasoning (“to help me think of something”). Paper presented at the NCTM Research
Pre-Session.
Carpenter, T. P., & Fennema, E. (1999). Children’s Mathematics: Cognitively Guided Instruction.
Portsmouth, NH: Heinemann.
Carpenter, T., Franke, M., & Levi, L. (2003). Thinking Mathematically: Integrating Arithmetic and
Algebra in Elementary School. Portsmouth, NH: Heinemann.
Carraher, D. W., Schliemann, A. D., & Schwarz, J. L. (2008). Early algebra = algebra early. In J.
Kaput, D. Carraher, & M. Blanton (Eds.), Algebra in the Early Grades. Mahwah, NJ: Lawrence
Erlbaum/Taylor & Francis Group & National Council of Teachers of Mathematics.
Confrey, J., & Smith, E. (1991). A framework for functions: Prototypes, multiple representations,
and transformations. In R. G. Underhill (Ed.), Proceedings of the 13th Annual Meeting of the
North American Chapter of the International Group for the Psychology of Mathematics Edu-
cation, Blacksburg, VA (Vol. 1, pp. 57–63).
Cramer, K. (2001). Using models to build an understanding of functions. Mathematics Teaching in
Middle School, 2(4), 310–318.
Franke, M. L., Carpenter, T. P., Levi, L., & Fennema, E. (2001). Capturing teachers’ generative
growth: A follow-up study of professional development in mathematics. American Educational
Research Journal, 38(3), 653–690.
Freudenthal, H. (1982). Variables and functions. In G. V. Barneveld & H. Krabbendam (Eds.), Pro-
ceedings of Conference on Functions (pp. 7–20). Enschede, The Netherlands: National Institute
for Curriculum Development.
Goldin, G., & Shteingold, N. (2001). Systems of representations and the development of mathe-
matical concepts. In A. A. Cuoco & F. R. Curcio (Eds.), The Roles of Representation in School
Mathematics. NCTM 2001 Yearbook (pp. 1–23). Reston, Virginia: National Council of Teachers
of Mathematics.
Hamley, H. R. (1934). Relational and functional thinking in mathematics. In The 9th Yearbook of
NCTM. New York: Bureau of Publications, Teachers College, Columbia University.
Kaput, J. (1995). A research base for algebra reform: Does one exist. In D. Owens, M. Reed, & G.
M. Millsaps (Eds.), Proceedings of the 17th Annual Meeting of the North American Chapter
of the International Group for the Psychology of Mathematics Education (Vol. 1, pp. 71–94).
Columbus, OH: The ERIC Clearinghouse for Science, Mathematics, and Environmental Edu-
cation.
Kaput, J. (2008). What is algebra? What is algebraic reasoning? In J. Kaput, D. Carraher, & M.
Blanton (Eds.), Algebra in the Early Grades. Mahwah, NJ: Lawrence Erlbaum/Taylor & Fran-
cis Group & National Council of Teachers of Mathematics.
Kaput, J., & Blanton, M. (2005). Algebrafying the elementary mathematics experience in a teacher-
centered, systemic way. In T. A. Romberg, T. P. Carpenter, & F. Dremock (Eds.), Understanding
Mathematics and Science Matters. Mahwah, NJ: Lawrence Erlbaum Associates.
Lins, R., & Kaput, J. (2004). The early development of algebraic reasoning: The current state of
the field. In H. Chick & K. Stacy (Eds.), The Future of the Teaching and Learning of Algebra:
The 12th ICMI Study. London: Kluwer.
Mason, J. (2008). Making use of children’s powers to produce algebraic thinking. In J. Kaput,
D. Carraher, & M. Blanton (Eds.), Algebra in the Early Grades. Mahwah, NJ: Lawrence Erl-
baum/Taylor & Francis Group & National Council of Teachers of Mathematics.
Moss, J., Beatty, R., Shillolo, G., & Barkin, S. (2008). What is your theory? What is your rule?
Fourth graders build their understanding of patterns and functions on a collaborative database.
In C. Greenes (Ed.), Algebra and Algebraic Thinking in School Mathematics: The National
Council of Teachers of Mathematics 70th Yearbook (2008) (pp. 155–168). Reston, VA: NCTM.
National Research Council (2001). Adding It Up: Helping Children Learn Mathematics. Washing-
ton, DC: National Academy Press.
Romberg, T., & Kaput, J. (1999). Mathematics worth teaching, mathematics worth understanding.
In E. Fennema & T. Romberg (Eds.), Mathematics Classrooms that Promote Understanding
(pp. 3–32). Mahwah, NJ: Lawrence Erlbaum Associates.
Schliemann, A. D., & Carraher, D. W. (2002). The evolution of mathematical understanding: Ev-
eryday versus idealized reasoning. Developmental Review, 22(2), 242–266.
Schliemann, A. D., Carraher, D. W., & Brizuela, B. (2001). When tables become function tables.
In M. Heuvel-Panhuizen (Ed.), Proceedings of the Twenty-fifth International Conference for
the Psychology of Mathematics Education, Utrecht, The Netherlands (Vol. 4, pp. 145–152).
Schwartz, J. (1990). Getting students to function in and with algebra. In G. Harel & E. Dubin-
sky (Eds.), The Concept of Function: Aspects of Epistemology and Pedagogy (pp. 261–289).
Washington, DC: Mathematics Associations of America.
Smith, E. (2003). Stasis and change: Integrating patterns, functions, and algebra throughout the K-
12 curriculum. In J. Kilpatrick, W. G. Martin, & D. Schifter (Eds.), A Research Companion to
Principles and Standards of School Mathematics (pp. 136–150). Reston, VA: National Council
of Teachers of Mathematics.
Smith, E. (2008). Representational thinking as a framework for introducing functions in the el-
ementary curriculum. In J. Kaput, D. Carraher, & M. Blanton (Eds.), Algebra in the Early
Grades. Mahwah, NJ: Lawrence Erlbaum Associates/Taylor & Francis Group and National
Council of Teachers of Mathematics.
Thompson, A. G., Philipp, R. A., Thompson, P. W., & Boyd, B. A. (1994). Calculational and
conceptual orientations in teaching mathematics. In A. Coxford (Ed.), 1994 Yearbook of the
NCTM (pp. 79–92). Reston, VA: NCTM.
Vygotsky, L. (1962). Thought and Language. Cambridge, MA: Massachusetts Institute of Tech-
nology (E. Hanfmann and G. Vakar, Trans. Original work published in 1934).
Developing Students’ Algebraic Thinking
in Earlier Grades: Lessons from China
and Singapore
Jinfa Cai, Swee Fong Ng, and John C. Moyer
Abstract In this chapter, we discuss how algebraic concepts and representations

are developed and introduced in the Chinese and Singaporean elementary curric-
ula. We particularly focus on the lessons to be learned from the Chinese and Sin-
gaporean practice of fostering early algebra learning, such as the one- problem-
multiple-solutions approach in China and pictorial equations approach in Singapore.
Using the lessons learned from Chinese and Singaporean curricula, we discuss four
issues related to the development of algebraic thinking in earlier grades: (1) To what
extent should we expect students in early grades to think algebraically? (2) What
level of formalism should we expect of students in the early grades? (3) How can
Authors’ note: The preparation of this paper has been supported by grants from the National
Science Foundation (ESI-0114768; ESI-0454739). Any opinions expressed herein are those of the
authors and do not necessarily represent the views of the National Science Foundation. This
chapter is a revised version of an article published in ZDM (Cai, J., Lew, H. C., Morris, A.,
Moyer, J. C., Ng, S. F. & Schmittau, J. (2005). The Development of Students’ Algebraic thinking
in Earlier Grades: A Cross-Cultural Comparative Perspective. Zentralblatt fuer Didaktik der
Mathematik (International Journal on Mathematics Education), 37(1), 5–15, DOI
10.1007/BF02655892), where we summarized findings from analyses of Chinese, Russian, South
Korean, Singaporean, and U.S. elementary curricula. In this chapter, we only focused on the
Chinese and Singaporean curricula. We are grateful for the productive collaboration with Drs.
H. C. Lew, A. Morris, & J. Schmittau.
J. Cai ()
Department of Mathematical Sciences, University of Delaware, Newark, USA
S.F. Ng
National Institute of Education, Nanyang Technological University, Singapore, Singapore
J.C. Moyer
Department of Mathematics, Statistics and Computer Science, Marquette University, Milwaukee,
USA

26 J. Cai et al.
we help students make a smooth transition from arithmetic to algebraic thinking?

and (4) Are authentic applications necessary for students in early grades?
Introduction
Algebra has been characterized as the most important “gatekeeper” in mathematics.
It is widely accepted that to achieve the goal of “algebra for all,” students in elemen-
tary school should have experiences that prepare them for the more formal study of
algebra in the later grades (National Council of Teachers of Mathematics [NCTM]
2000). However, curriculum developers, educational researchers, and policy mak-
ers are just beginning to explore the kinds of mathematical experiences elementary
students need to prepare them for the formal study of algebra at the later grades
(Britt and Irwin, this volume; Carpenter et al. 2003; Carraher and Schliemann 2007;
Kaput 1999; Mathematical Sciences Education Board 1998; NCTM 2000; Schifter
1999; Stacey et al. 2004).
For example, there is evidence that U.S. students are ill-prepared for the study
of algebra (Silver and Kenney 2001). One of the challenges teachers in the United
States face is the lack of a coherent K-8 curriculum that can provide students with
algebraic experiences that are both early and rich (Schmidt et al. 1996). Customarily,
algebra has not been treated explicitly in the school curriculum until the traditional
algebra course offered in middle school or high school (NCTM 2000). Moreover,
according to a rigorous academic analysis by the American Association for the Ad-
vancement of Sciences (AAAS 2000), the majority of textbooks used for algebra in
the United States have serious weaknesses. In addition, there is evidence that most
elementary school teachers in the United States are not adequately prepared to in-
tegrate algebraic reasoning into their instructional practices (e.g., van Dooren et al.
2002). By comparison, it is plausible that the preparation of Chinese and Singa-
porean elementary school teachers benefits from their own elementary school edu-
cation, in which the formal study of algebra begins much earlier than in the United
States.
Most U.S. students do not start the formal study of algebra until eighth or ninth
grade, and many of them experience difficulties making the transition from arith-
metic to algebra because they have little or no prior experience with the subject
(Silver and Kenney 2001). The unsatisfactory findings from national and interna-
tional assessments (e.g., NAEP, TIMSS) indicate a need to develop U.S. students’
algebraic thinking in the early grades. In several countries (e.g., China and Singa-
pore), students begin the formal study of algebra much earlier. However, it is likely
that, across these settings, curriculum developers, educational researchers, and pol-
icy makers faced challenges similar to those of their counterparts in the United
States and other countries as they attempted to develop early and appropriate alge-
braic experiences for younger children (Cai and Knuth 2005; Carpenter et al. 2003;
Kieran 2004; Stacey et al. 2004).
In this chapter, we discuss how algebraic concepts and representations are de-
veloped and introduced in the Chinese and Singaporean elementary curricula. We
Developing Students’ Algebraic Thinking in Earlier Grades 27
particularly focus on the lessons to be learned from the Chinese and Singaporean
practice of fostering early algebra learning. Knowledge of the curricula and instruc-
tional practices of different nations can increase educators’ and teachers’ abilities to
meet the challenges associated with the development of students’ algebraic thinking.
We should state at the outset that it is not our intent in this paper to evaluate
the curricula or the instructional practices used in China and Singapore. Instead,
our focus is on studying and understanding how the curricula (and instruction) in
the two countries contribute to the development of students’ algebraic thinking. We
believe that the understanding we gain by taking this international perspective will
increase our ability to address the issues related to the development of students’
algebraic thinking in elementary school. We need to support students’ development
of algebraic thinking in the early grades and help them appreciate the usefulness
of algebraic approaches in solving various problems. Therefore, we are especially
interested in the way the two curricula prepare students to make smooth transitions
from informal to formal algebraic thinking.
In the sections that follow, we first present the unique features of the two curricula
we studied. Then we discuss the lessons we can learn from China and Singapore
regarding the development of students’ algebraic thinking in earlier grades.
Features of the Chinese and Singaporean Curricula

We analyzed the national curricula of China and Singapore to understand how
the authors intended to help students develop algebraic thinking in elementary
school (Curriculum Planning & Development Division 2000; Division of Elemen-
tary Mathematics 1999). We were interested in the algebraic concepts included in
the curriculum and in how these concepts are developed and represented. In particu-
lar, we analyzed each curriculum along three dimensions: (1) goal specification, (2)
content coverage, and (3) process coverage (Cai 2004a). Results from the analysis
can be found in Cai (2004b) and Ng (2004). The focus of this chapter is to identify
the unique features of Chinese and Singaporean curricula and then to illustrate how
mathematics teachers and researchers can use the international perspective to tackle
difficult issues related to the development of algebraic thinking in early grades.
Algebra Emphases in the Chinese and Singaporean Curricula
The four goals of the algebra standard in the Principles and Standards for School
Mathematics (NCTM 2000) are: Goal 1—Understand patterns, relations, and func-
tions; Goal 2—Represent and analyze mathematical situations and structures using
algebraic symbols; Goal 3—Use mathematical models to represent and understand
quantitative relationships; and Goal 4—Analyze change in various contexts. We
employed our previously developed case studies of the Chinese and Singaporean
curricula (Cai 2004b; Ng 2004) to compare the emphases of each curriculum with
28 J. Cai et al.
these four goals. Our criteria for determining whether a curriculum emphasizes the
development of a certain NCTM goal are that the curriculum must explicitly state
a similar learning goal or must include extensive instructional activities clearly in-
tended to help students achieve the goal.
There is a consistency between the curricular emphases of both the Chinese and
Singaporean curricula. They both explicitly state that their main goal in teaching al-
gebraic concepts is to deepen students’ understanding of quantitative relationships.
Furthermore, both curricula include extensive use of algebraic models that are in-
tended to help achieve the overriding goal of deepening students’ understanding of
quantitative relationships. These two features indicate that both curricula emphasize
the development of goal 3 of NCTM’s algebra standard.
In fact, the algebraic emphasis in both Chinese and Singaporean elementary
school mathematics is consonant with all but the fourth goal of the NCTM alge-
bra standard. In China, the fourth goal is not addressed fully until the concept of
function is formally introduced in junior high school, although qualitative analysis
of change is done in elementary school. In Singapore, the fourth goal is not devel-
oped until students are in secondary school.
The Chinese Curriculum
The overarching algebra-related goal in the Chinese elementary curriculum is for

students to better represent and understand quantitative relationships, numerically
and symbolically. The main focus is on equations and equation solving. Variables,
equations, equation solving, and function sense permeate the curriculum in grades
1 to 4. Equations and equation solving are formally introduced in the first half of
grade 5. Once equation solving is introduced, it is applied to the learning of mathe-
matical topics, such as fractions, percents, statistics, and proportional reasoning, in
the second half of grade 5 and grade 6.
The term “variable” is not formally defined in Chinese elementary school math-
ematics. However, in the teacher’s guide for the national curriculum, teachers are
reminded that variables can represent many numbers simultaneously, that they have
no place value, and that representations of variables can be selected arbitrarily. In
Chinese elementary school mathematics, variable ideas are used in three different
ways. First, they are used as place holders for unknowns in equation solving. In
grades 1–2, for example, a question mark, a picture, a word, a blanket, or a box
is used to represent the unknowns in equations. Second, variables are viewed as
pattern generalizers or as representatives of a range of values. Specifically, words,
rather than letters, are used to represent variables in order to help 3rd grade students
understand the meanings of formulas. For example, after examining several specific
examples, the formulas for the areas of rectangles and squares are represented as
area of a rectangle = its length × its width and area of a square = its side × its
side. However, teachers are counseled to emphasize the generalizable nature of the
formulas. That is, for any rectangle, its area can be found by multiplying its length
and width. In grades 5 and 6, letters are used to represent formulas for finding areas
of squares, triangles, rectangles, trapezoids, and circles. The third use of variables
is to represent relationships, such as direct proportionality (y/x = k) and inverse
proportionality (xy = k).
The function concept is not formally introduced in the Chinese elementary cur-
riculum. However, function ideas permeate the curriculum so that students’ function
sense can be informally developed. According to the curriculum guide, the perva-
sive use of function ideas in various content areas not only fosters students’ learning
of the content topics, but also provides a solid foundation for the future learning of
advanced mathematical topics in middle and high schools. In the early grades, the
Chinese curriculum provides students with many opportunities to develop function
sense at a concrete, intuitive level. Function sense is first introduced in the context of
comparing and operating with whole numbers in grade 1 using a one-to-one map-
ping. In grade 6, multiple representations (pictures, diagrams, tables, graphs, and
equations) are used to represent functional relationships between two quantities.
These functional relationships are embedded in the curricular treatments of circles,
statistics, and proportional reasoning.
The Chinese elementary school curriculum is intended to develop at least three
thinking habits in students. The first thinking habit is to examine quantitative rela-
tionships from different perspectives. Students are consistently encouraged, and pro-
vided with opportunities, to represent a quantitative relationship in different ways.
Throughout the Chinese elementary school curriculum, there are numerous exam-
ples and problems that require students to identify quantitative relationships and
represent them in multiple ways (Cai 2004b). For example, in the following prob-
lem from Grade 2, the quantitative relationship involves the amount of money paid
to the cashier, the change, and the cost of two batteries: Xiao Qing purchased two
batteries. She gave the cashier 6 Yuans and got 4 Yuans in change back. How much
does each battery cost? The teacher’s reference book recommends that teachers al-
low students to represent the quantitative relationship in different ways, such as the
following:
The amount of money paid to the cashier − the cost of the two batteries = the change.
The cost of the two batteries + the change = the amount of money paid to the cashier.
The amount of money paid to the cashier − the change = the cost of the two batteries.
The second thinking habit is to solve a problem using both arithmetic and alge-
braic approaches. Expectations that students solve problems arithmetically in multi-
ple ways and also algebraically in multiple ways can clearly be seen in the Chinese
curriculum, and such expectations are common in Chinese classrooms by teachers
(Cai 2004c). Furthermore, students are asked to make comparisons between arith-
metic and algebraic ways of representing quantitative relationships. For example, in
Grade 5, teachers and students can discuss and compare different ways to solve the
following problem: Liming elementary school has funds to buy 12 basketballs at 24
Yuans each. Before buying the basketballs, they decided to spend 144 Yuans of the
funds for some soccer balls. How many basketballs can they buy?
30 J. Cai et al.
Arithmetic solutions:
Solution 1: Begin by computing the original funding and subtract the money spent
on soccer balls: (24 × 12 − 144) ÷ 24 = 144 ÷ 24 = 6 basketballs.
Solution 2: Begin by computing the number of basketballs that can no longer be
bought: 12 − (144 ÷ 24) = 6 basketballs.
Algebraic solutions:
Solution 3: Assume that the school can still buy x basketballs: (24 × 12 − 144) =
24x.
Therefore, x = 6 basketballs.
Solution 4: Assume that the school can still buy x basketballs: 24×12 = 24x +144.
Solution 5: Assume that the school can still buy x basketballs. 12 = (144 ÷ 24) + x.
In the Chinese curriculum, after equation solving is introduced, students have
opportunities to use an equation-solving approach to solve application problems
as they learn statistics, percents, fractions, and ratios and proportions (Cai 2004b).
This arrangement is built into the curriculum to deepen the students’ understanding
of quantitative relationships and to help students appreciate the equation-solving
approach. For example, students in Grade 6 are encouraged to use four different
methods to solve the following percent problem: A factory modified its production
procedures. After that, the cost of making one product was 37.40 Yuans which is 15%
lower than the cost before the production procedures were modified. What was the
cost of making the same product before the production procedures were modified?
Solution 1: If the pre-modification cost is viewed as the unit 1, then the current cost
is 15% less than the cost before modification.
Let x = the cost before the modification.
x − 15%x = 37.4
(1 − 15%)x = 37.4
85%x = 37.4
x = 44
Answer: The cost before the modification was 44 Yuans.
Solution 2: If the pre-modification cost is viewed as the unit, then 37.40 is 15% less,
or 1 − 15% = 85% of the cost before the modification. Therefore, the cost before
the modification is 37.4 ÷ 85% = 44.
Solution 3: Similarly, the figure below can be used to represent the problem if the
pre-modification cost is viewed as the unit. Since the current cost is 15% less than
the cost before the modification, the current cost is 85% of the previous cost. There-
fore, the cost before the modification was 37.4 ÷ 85% = 44.
| |
The cost before the modification
37.40 Yuans .
| | . . . . . . . . . . . ..
The current cost 15%
Solution 4: Because the 85% also represents a ratio, the ratio of costs before and
after the modification is 85 to 100. It must be the same as 37.40 Yuans to x, where
x is the cost before the modification.
Hence, 37.4/x = 85/100, x = 44.
Undoubtedly, the approach of asking students to use and compare these four
different types of solutions (algebraic, arithmetic, pictorial, and ratio) is based on
the principle that considering multiple perspectives can foster a deep understanding
of the relationship between quantities.
According to the Chinese curriculum, solving a problem using both an arithmetic
approach and an algebraic approach helps students build arithmetic and algebraic
ways of thinking about problem solving. At the elementary school level, Chinese
students solve problems like the examples given in this section, all of which can
be solved arithmetically. As might be expected, it is common at the beginning of
their transition to algebraic problem solving for students to wonder why they need
to learn an equation-solving approach. However, after a period of time using both
approaches, students come to see the advantages of using equations to solve these
types of problems. In recent years, several researchers have discussed the notion
of “algebra in arithmetic” (e.g., Britt and Irwin, this volume; Russell et al., this
volume). While Chinese school mathematics does not explicitly claim this notion,
the use of both arithmetic and algebraic approaches to solve problems can help show
students the algebra in arithmetic. Although this practice contrasts the arithmetic and
algebraic approaches, it also helps soften the boundaries between them.
There are three objectives in teaching students to solve problems both arith-
metically and algebraically: (1) to help students attain an in-depth understanding
of quantitative relationships by representing them both arithmetically and alge-
braically; (2) to guide students to discover the similarities and differences between
arithmetic and algebraic approaches, so they can understand the power of a more
general, algebraic approach; and (3) to develop students’ thinking skills as well as
flexibility in using appropriate approaches to solve problems. Post et al. (1988) in-
dicated that “first-describing-and-then-calculating” is one of the key features that
make algebra different from arithmetic. Comparisons between the arithmetic and
algebraic approaches can highlight this unique feature.
The third thinking habit in Chinese school mathematics is to use inverse oper-
ations to solve equations. Starting in the first grade, subtraction is defined as the
inverse of addition. Although the term “solve” is not used at grade 1, students learn
32 J. Cai et al.
to solve equations starting at grade 1 and they continue solving equations through-
out the entire curriculum. For example, students in the first grade are guided to think
about the following question: “If 1 + ( ) = 3, what is the number in ( )?” In order to
find the number in ( ), the subtraction 3 − 1 = 2 is introduced. Throughout the first
grade, students are consistently asked to solve similar problems. In the second grade,
multiplication and division with whole numbers are introduced in Chinese elemen-
tary school mathematics. Division is first introduced using equal sharing. Division
is also presented as the inverse of multiplication: “What multiplied by 2 = 8?” That
is, “If ( ) × 2 = 8, what is the number in ( )?”
In addition, the Chinese curriculum emphasizes generalizing from specific ex-
amples. By examining specific examples, students are guided to create generalized
expressions. Students develop this habit of mind at a variety of points in the curricu-
lum, but especially when formulas for finding perimeters, areas, and volumes are
introduced, when operational laws are presented, or when the averaging algorithm
is discussed. In particular, generalizing is intertwined in the three habits of minds
mentioned above.
The Singaporean Curriculum
In Singapore, some algebraic concepts are formally introduced in elementary grade

six (age 12+). At this level children are taught how to construct, simplify, and eval-
uate algebraic expressions in one variable. The notion of letters as variables is intro-
duced at this level. The concept of equations and other structural aspects of algebra
are developed in lower secondary years (age 13+ onwards). However, the Singa-
porean elementary mathematics curriculum provides a wide variety of experiences
to help younger children develop algebraic thinking, and this development is made
possible by using “model methods” or “pictorial equations” to analyze parts and
wholes, generalize and specify, and do and undo.
Solving arithmetic and algebra word problems is a key component at every level
of the Singapore elementary mathematics curriculum (Curriculum Planning & De-
velopment Division [CPDD] 1999, 2000). In 1983, the Singapore Ministry of Ed-
ucation officially introduced into the elementary mathematics curriculum a heuris-
tic involving diagram- or model-drawing. This heuristic was intended as a tool for
solving arithmetic, as well as algebraic, word problems involving whole numbers,
fractions, ratios and percents (Kho 1987). It was believed that if students were pro-
vided with the means to visualize a word problem—be it a simple arithmetic word
problem or an algebra word problem—the structural underpinning of the problem
would be made overt. Once children understood the structure of the problem, they
were more likely to solve it (Kho 1987).
In the earlier grades, pictures of real objects are initially used to model problem
situations, but then the pictures are replaced by the more abstract rectangles. For
example, actual pictures of cars, and then rectangles, are used to solve the following
problem in second grade: Ali has 8 toy cars and David has 6 toy cars. How many
Fig. 1 Pictorial equation Raju and Samy shared $410 between them. Raju re-
solving ceived $100 more than Samy. How much money did
Samy receive?
2 units = $410 - $100
= $310
1 unit = $155
Samy received $ 155
toy cars do they have altogether? This pictorial approach, which becomes increas-
ingly more complex as the grade level progresses is introduced in the first grade’s
teacher’s guide. As students advance through the primary years, the model method
is used to solve algebra problems involving unknowns, the part-whole concept and
proportional reasoning. In each case, the rectangles allow students to treat unknowns
as if they are knowns. This is because the unknowns are represented by unit rect-
angles that can be treated as if they are knowns, even though the unit represents an
unknown number of objects.
Figure 1 is an example from Grade 5. Samy’s rectangle or unit is the genera-
tor of all the relationships presented in the problem. Raju’s rectangle is dependent
upon Samy’s, with Raju’s share represented by a unit identical to Samy’s plus an-
other rectangle representing the relational portion of $100 more. Using the model
drawing, a pictorial equation representing the problem is formed, and if the letter x
replaces Samy’s unit, then the algebraic equation x + x + $100 = $410 is produced.
The use of the rectangle as a unit representing the unknown provides a pictorial link
to the more abstract idea of letters representing unknowns. The entire structure of
the model can be described as a pictorial equation.
In summary, children solve word problems using the “model method” to con-
struct pictorial equations that represent all the information in word problems as a
cohesive whole, rather than as distinct parts. To solve for the unknown, children
undo the operations that are implied in the pictorial equation. This approach helps
further enhance their knowledge of the properties of the four operations. The intent
of using the “model method” described above is to provide a smooth transition from
working with unknowns in less abstract form to the more abstract use of letters in
formal algebra in secondary school.
Besides the use of pictorial equations to analyze part-whole relationships, the
second big idea in the Singaporean curriculum is developed by exploring the struc-
tures in patterns. Students are provided with both numeric and geometric pattern
recognition activities. Such activities require students to specify and then generalize
the rule they construct to continue the pattern they see. The inclusion of many such
activities in the curricular materials suggests that the Singapore primary mathemat-
ics syllabus places great importance on the thinking processes—generalizing and
specifying.
34 J. Cai et al.
Fig. 2 Inverse operations in

the second grade Singaporean
curriculum
Furthermore children are provided a variety of activities that foster the devel-
opment of other algebraic thinking habits—doing and undoing, building rules to
represent functions, and also abstracting from computation (Driscoll 1999). For ex-
ample, the process of “doing-undoing” is emphasized in the Singaporean teacher’s
guide, specifically when teachers first introduce the four operations. In particular,
in the unit “Addition and Subtraction,” notes in the second grade teacher’s guide
suggests that teachers highlight the relationship between addition and subtraction as
well as multiplication and division, as evidenced by the example in Fig. 2 (TG2A,
1995, pp. 23–24).
The habit of mind, building rules to represent functions, as well as the notion
of letters as variables, is developed using a functional approach. Through this ap-
proach, children engage in activities that help them develop the pointwise notion of
function by looking for the relations exhibited in sets of ordered pairs representing
the input and output of problem situations (e.g. (1, 2), (2, 4), (3, 6), . . . , (x, 2x)).
Tasks move from simple numerical activities where children are engaged in doing
and undoing to using letters to generalize the operation that produces the output
from a given input. As children perform these tasks, they address the question,
“What’s the rule for this pattern?” This notion is first developed through supple-
mentary activities in the second grade teacher’s guide (Ng 2004) designed to help
students memorize the multiplication facts and solve word problems (see Fig. 3).
The functional approach provides students in Grade 3 with experiences extending
sequentially written number patterns, and with tasks that require them to look for
number patterns presented in tables. Here students devise the rule linking numbers
in one row/column with numbers in another row/column, thus continuing the in-
formal introduction to pointwise functions begun in second grade. In the Singapore
mathematics curriculum, these activities are generally presented to students within
a context that is familiar to them.
The complexity of the doing and undoing process increases in Grade 3. For ex-
ample, the teacher’s guide (Ng 2004) challenges students to determine the input
number from the output after two given operations. Add 30 to a number. Then sub-
tract 35 from the sum. The answer is 27. What is the original number?
At Grade 6, a developmental approach to the informal introduction of functions
(pictorial to symbolic) is used to introduce non-recursive rules and the concept of
letter as variable.
Lessons from Chinese and Singaporean School Mathematics

What lessons can we learn from Chinese and Singaporean curricula about devel-
oping students’ algebraic thinking in earlier grades? In the sections that follow we
Fig. 3 Informal introduction

to functions in the second
grade Singaporean
curriculum
present insights from our curricular analyses. In particular, we address the follow-
ing issues: the reason curricula should expect students in early grades to think al-
gebraically, the level of formalism and generalization expected of students, the na-
ture of support for helping students make a smooth transition from arithmetic to
algebraic thinking, and the role authentic applications play in fostering algebraic
thinking.
Why Should Curricula Expect Students in Early Grades to Think

Algebraically?
We realize that the question of whether we should expect students in early grades to
think algebraically is not an issue these days. Based on recent research on learning,
there are many obvious and widely accepted reasons for maintaining this expecta-
tion. However, we raise the question of why in order to offer a less obvious reason
for developing algebraic ideas in the earlier grades, namely that resistance to algebra
would be reduced if we could remove the misconception that arithmetic and algebra
are disjointed subjects.
Although one can make an eloquent argument in favor of studying algebra at
the secondary level (e.g., Usiskin 1995), in reality, the need to learn algebraic ideas
currently is not as universally accepted as the need to learn arithmetic, history, or
writing (Usiskin 1995). Even those who have taken an algebra course and have done
well can live productive lives without ever using it. Therefore, many middle and high
school students are not motivated to learn algebra. We believe that resistance to al-
gebra can be more effectively addressed by helping students form algebraic habits
of thinking starting at elementary school. If students and teachers routinely spent
the first five or six years of elementary school simultaneously developing arith-
metic and algebraic thinking (with differing emphases on both at different stages
of learning), arithmetic and algebra would come to be viewed as being inextricably
interconnected. We believe an important outcome would be that the study of algebra
in secondary school would become a natural and non-threatening extension of the
mathematics of the elementary school curriculum.
Although it is widely accepted that we should expect students in early grades to
think algebraically, the real question is how can we prepare students in earlier grades
36 J. Cai et al.
Fig. 4 Fifth grade

Furniture Problem: At a sale, Mrs. Tan spent $530 on a ta-
Singaporean student solution
ble, a chair and an iron. The chair cost $60 more than the
iron. The table cost $80 more than the chair. How much did
the chair cost?
to think algebraically? “Although there is some agreement that algebra has a place in
the elementary school curriculum, the research basis needed for integrating algebra
into the early mathematics curriculum is still emerging, little known, and far from
consolidated” (Carraher and Schliemann 2007, p. 671). Our analyses indicate that
the Chinese and Singaporean curricula could be useful references for those wishing
to help elementary students develop a stronger sense of the connections between
arithmetic and algebra. Specifically, the Chinese and Singaporean curricula provide
concrete examples of promising ways to integrate arithmetic and algebraic ideas in
the earlier grades.
Are Young Children Capable of Thinking Algebraically?
Our research clearly shows that elementary Chinese and Singaporean students are
capable of using algebraic approaches to solve problems (Cai 2003, 2004b, 2004c;
Ng and Lee 2005). For example, when 151 5th grade Singaporean students were
asked to solve the “Furniture Problem,” shown in Fig. 4, nearly a half of them cor-
rectly used the pictorial equation approach to solve the problem.
In comparative studies involving Chinese and U.S. 6th grade students, we found
that U.S. 6th grade students tended to use concrete problem-solving strategies, while
Chinese students tended to use generalized problem-solving strategies involving
letter symbols as generalized representatives of ranges of values (Cai and Hwang
2002).
For example, in solving the Odd Number Pattern Problem shown in Fig. 5, the
U.S. and Chinese students had almost identical success rates (70%) when they were
asked to find the number of guests who entered on the 10th ring. However, the suc-
cess rate for Chinese students (43%) was higher than that of the U.S. students (24%)
when they were asked to find the ring number on which 99 guests would enter the
party (χ2(1, N = 253) = 10.23, p < .01). This appears to be due to the fact that
more Chinese than U.S. students used abstract strategies to answer the question. In-
deed, fully 65% of Chinese students choosing an appropriate strategy for Question
C used an abstract strategy, compared to only 11% for the U.S. sample. In contrast,
the majority (75%) of U.S. students chose concrete strategies, compared to 29% of
Sally is having a party.

The first time the doorbell rings, 1 guest enters.
The second time the doorbell rings, 3 guests enter.
The third time the doorbell rings, 5 guests enter.
The fourth time the doorbell rings, 7 guests enter.
The guests keep arriving in the same way. On the next ring a group enters that has 2 more
persons than the group that entered on the previous ring.
A. How many guests will enter on the 10th ring? Explain or show how you found your answer.
B. Write a rule or describe in words how to find the number of guests that entered on each
ring.
C. 99 guests entered on one of the rings. What ring was it? Explain or show how you found
your answer.
Fig. 5 Odd number pattern problem
the Chinese students. Abstract strategies (e.g. solve for n if 99 = 2n − 1) are more
efficient than concrete strategies (e.g., repeatedly adding 2 until 99 is reached or
making an exhaustive table or list) to answer the third question, which involves “un-
doing” (i.e., finding the ring number when the number of entering guests is known).
This Odd Number Pattern Problem was administered along with other tasks to some
4th , 5th , and 6th Singaporean students (Cai 2003). It was found that 12% of the 4th
graders, 16% of the 5th graders, and 37% of the 6th graders used abstract strategies.
The findings from these studies about Chinese and Singaporean students’ use of
abstract strategies to solve problems like the Odd Number Pattern suggest that young
children are capable of thinking algebraically. The findings also suggest that the
Chinese and Singaporean approaches are beneficial for helping elementary students
develop algebraic thinking.
How Can We Help Students to Think Arithmetically and

Algebraically?
According to Kieran (2004), in the transition from arithmetic to algebra, students

need to make many adjustments in the way they think, even those students who are
quite proficient in arithmetic. Kieran particularly suggested the following five types
of adjustments in developing an algebraic way of thinking: (1) Focus on relation-
ships and not merely on the calculation of a numerical answer, (2) Focus on inverses
of operations, not merely on the operations themselves, and on the related idea of
doing/undoing, (3) Focus on both representing and solving a problem rather than
on merely solving it, (4) Focus on both numbers and letters, rather than on num-
bers alone, (5) Refocus on the meaning of the equal sign. Helping students make
a smooth transition from arithmetic to algebraic thinking is a common goal in the
two Asian curricula. There are at least three ideas in the Chinese and Singaporean
curricula that help students make the adjustments needed to develop algebraic ways
of thinking.
38 J. Cai et al.
The first is related to the use of inverse operations to solve equations. For exam-
ple, in Chinese elementary schools, addition and subtraction are introduced simul-
taneously at the first grade, and subtraction is introduced as the inverse operation
of addition (Cai 2004a, 2004b, 2004c). The idea of equation and equation solv-
ing permeates the introduction of both subtraction and division. Similar approaches
are taken in the Singaporean curricula (see Fig. 2). There is no doubt that this in-
verse operation approach to subtraction and division can help students make two of
the adjustments suggested by Kieran: (1) Focus on relationships and not merely on
the calculation of a numerical answer, and (2) Focus on inverses of operations, not
merely on the operations themselves, and on the related idea of doing/undoing.
The second idea is the use of pictorial equation solving in the Singaporean cur-
riculum. Pictorial equation solving clearly can help students to focus on both rep-
resenting and solving a problem rather than on merely solving it, as suggested by
Kieran. Before the calculation of the numerical answer can begin, the model draw-
ing has to make clear the relationships between the different objects. The pictorial
equation solving approach also focuses on both numbers and letters, rather than on
numbers alone, as well as on both representing and solving a problem rather than
on merely solving it. Although rectangles, rather than letters, are used to represent
variables in the pictorial equation solving approach, the model drawings show how
children must learn to be flexible in their use of the rectangles, just as they do in later
years when they use letters to represent variables. In a given question, some rectan-
gles can be used to represent unknown values and others can be used to represent
known parameters.
The third idea is that solving problems is done using both arithmetic and al-
gebraic approaches in Chinese curriculum. This idea, incorporating the Chinese
practice of using both arithmetic and algebraic approaches to solve problems, can
help students to focus on both numbers and letters, rather than on numbers alone.
Through the comparisons of the approaches without using letters (arithmetic ap-
proach) and with letters (algebraic approach), students are able to see the role of
letters in the algebraic approaches. Incorporating the Chinese practice of using both
arithmetic and algebraic approaches to solve problems can also focus on relations
and not merely on the calculation of numeric answers, as well as focus on both
representing and solving a problem rather than on merely solving it. In fact, the
two different approaches can be viewed as different ways to represent quantitative
relationships.
Are Authentic Applications Necessary for Students in Early

Grades?
Some researchers and educators believe that the learning of algebraic ideas should
always be anchored in real-world situations that the students are familiar with (e.g.,
Bell 1996). Some U.S. Standards-based curricula reflect this view (e.g., Senk and
Thompson 2003). These curricula engage U.S. students in mathematical problems
embedded in authentic contexts. The applied problem solving activities require U.S.
students to explore contextualized problems in depth, construct strategies and ap-
proaches based on their understanding of mathematical relationships, utilize a va-
riety of tools (e.g., manipulatives, computers, calculators), and communicate their
mathematical reasoning through drawing, writing, and talking.
Others believe algebra does not have to be learned using real-world situations
because the essence of algebra is not applied (Kieran 1992; Usiskin 1995). Rather,
at its core, algebraic knowledge is an understanding of mathematical structures and
relationships. So the work of algebra should be to abstract properties of operations
and structures, and the goal should be to learn the abstract structures themselves,
rather than to learn how the structures can be used to describe the real world.
While applications are important in both the Chinese and Singaporean curric-
ula, the contexts of application problems are not truly authentic. In the case of the
Chinese curriculum, for example, the main focus is on equations and the process of
equation solving itself, rather than on the use of applications to provide insight into
the equation solving process. As a result, the development of equation and equation
solving ideas in the Chinese elementary mathematics curriculum is done in three
interrelated stages: (1) the intuitive stage, (2) the introduction stage, and (3) the ap-
plication stage. After Chinese students have been formally introduced to equations
and equation solving, there are opportunities to use an equation-solving approach
to solve application problems as they learn statistics, percents, fractions, and ratios
and proportions (Cai 2004b). This arrangement is desirable in order to deepen the
students’ understanding of quantitative relationships and to help students appreciate
the equation-solving approach.
Conclusion
This study analyzes and compares how algebraic concepts and representations are
introduced and developed throughout the Chinese and Singaporean curricula. It
provides an international perspective to the question of the kinds of algebraic ex-
periences elementary school students should have. In particular, the study identi-
fies unique features of each curriculum. Both curricula state that their main goal
in teaching algebraic concepts is to deepen students’ understanding of quantitative
relationships, but the emphases and approaches to helping students deepen their un-
derstanding of quantitative relationships differ.
The Chinese elementary school curriculum emphasizes the examination of quan-
titative relationships from various perspectives. Students are consistently encour-
aged and provided with opportunities to represent quantitative relationships both
arithmetically and algebraically. Furthermore, students are asked to make compar-
isons between arithmetic and algebraic ways of representing a quantitative relation-
ship.
In Singapore, students are provided ample opportunities to make generalizations
through number pattern activities. Equations are not introduced symbolically; in-
stead, they are introduced through pictures. Such “pictorial equations” are used ex-
tensively to represent quantitative relationships. The “pictorial equations” not only
40 J. Cai et al.
provide a tool for students to solve mathematical problems, but they also provide a
means for developing students’ algebraic ideas.
In this chapter, we addressed four questions related to the development of alge-
braic thinking in earlier grades: (1) Why should curricula expect students in early
grades to think algebraically? (2) Are young children capable of thinking alge-
braically? (3) How can we help students make a smooth transition from arithmetic
to algebraic thinking? (4) Are authentic applications necessary for students in early
grades? We found that both Chinese and Singaporean students in elementary school
are expected to think algebraically. In fact, students in earlier grades are capable of
thinking algebraically to solve problems. The earlier emphasis on algebraic ideas
may indeed help students develop arithmetic and algebraic ways of thinking about
problems.
Regarding the use of authentic applications in elementary school, both Chinese
and Singaporean curricula include many application problems, but the contexts are
not necessarily authentic. The Chinese elementary curriculum uses formal algebraic
symbolism, but the Singaporean elementary curriculum does not.
It is important to indicate that any curriculum has a complex relationship to what
actually occurs in classrooms. In this paper, the focus of our discussion has been on
the intended treatment of algebraic ideas in the Chinese and Singaporean curricula.
Nevertheless, the Chinese and Singaporean curricula may serve as concrete exam-
ples of what can be implemented when we try to address the issues related to the
development of students’ algebraic thinking in earlier grades.
References
AAAS (2000). Algebra for all: Not with today’s textbooks. See www.project2061.org/newsinfo/
press/r1000426.htm.
Bell, A. (1996). Problem-solving approaches to algebra: Two aspects. In N. Bernardz, C. Kieran,
& L. Lee (Eds.), Approaches to Algebra. Perspectives to Research and Teaching (pp. 167–187).
Dordrecht, The Netherlands: Kluwer Academic Publishers.
Cai, J. (2003). Singaporean students’ mathematical thinking in problem solving and problem pos-
ing: An exploratory study. International Journal of Mathematical Education in Science and
Technology, 34(5), 719–737.
Cai, J. (2004a). Introduction to the special issue on developing algebraic thinking in the earlier
grades from an international perspective. The Mathematics Educator (Singapore), 8(1), 1–5.
Cai, J. (2004b). Developing algebraic thinking in the earlier grades: A case study of the Chinese
mathematics curriculum. The Mathematics Educator (Singapore), 8(1), 107–130.
Cai, J. (2004c). Why do U.S. and Chinese students think differently in mathematical problem
solving? Exploring the impact of early algebra learning and teachers’ beliefs. Journal of Math-
ematical Behavior, 23, 135–167.
Cai, J., & Hwang, S. (2002). Generalized and generative thinking in U.S. and Chinese students’
mathematical problem solving and problem posing. Journal of Mathematical Behavior, 21(4),
401–421.
Special Issue.
Carpenter, T. P., Franke, M. L., & Levi, L. (2003). Thinking Mathematically: Integrating Arithmetic
and Algebra in Elementary School. Portsmouth, NH: Heinemann.
Carraher, D. W., & Schliemann, A. D. (2007). Early algebra and algebraic reasoning. In F. Lester
(Ed.), Second Handbook of Research on Mathematics Teaching and Learning: A Project of the
National Council of Teachers of Mathematics (Vol. II, pp. 669–705). Charlotte, NC: Informa-
tion Age Publishing.
Curriculum Planning & Development Division (1999). Mathematics Syllabus Primary. Singapore:
Ministry of Education.
Curriculum Planning & Development Division (2000). Mathematics Syllabus Primary. Singapore:
Ministry of Education.
Division of Elementary Mathematics (1999). Mathematics: Elementary School Textbook (number
1). Beijing, China: People’s Education Press.
Driscoll, M. (1999). Fostering Algebraic Thinking: A Guide for Teachers Grades 6–10.
Portsmouth, NH:: Heinemann.
Kaput, J. (1999). Teaching and learning a new algebra. In E. Fennema & T. Romberg (Eds.),
Mathematics Classrooms that Promote Understanding (pp. 133–155). Mahwah, NJ: Erlbaum.
Kho, T. H. (1987). Mathematical models for solving arithmetic problems. In Proceedings of Fourth
Southeast Asian Conference on Mathematical Education (ICMI-SEAMS). Mathematical Edu-
cation in the 1990’s (pp. 345–351). Singapore: Institute of Education.
Kieran, C. (1992). The learning and teaching of school algebra. In D. A. Grouws (Ed.), Handbook
of Research on Mathematics Teaching and Learning (pp. 390–419). New York: Macmillan.
Kieran, C. (2004). Algebraic thinking in the early grades: What is it? The Mathematics Educator
(Singapore), 8(1), 139–151.
Mathematical Sciences Education Board (1998). The Nature and Role of Algebra in the K-14 Cur-
riculum: Proceedings of a National Symposium. Washington, DC: National Research Council.
Ng, S. F. (2004). Developing algebraic thinking: A case study of the Singaporean primary school
curriculum. The Mathematics Educator (Singapore), 8(1), 39–59.
Ng, S. F., & Lee, K. (2005). How primary five pupils use the model method to solve word problems.
The Mathematics Educator, 9(1), 60–83.
Post, T. R., Behr, M. J., & Lesh, R. (1988). Proportionality and the development of prealgebra
understandings. In A. Coxford & A. Shulte (Eds.), The Ideas of Algebra, K-12, 1988 Yearbook
(pp. 78–90). Reston, VA: NCTM.
Schifter, D. (1999). Reasoning about operations: Early algebraic thinking in grades K-6. In L. V.
Stiff & F. R. Curcio (Eds.), Developing Mathematical Reasoning in Grades K-12, 1999 NCTM
Yearbook (pp. 62–81). Reston, VA: NCTM.
Schmidt, W. H., McKnight, C. C., & Raizen, S. A. (1996). Characterizing Pedagogical Flow: An
Investigation of Mathematics and Science Teaching in Six Countries. The Netherlands: Kluwer
Academic Publishers.
Schmittau, J., & Morris, A. (2004). The development of algebra in the elementary mathematics
curriculum of V. V. Davydov. The Mathematics Educator (Singapore), 8(1), 60–87.
Senk, S. L., & Thompson, D. R. (Eds.) (2003). Standards-Based School Mathematics Curricula:
What Are They? What Do Students Learn? Mahwah, N.J: Lawrence Erlbaum Associates.
Silver, E. A., & Kenney, P. A. (2001). Results from the Sixth Mathematics Assessment of the Na-
tional Assessment of Educational Progress. Reston, VA: NCTM.
Stacey, K., Chick, H., & Kendal, M. (2004). The Future of the Teaching and Learning of Algebra:
The 12th ICMI Study. The Netherlands: Kluwer Academic Publishers.
Usiskin, Z. (1995). Why is algebra important to learn. In B. Moses (Ed.), Algebraic Thinking in
Grades K-12: Readings from NCTM’s School-Based Journals and Other Publications (pp. 16–
21). Reston, VA: NCTM.
van Dooren, W., Verschaffel, L., & Onghena, P. (2002). The impact of preservice teachers’ content
knowledge on their evaluation of students’ strategies for solving arithmetic and algebra word
problems. Journal for Research in Mathematics Education, 33, 319–351.
Developing Algebraic Thinking in the Context
of Arithmetic
Susan Jo Russell, Deborah Schifter, and Virginia Bastable
Abstract Using classroom episodes from grades 2–6, this chapter highlights four
mathematical activities that underlie arithmetic and algebra and, therefore, provide
a bridge between them. These are:
• understanding the behavior of the operations,
• generalizing and justifying,
• extending the number system, and
• using notation with meaning.
Analysis of each episode provides insight into how teachers recognize the opportu-
nities to pursue this content in the context of arithmetic and how such study both
strengthens students’ understanding of arithmetic operations and enables them to
develop ideas foundational to the study of algebra.
In recent years, the question, “What can be done in the elementary grades to pre-
pare students for algebra?” has received a great deal of attention. The form of the
question sometimes leads to a conception of preparation for algebra that focuses on
doing formal algebra—or aspects of formal algebra—in lower grades. Rather, one
might reframe the question as, “What are ways of thinking, modes of reasoning, and
essential understandings that have their roots in arithmetic and are essential to alge-
bra? What are the underlying connections between arithmetic and algebra?” These
S.J. Russell ()

Education Research Collaborative, TERC, Cambridge, USA
D. Schifter
Education Development Center, Newton, USA
V. Bastable
SummerMath for Teachers, Mount Holyoke College, South Hadley, USA

44 S.J. Russell et al.
questions lead to a focus on finding instructional emphases that both serve the ele-
mentary grade goals of computational fluency and support students to develop the
kind of reasoning that will lead to the need for, and meaningful use of, algebraic
tools.
Several research teams have been pursuing these questions, as is represented by
Kaput et al.’s (2008) anthology, Algebra in the Early Grades. Some of the groups
represented in this collection of current work focus on introducing the concept of
functions, providing tasks which invite students to create function rules to describe
patterns of growth (e.g., Blanton 2008). Others organize their work around general-
izations in the number system. For example, Carpenter et al. (2003) describe class
discussion about true and false number sentences. Over the past decade, the authors
of this chapter have been developing K-5 student curriculum and professional de-
velopment materials for teachers in grades K-8 that address both of these strands of
early algebra (Russell et al. 2008; Schifter et al. 2008a, 2008b). This paper draws
from the part of our research that focuses on how students engage with generaliza-
tions about the behavior of the operations.
From our work with elementary and middle grade teachers, we have identified
four mathematical activities that underlie both arithmetic and algebra and, therefore,
provide a bridge between the two. These are:
• understanding the behavior of the operations,
• generalizing and justifying,
• extending the number system, and
• using notation with meaning.
These themes emerge from content at the heart of the elementary mathematics pro-
gram, and can be highlighted and pursued by teachers who learn to recognize the
opportunities that arise in their classrooms. Focusing on these aspects of arithmetic
addresses two major goals: (1) It enables students to grow from arithmetic towards
algebra, and (2) it strengthens their understanding of arithmetic operations and con-
tributes to computational fluency.
In collaboration with teachers in grades K-8, we have been investigating how stu-
dents articulate, represent, and justify general claims about the operations. We have
also been examining how teachers can recognize the implicit generalizations that
arise in the course of students’ study of arithmetic and make them explicit objects
of study in the classroom (Russell et al. 2006; Schifter et al. 2008c). An impor-
tant component of this research is the close observation of classroom discourse by
teachers, who carefully document and write about learning episodes in their own
classrooms. Through discussion and analysis of these episodes at regular project
meetings and via an electronic web-board, we consider evidence and develop ideas
about students’ early algebra experience.
In each of the next four sections of this chapter we focus on one of the four ar-
eas that links arithmetic and algebra. The examples come from videotaped lessons,
lessons observed by project staff, and narratives written by teachers based on tran-
scripts from their teaching.
Developing Algebraic Thinking in the Context of Arithmetic 45
Understanding the Behavior of the Operations
Computational fluency with the four basic arithmetic operations is a core of the el-
ementary curriculum. In these years, students move from counting to computation.
It is an expectation that students enter middle school with a firm grasp of addition,
subtraction, multiplication, and division of, at least, whole numbers. Most students
come into the secondary grades with procedures for solving basic arithmetic prob-
lems. Yet, even among students who carry out these procedures correctly, there are
persistent problems as they make the transition from arithmetic to algebra. Many of
these problems can be traced to lack of knowledge about the properties and behav-
iors of the operations. At best, these students may understand these properties in the
context of arithmetic, but not access their knowledge in the new context of algebra.
At worst, these students use memorized procedures correctly, but do not understand
why they work or how they are based on properties of the operations.
What does it look like when students don’t have sufficient experience with the
behavior and properties of the operations when they reach algebra? What hap-
pens when only speed with computation and memorization of algorithms are fore-
grounded, while understanding falls into the background? Many teachers of algebra
in the middle and high school note that students repeatedly make the same errors,
for example:
−3 + −5 = 8
(a + b)2 = a 2 + b2
2(xy) = (2x)(2y)
Student Errors
Such errors can be persistent, even in the face of repeated correction. It is likely
that students who make them see a resemblance in the patterns of the symbols to
other, correct rules. For example, students who rely on memorization of calculation
procedures may remember a rule informally expressed as “two negatives make a
positive,” but don’t have other tools that help them determine that this rule applies
to the product of two negative numbers, but not to the sum. Students who make
the second error may incorrectly interpret the exponent as a number that behaves
like a factor, so that (a + b)2 is interpreted in the same way they would interpret
2(a + b). Or, if they do understand the meaning of the exponent, they are not able
to access and apply the distributive property from their knowledge of multiplying
whole numbers. In the third example above, students may be applying a rule to
“multiply everything inside the parentheses by the number outside the parentheses,”
which would work for 2(x + y), but not for 2(xy). They incorrectly apply what
they think is the distributive property and do not recognize an application of the
associative property. In each case, properties of operations are over-generalized or
misapplied.
In our work with elementary and middle grades teachers, we have been investi-
gating how their students benefit from explicit study of the operations, for example,
by examining calculation procedures as mathematical objects that can be described
generally in terms of their properties and behaviors. By this study, we do not mean
that students should learn the names of properties and state them as rules, as oc-
curred in some curricula in the 1960s. Some of us who went to school at that time
remember that we learned, for example, what the commutative, associative, and dis-
tributive properties were, but weren’t quite sure why we were learning them or why
they were so important. Rather, students use representations or story contexts to de-
scribe the behavior of the operations. For example, students might join two sets of
cubes to illustrate addition, switching positions of the sets to show that changing
the order of addends does not affect the sum. They might draw an image of some
amount removed from a larger amount to demonstrate that as the amount removed
(the subtrahend) increases, the result (the difference) decreases. Similarly, students
might use arrays or equal groups of objects to illustrate the behavior of multiplica-
tion and division.
The following classroom episodes illustrate a grade 2 class investigating addition
and subtraction and a grade 5 class investigating multiplication in this way.
Episode A: How Are Addition and Subtraction Different? (Grade 2)
In prior lessons in this second grade class (Schifter et al. 2008a, p. 114), the students
had noticed that if you change the order of the numbers in an addition expression, the
sum remains the same. Many students had been using this idea in their computation,
but the teacher, Maureen Johnson, wanted them to consider this property of addition
explicitly. During this class session, Ms. Johnson asked students to find pairs of
numbers that add to 25. Then she brought students’ attention to the question of
whether the order of two addends can always be changed without affecting the sum.
Teacher: These two numbers that we used, can we switch them around? Can we change the
order and still get 25? I hear a lot of yeses. Who’s not sure? So someone’s not sure? Two
people aren’t so sure? If you feel sure, how would you explain that? Kwame?
Kwame: 18 + 7. Change it around. That’s 7 + 18.
Teacher: So what do you want to say about that?
Kwame: It will still be 25.
Teacher: How come that’s still 25?
Kwame: We didn’t change the numbers.
Teacher: Does someone have another one they want to talk about? Kamika?
Kamika: 19 + 6.
Teacher: OK. If I put the 6 first and then the 19, what will it be?
Kamika: 25, because you’re just switching the numbers. You’re not adding any more and
you’re not taking away any numbers. You’re just changing them around.
Ms. Johnson then asked the class if they were sure this would work for all numbers.
When they said yes, she asked if they could prove it: “Can anyone show me some-
thing that would prove it or explain it better?” She built two towers out of connecting
cubes, one with 23 cubes and one with 2 cubes.
Latifa took the 2-cube tower and moved it rapidly back and forth from one side
of the 23-cube tower to the other.
Latifa: If you keep on switching it around, it will still make 25. Because you’re not taking
away or adding anything to it, so it will still be the same number.
Other students showed that they understood and agreed with Latifa’s actions and
words. Latifa used a representation of joining two sets of cubes to show that 23+2 =
2 + 23, but she also used language to explain why this relationship would hold for
any pair of numbers: If you change the order, nothing more is added and nothing is
taken away, so the total stays the same.
Latifa’s demonstration is an example of a phenomenon we see in many of our
classroom examples: a representation showing specific quantities is talked about
and thought about by students as representing a class of numbers. Although there
are a specific number of cubes in each cube tower, the students can hold this model
in their imagination to represent any pair of numbers—or any pair of numbers they
can imagine (which, for second graders, may be the set of whole numbers or, at
least, the whole numbers with which they are familiar and comfortable).
To find out whether students were, in fact, talking about any pair of numbers
and not just those that sum to 25, Ms. Johnson asked them to consider numbers
larger than they could easily add: “What about 175 + 266?” Her students argued
that 175 + 266 and 266 + 175 must both have the same sum, even though they had
not attempted to carry out the addition. “It doesn’t matter,” they said. “You’re not
adding anything or taking anything away.”
By now Ms. Johnson felt assured that the students in the class were, indeed,
thinking in terms of a generalization, beyond the specific numbers of their examples,
and they were able to describe the essential aspects of a representation to justify the
claim. But she was also concerned that they should not overgeneralize. Were they
thinking about a property that applies to addition, or were they thinking that this
property would apply to any operation? She asked them whether they could apply
their generalization to 7 − 3: does 7 − 3 equal 3 − 7?
Latifa: If you have 3 take away 7, but 3 doesn’t have 7. So you can only do 7 and 3, because
3 is not a 7.
Teacher: There is not enough in 3 to take away 7? Is that what you’re saying? What if I had
3 and I want to take away 7, then how many could I take away?
Latifa: You could only take away 3, to make 0.
Kamika: After you use the 3, it’s 3, 2, 1, 0, 0, 0. The 0 is going to keep on repeating itself
until it gets to 7.
The question of what happens when one changes the order of the numbers in sub-
traction allows the possibility of introducing negative numbers. In fact, at a later
point in the discussion, one student did raise this idea. Antoine stated, “That won’t
be 0, it would be negative 4. . . That means it’s going lower. When you go lower
than 0, that means negative 1, negative 2, negative 3, . . . ” However, most students
in the class, basing their ideas on their familiarity with positive numbers and a “take
away” or removal model of subtraction, came to the conclusion that subtracting 7
from 3 is not possible. If you have 3 cookies and try to eat 7, you can only eat 3; then
you have 0, and no more can be removed. As Latifa says, “you could only take away
3, to make 0.” This reasoning was sufficient to convince students that 7 − 3 = 3 − 7,
and that the commutative property applies to addition, but not to subtraction, which
was the teacher’s purpose for this part of the lesson.
In this class, as in many primary classes, students noticed a regularity as they
solved addition problems: 4+3 and 3+4 are both equal to 7; 5+8 and 8+5 are both
equal to 13; and so forth. Students who notice such a regularity may be convinced it
will always hold because they have encountered many examples and may apply the
rule they have formulated in their computation. This teacher took the opportunity to
make this regularity an explicit focus of investigation. She challenged her students
to think about whether changing the order of the addends maintains the sum only
for specific cases or whether it is true more generally and to explain how they knew.
Keeping the symbols connected to a representation that demonstrates the action
of addition allowed them to explain why their claim must be true. By presenting
a contrasting case of subtraction, she checked to make sure they understood that
their generalization applied specifically to the operation of addition. The students’
explanation of the effect of changing the terms of a subtraction problem was, again,
tied to their understanding of a model of the action of subtraction.
Episode B: Rounding Factors in a Multiplication Problem (Grade 5)
In order to focus on the behavior of the operations, teachers can pay attention to what
regularities students are noticing, as the teacher did in the example above. Another
site for determining which behaviors of the operations might be an important focus
for a particular group of students is student errors, since errors are often related to
the misapplication of basic properties of the operations. In the following example,
students had been working on the problem, 17 × 36. After solving the problem
and comparing results, students in the class knew that the correct product was 612.
However, one student, Thomas, solved the problem this way:
I round 17 to 20 and 36 to 40. I know that 20 × 40 is 800. Then I need to subtract the extra
3 (from rounding 17 to 20) and the extra 4 (from rounding 36 to 40). 800 − 3 − 4 = 793.
The answer is 793.
At this point, the teacher, Liz Sweeney, asked Thomas to put his method on the
board and explain it to the class. Once Thomas—who also knew that the answer
he had was incorrect—had finished his explanation, Ms. Sweeney asked the class
to think through Thomas’s method for homework, to consider how Thomas had
been thinking about the problem, and why his reasoning didn’t lead to the correct
answer.
Why would Ms. Sweeney do this? This particular classroom episode is taken
from a videotape (Schifter et al. 1999) that is used in a professional development
seminar. Some teachers who watch this tape are horrified by the teacher’s move—
that she would focus on this incorrect solution and, even worse, ask students to work
on it at home! While we might debate whether, strategically, we would or wouldn’t
send such an assignment home where it might be misinterpreted, the teacher’s rea-
soning is clear, as she explained to the class. She saw that even students who eas-
ily computed the correct product were somewhat persuaded by Thomas’s reason-
ing. This method looks like it should work—from the point of view of addition:
students didn’t automatically see why his method does not lead to the correct an-
swer.
We might ask, then, what is right or sensible about Thomas’s method? In fact,
in the operation of addition his idea works; one might add some amount to one or
more addends, add the numbers, then subtract those amounts that had been added,
for example:
17 + 36 = (17 + 3) + (36 + 4) − 3 − 4 = 20 + 40 − 3 − 4 = 60 − 7 = 53
Thomas’s method is an example of taking a behavior of one operation and apply-

ing it to another operation where it doesn’t work. By explicitly studying Thomas’s
method and why it doesn’t work, students have to think through the properties of
multiplication—in particular, the distributive property—in order to understand the
role of the 3 and the 4 that Thomas added. In fact, using this problem with adults
over many years, we have found that the exercise of starting with Thomas’s steps
of changing 17 to 20 and 36 to 40, and then figuring out how to complete the prob-
lem correctly (answering the question, what is it you have to subtract from 800?)
is an excellent way for adults to revisit their understanding of multiplication and its
properties.
By having teachers or students examine Thomas’s strategy, we are not advocat-
ing that his procedure (completed in a way that it results in the correct product) is
one that should be learned and used to solve multiplication problems. In Thomas’s
class, the teacher was not hoping that students would routinely alter multiplication
problems in the way he had in order to solve them. His method does not necessar-
ily make the problem easier to solve in the long run. However, figuring out what
has been added to the product by changing the two factors gets at the heart of the
meaning of multiplication and the distributive property, making this procedure worth
studying. One way of representing the effect on the product of increasing the factors,
as Thomas does, is illustrated below:
This analysis requires representing the operation of multiplication in a way that

manifests the distributive property (which may be hidden from students by some of
the algorithms they use). Such visualization of the way factors are pulled apart and
multiplied by parts of other factors applies to both arithmetic and algebraic contexts.
The reasoning that students might engage in to decode Thomas’s error is similar to
the reasoning they might engage in to justify why (a + b)2 is not equal to a 2 + b2 .
Their understanding of the distributive property can be explicitly called upon, so
that they can visualize that (a + b)2 cannot possibly be equivalent to a 2 + b2 unless
a or b is equal to 0.
Ms. Sweeney reported that Thomas’s error led to three days of deep thinking en-
gaging the entire class. The students drew pictures of groups and presented arrays
to explain what happens when the two factors are increased.
Generalizing and Justifying
A second mathematical activity that connects arithmetic to algebra is articulating,

representing, and justifying generalizations about the operations. As seen in the
episodes in the previous section, general ideas arise frequently in the course of
students’ study of arithmetic. For example, young students notice that when they
change the order of addends, the sum does not change. Older students notice the
same thing about multiplication expressions. Throughout the elementary grades,
opportunities arise to investigate general claims about the operations that can be
brought to the explicit attention of the students.
There are two aspects of engaging with general claims that we see teachers de-
veloping in the elementary grades:
• articulating particular general claims based on the regularities students notice in
the behavior of numbers and operations
• developing a mathematical argument to justify a general claim for a class of num-
bers
The three classroom episodes in this section are examples of (1) a teacher helping
her third graders focus on the articulation of a general claim; (2) a group of fifth
graders who are developing both articulation and justification as they investigate
equivalent addition expressions; and (3) fifth graders’ representation-based proof of
a generalization about multiplication.
Articulating General Claims
As students in the elementary grades are given opportunities to notice and discuss
generalizations about number and operations, they encounter the need for language
to describe the generalizations they are investigating. Young students often use
words like “it” or “that,” or use a gesture such as pointing, to indicate what they
are describing. In math class, when a student says, for example, “I think it’s true,”
it is important to clarify exactly what “it” means, both so that the student offering
the idea can clarify his or her own thinking and so that other students do not make
different assumptions about the nature of the assertion being considered. Putting
reasoning into words can be challenging, for students or adults, but clarification
of the language and clarification of the ideas appear to develop together for young
students, as illustrated in the next example.
Episode C: Equivalent Expressions in Addition and Subtraction (Grade 3)
Alice Kaye’s third graders had formulated a general claim about addition, which
had been expressed by one of the students, Clarissa, as: “When you’re adding two
numbers together, you can take some amount from one number and give it to the
other, and if you add those up, it will still equal the same thing.” A few weeks later,
Alice asked the class to consider subtraction: “By Clarissa’s statement, we could
say that we know this equation is true: 57 + 21 = 58 + 20. Without even doing the
addition, we would know that whatever 57 + 21 equals, 58 + 20 also equals that
same total. Would it also be true to say that 57 − 21 = 58 − 20?” Students quickly
computed 57 − 21 and 58 − 20 and concluded that the differences are not equal, but
students were puzzled about why this was true. As one student put it, “why wouldn’t
they be the same?”
After a couple of days of investigating this question and coming up with story
contexts to illustrate their ideas, the class was considering two series of equations:
25 + 0 = 25 25 − 0 = 25
24 + 1 = 25 26 − 1 = 25
23 + 2 = 25 27 − 2 = 25
22 + 3 = 25 28 − 3 = 25
21 + 4 = 25 29 − 4 = 25
20 + 5 = 25 30 − 5 = 25
19 + 6 = 25 31 − 6 = 25
The set of subtraction equations had been generated using a story context that Todd
had come up with:
If Todd had 26 baseball cards, and his little brother stole 1, he’d have 25 cards left. What
are other numbers of baseball cards Todd could start with, and how many would his little
brother have to steal so that he would always have 25 cards left?
In the course of the discussion about the two sets of equations, the teacher repeatedly
asked her class to clarify what their general claim was as they were developing
arguments to support it:
Nora [looking at the chart]: So I guess it only works with adding, not subtracting.
Teacher: What only works with adding? What’s the “it?”
Nora: The. . . um. . . the. . . the. . . [a long pause, but she cannot yet put into words what she
was thinking was “working with addition”]
Carl: There’s both the same thing in the middle. . . 0, 1, 2, 3, 4, 5, 6 [pointing to the subtrac-
tion sequence] and 0, 1, 2, 3, 4, 5, 6 [pointing to the addition sequence]. But 26, 27, 28, 29,
30 is the other way from 24, 23, 22, 21, 20, 19.
Clarissa: I noticed that’s because it’s going down, and this is going up. . . because in order
to minus, you usually have to go up because if you did like 25 − 0 = 25, 24 − 1 would be
23. That would be if you did the same thing as this [pointing to the addition sequence].
Teacher: And what’s the “this” you’re talking about?
Clarissa: 25 + 0, 24 + 1, 23 + 2. Because that would be adding one on, but you’re subtracting
one off.
Todd: Since this one is going down [pointing to the first addend in the addition series of
equations] this one [pointing to the second addend in the addition series of equations] has
to go up, too. This column is going down, so this one has to go up.
Teacher: What’s the idea about addition and subtraction that’s being revealed here?
Jonah: I think the reason that both of these columns are going up in value is because if you
want to get the same thing if you have a higher number to minus, you need a higher number
to minus it to 25. But if you have a lower number to start with, you don’t need as many
numbers to get to 25.
Many kids: Ohhhhh! I get it!
Teacher: So can you use Todd’s example to talk us through your idea? Todd’s talking about
always wanting to make sure he has 25 cards. Can you use that?
Jonah: If he starts with more, his brother has to take more to get to 25, because there’s more
cards to take.
Frannie: It sounds simple, but it really isn’t.
Manuel: It’s just like. . . he has a bigger number here. So he has to take away more in order
to get to the number he wants to get.
Jamie: That’s what I was going to say.
Sierra: Yeah. . . We knew that, and we thought everyone knew that, but now we just sort of
figured it out.
Helen: Knew what?
Teacher: What is the idea, Sierra?
Sierra: The idea is that you need more to take away and get the same amount. If I had 26
and I minus 1, if you want the. . . That would be the same as if you wanted to have the same
answer, . . . [you] could start with 27 and take away 2.
Addison: The reason why they’re both going up is. . . Since it’s higher, then you have to
subtract more to get to that, but if it’s less, you don’t need as much to get to that number.
It’s less numbers to get to it.
In this episode, Ms. Kaye urges students to clarify what they mean by “it” and
“this” as they are articulating their claim and explaining why they think the pattern
for subtraction differs from the pattern for addition. As they build on each other’s
thinking to articulate why adding the same amount to both numbers in a subtraction
expression results in an equivalent expression, they are simultaneously articulating
more clearly what their generalization about subtraction is.
Developing a Mathematical Argument to Justify

a General Claim
Articulation, representation, and justification of general claims do not occur for

young students in a predictable sequence; rather, they develop together in the course
of students’ work. Representing particular instances of a regularity students have
noticed leads to a clearer description of the claim as well as images of the mathe-
matical relationships and structure that inform justification.
Episode D: Equivalent Addition Expressions (Grade 5)
In the following episode, students work with a representation at the same time that
they are sorting out and stating a general claim. The teacher, Meg Lawson, has asked
the students to write expressions equivalent to 32 using 2 addends. Not surprisingly,
for fifth graders, they come up with many, for example:
16 + 16
30 + 2
28 + 4
10 + 22
15 + 17
Ms. Lawson then writes on the overhead: 16 + 16 = 15 + 17

Teacher: I know you can calculate each side of this equation to find that each side equals 32.
But if you didn’t add each side, how would you know for sure that 16 + 16 equals 15 + 17?
Think about explaining this to someone who couldn’t add up these sums. Show with words
and pictures how you know that 16 + 16 = 15 + 17.
In one small group, Fred, Carlson, and Laura work together.

Laura [very excited]: The total just doesn’t change. One number is just passing one over to
the other number. See, this 16 gave 1 away and became 15 and the other one took it and is
17.
Carlson: I have no idea what you just said.
Teacher: Laura, instead of saying that again, can you show what you mean? Is there anything
you can use or draw that would show what you understand?
Laura puts together 2 sticks of 8 connecting cubes and demonstrates taking one cube from
one stick and putting it onto the other.
Laura: Look. This is 8 plus 8 which is 16. I can take one cube off of this stick and put it on
the other stick and the total is the same.
Carlson and Fred are very quiet so the teacher asks, “Does this help us with
16 + 16 = 15 + 17?”
Fred: Now it shows that 8 plus 8 is the same amount as 7 plus 9. But it works the same way.
One number is smaller and the other number gets bigger.
Laura: And by the same amount! Look—I could move 2 over to the other side and it would
still work!
Teacher: Carlson, what do you think about this? Can you make any sense out of what Laura
and Fred are saying?
Carlson: I think it’s like the stuff is moving back and forth but the whole amount is staying
the same. So you can take some away from an amount and the same plussed to another
amount. . .
Carlson trails off, getting tangled in the words, and Ms. Lawson leaves them while
they are working on the wording for their paper. They are very excited and almost
laughing as they stumble over how complicated the words are. When Ms. Lawson
returns a couple of minutes later, Laura has the cubes out again and is explaining to
Carlson:
Laura: Look, it doesn’t only work for 1 number of change. I can take any amount away as
long as I add it to the other number so the total cubes don’t change.
In her effort to convince Carlson of the generalization she has recognized, Laura has
created a representation that proves that, for any two (whole number) addends, she
can remove some amount from one addend and add it to the other without changing
the total.
The task that the teacher gave her class has several characteristics that we have
identified consistently, across grades, as helpful in engaging students in articulating
and justifying general claims about the operations:
1. The task involves numbers and operations easily accessible to the students.
2. Students are asked to develop explanations about equivalence that do not rely on
computing. (Even if students initially compute to convince themselves, they then
move on to a different way of thinking about justification.)
3. Students are asked to use a representation of the operation as the basis for a
general argument.
With these constraints and requirements, the students in this example began to shift
from talking about specific numbers to talking in general terms. The first indication
of this is how Laura used arbitrary numbers, 8 and 8 changed to 7 and 9, to represent
16 and 16 changed to 15 and 17. It is as if the particular numbers do not matter to
her. At first her choice was confusing to Fred and Carlson, but when the teacher
asked them, “does this help us with 16 + 16 = 15 + 17?”, Carlson talked in very
general terms: “the stuff is moving back and forth but the whole amount is staying
the same. So you can take some away from an amount and the same plussed to
another amount. . . ” By the end of small group time, Laura was able to articulate
the claim clearly in general terms.
Reflecting on this episode, Ms. Lawson wrote: “I wasn’t sure if this question was
going to be interesting to the 5th graders. I wondered if the idea would be so obvious
that they wouldn’t be able to engage. But most kids seemed very excited to show me,
and each other, that they could see and understand what was happening. They looked
like they felt very important as each group had a chance to share their findings.” She
noted that all of the small groups moved from explaining why 16 + 16 = 15 + 17
to a more general argument for any addends, and two groups also realized that the
amount being added/subtracted could be any amount.
Representation-Based Proof: Tools for Proving in the Elementary

Grades
Despite some decades of emphasis on reasoning in national documents, many stu-

dents expect mathematics to be about finding answers. They don’t know what it
means to state a general claim or, if they do, they don’t know what it means to argue
that the claim is true. It is not surprising that younger students might think that a few
examples are sufficient to show that a general claim is true. For example, a second
grader might argue that she has tried changing the order of addends lots of times,
and the sum always stays the same, “so I think it’s true for any numbers.” Many
students, as they progress through the grades, continue to believe that trying many
examples is sufficient to prove a generalization. They never develop an understand-
ing of what it might mean to state something general about how a class of numbers
behaves under a particular operation or to justify such a claim in mathematics. The
use of examples by both students and adults as sufficient proof is well-documented;
even at the college level, many students are satisfied to accept a general claim on
the evidence of a few examples (Harel and Sowder 1998, 2007; Kieran et al. 2002;
Martin and Harel 1989; Recio and Godino 2001).
For example, in a 5th grade class, students have noticed, through many examples,
that if you double one factor in a multiplication expression and halve another factor,
the product remains the same. They have come to accept this idea and routinely ap-
ply it as they solve multiplication problems. However, when their teacher asks them
why doubling one factor and halving another results in an equivalent multiplication
expression, their responses are largely assertions:
Adele: It is the same product because they are equivalent. If you double one factor and halve
the other it will result in the same product because it will stay the same product and not a
wrong product.
Therèse: When you double one number and you halve the other the result is the same prod-
uct because they are equivalents, or the other way to say it is that they are in the same
family.
Gloria: When you double a factor and leave the other alone, the product becomes doubled.
If you double one factor and halve the other factor the product stays the same but if you
double one factor and not halve the other it will be wrong and if you halve both numbers it
will be wrong.
Kamala: There are no limitations to doubling and halving because you can halve any num-
ber to get a whole number or a mixed number and you can double any kind of number. For
example I did 2 × 12 = 4 × 6, and 7 × 5 = 14 × 2.5. I think doubling and halving works
with all numbers.
The students are convinced that halving and doubling will always work to maintain
the same product, but, inexperienced with the kind of question the teacher is asking,
their responses do not move in the direction of justification. Adele and Therèse
assert, correctly, that if the expressions are equivalent, the products must be the
same, but they do not show or describe why the new expression must be equivalent
to the original expression. Gloria is correct that doubling only one factor results in
doubling the product. If she could show why this is true, that could lead her towards
an argument for doubling and halving. Kamala seems convinced that their general
claim can work with both whole numbers and rational numbers, but she offers only
examples to justify her assertions.
These students are typical of students just beginning to justify general claims;
they have no experience with constructing mathematical arguments, but rely on ex-
amples or assertions. Within their statements there are some glimpses of ideas about
multiplication that, if taken further, could lead to more complete mathematical argu-
ments. How can they take the next step towards developing a justification for their
claim?
Students in these grades do not have available to them the tools of formal
proof. What they do have available to them are representations of the operations—
drawings, models, or story contexts that can be used to represent specific numerical
expressions, but can also be extended to model and justify general claims. In or-
der to use representations to make mathematical arguments, students must develop
strong images of the operations, images that embody their properties.
Elsewhere we have described and defined representation-based proof as the
means for elementary and middle grade students to justify general claims (Schifter
2009) by reasoning from visual representations. As students gain experience in ar-
ticulating, representing, and justifying generalizations in the context of number and
operations, they learn to develop pictures, models, diagrams, or story contexts that
represent the meaning of the operations, can accommodate a class of instances (for
example, all whole numbers), and demonstrate, in the structure of the representation,
how the conclusion of the claim follows from the premise.
For example, in the first episode in this paper, second graders developed an ar-
gument for the commutative property of addition, based on a model of two cube
towers. The teacher deliberately introduced that model into the discussion. The third
graders investigating equivalent subtraction expressions used a story context about
baseball cards to explain why one must increase both terms of a subtraction problem
by the same amount to keep the same difference. Students typically begin by rep-
resenting a particular instance of a general claim, then expand it to other instances,
and, finally, modify the representation itself and the language they use to describe it
so that it represents an infinite class of numbers. Fred, Carlson, and Laura use such
general language in describing their cube towers. In the following episode, students
who have been working on making and justifying general claims throughout the
school year develop a representation-based proof.
Episode E: Equivalent Multiplication Expressions (Grade 5)
In the fifth grade described above, students had noticed the doubling/halving rule for
multiplication, but were at the very beginning of work on justifying general claims.
In another fifth grade class, students made the same claim—that if one factor of a
multiplication expression is doubled and the other is halved, the product remains
the same. After investigating and representing individual instances of this claim, the
teacher presented the challenge to prove it:
Teacher: Can you come up with a representation that shows this will always be true, no
matter what numbers you start with? Make a picture, draw a model, but don’t use any
particular numbers.
In response, Trisha and Emily created the following poster.
In their rectangle marked “original,” they represent the multiplication, T × E, as the

area of a rectangle with sides of lengths T and E. In the second picture, they have
cut the rectangle in half and show ½T as a side equal in length to half of T . The
same area (T × E) is equal to two rectangles, each with area (½T × E). By moving
one of the smaller rectangles below the other, as shown in the third picture, they
now have a rectangle with sides ½T and 2E. Since its area (½T × 2E) is equal to
the area of the original rectangle, they have shown that ½T × 2E = T × E.
In later years, students might prove the same claim by invoking the commutative
and associative laws of multiplication together with the multiplicative inverse and
multiplicative identity. At this stage, they use as proof what they understand about
multiplication as represented by the area of a rectangle and conservation of area.
The development of representations for the operations is critical to connecting
arithmetic and algebra. Even students in upper elementary and middle grades who
are fluent with computational procedures may not have developed images of the
operations they can use when they encounter new contexts, for example, making
the transition from using only numerical expressions to using symbolic notation in
algebraic expressions. The use of pictures, diagrams, and story contexts to justify
general claims appears to be accessible, powerful, and generative for elementary
students.
Extending the Number System

In the second grade discussion of the commutative property of addition (Episode A),
the focus was on how addition and subtraction behave differently; one is commuta-
tive, the other is not. But the discussion also allowed ideas about a different kind of
number to be voiced. The second grade teacher made a decision not to pursue the
idea of negative numbers at that time. But as students get older, discussions about
generalizations provide openings for consideration of new kinds of numbers. Does
a property they have justified for whole numbers, and perhaps now take for granted,
still hold when expanding the number system to include fractions, decimals, or neg-
ative numbers?
As they consider new classes of numbers, students sort out which behaviors of
the operations must remain consistent as the number system expands and which only
appear to be general when considering certain classes of numbers. For example,
consider these two statements:
• When you subtract any amount except 0, you end up with less than your original
amount. (For any number b = 0, a − b < a.)
• If you add two numbers to get a third number, then you can subtract either addend
from the sum to get the other addend. (If a + b = c, then c − a = b and c − b = a.)
Students are likely to encounter both of these ideas when their view of numbers is
limited to positive numbers. As their number system expands to include new classes
of numbers, they need opportunities to examine which of the statements are still true.
Students will find that the first statement is not true when b is a negative number,
but the second statement is true for all numbers on the number line. The next two
episodes illustrate students expanding a general claim in this way.
Episode F: Subtracting Negative Numbers (Grade 5)
The next episode illustrates how studies of the operations can support students’ work
on calculation and reasoning about new number domains.
These fifth graders were investigating equivalent subtraction expressions, like the
third graders in Episode C. In this class, students began by generating expressions
equal to 50 such as 70 − 20 and 100 − 50. As in many of these examples, the teacher,
Marlena Diaz, chose numbers that were easy for the students because she wanted the
focus of the discussion to be on the relationships of the numbers, not on computing
results.
The students all knew that the difference in both expressions is 50. Implicit here
is a generalization—add an amount to 70 and add the same amount to 20; the dif-
ference remains unchanged. But how do you know that the difference will always
remain constant when the same amount is added to each number? This is the ques-
tion Ms. Diaz posed to her class.
One fifth grader, Alex, came to the board and drew a number line on which he
showed the distance between 100 and 50 and the distance between 70 and 20.
Then he explained, “You can see that the distance is the same. If you change
one number, you change the other the same way. As long as both numbers change
the same, you can make lots of new expressions.” He was visualizing sliding the
interval, which remains rigid, along the number line so that the beginning and end
points change by the same amount, but the difference between those two points does
not change.
Alex offered a representation of subtraction to justify his claim. Unlike the sec-
ond graders who thought of subtraction as a process of removal, Alex relied on
a different model of subtraction—finding the distance between two numbers on a
number line. Alex’s number line and explanation made sense to other students, and
they realized they could generate many more expressions. Using Alex’s image, his
classmates were thinking of sliding the interval of 50 down the number line and pro-
posed additional equivalent subtraction expressions, which Ms. Diaz listed on the
board:
100 − 50
90 − 40
80 − 30
70 − 20
60 − 10
50 − 0
As this list was being generated, two students commented, as follows:

Patricia: We could keep adding to our list by changing both the numbers, but we are going
to get to a point where we won’t be able to change the numbers. That will happen when we
get to 50.
Nicole: Yes, I agree with Trisha. Because if we look at Alex’s number line we are going to
get to zero and 50, and the jump will be 50, but then we are done.
As the discussion continued, additional ideas were offered.

Raul: But we could use the other numbers.
Teacher: What other numbers?
Raul: The negative numbers on the other side of zero.
Alex: I have one we can use. Let’s use 40 and negative 10.
Teacher: How do you want me to write that on the chart?

Alex: Put 40, then the subtraction sign and then a negative 10.
Ms. Diaz wrote on the board, “40 − −10 = 50.”

In her written account of this class session, Ms. Diaz reported, “At this point
many of the students were talking at once. . . Several were pointing to the large
classroom number line that extends to −40.”
Josh: No way; you can’t do that. How can you have a negative 10 and end up with 50?
Alex: It is like adding 10, because if you look on the number line you would have to jump
50 to get from negative 10 to 40. It is the same as we did with 100 and 50 and 70 and 20.
Teacher: So, Alex, how do you know that 40 minus negative 10 will give you 50?
Alex: Because you have to add 50 to negative 10 to get 40.
In this classroom, the students were discussing a generalization about subtraction

of whole numbers and used a number line to clarify and justify it. Alex could see
that their reasoning about whole numbers could extend to negative numbers. Fur-
thermore, his reasoning brought him to articulate what subtraction of a negative
number must mean. He applied what he understood about the relationship between
addition and subtraction, as well as the image of “distance between” on the num-
ber line, to argue that 40 − (−10) must equal 50. For example, Alex reasoned that
if −10 + 50 = 40, then 40 − (−10) must equal 50. As with whole numbers, if
a + b = c, then c − a = b.
On the other hand, students must reconsider some of the generalizations they
may have made about the behavior of subtraction in the context of whole numbers.
For example, Josh says “No way; you can’t do that. How can you have a negative 10
and end up with 50?” It is likely that Josh and other students hold an implicit belief,
based on their experience with positive numbers, that the result of subtraction (the
difference) is always less than the initial amount (the minuend). Josh may have been
asking, “How can you subtract something from 40 and end up with a number larger
than 40?” The students will need to reconcile these questions with the behavior of
the operations in this new domain.
The ideas brought up in this discussion generated a great deal of interest and pro-
vided the class with the opportunity to think about subtraction of negative numbers
and about the consistencies that should be maintained in the behavior of operations.
Again, as for the students in Ms. Lawson’s class (Episode D) who were consider-
ing 16 + 16 = 15 + 17, a high level of enthusiasm for this kind of challenge was
evidenced.
Episode G: Multiplication with Decimals (Grade 6)
A generalization can help students tie together ideas that at first seem unrelated and
thereby strengthen their understanding of the foundations of arithmetic. In Jeanette
McCorkle’s sixth grade class, students had formulated the same doubling and halv-
ing rule that students were working on in Episode E. They had expanded that claim
to include multiplying and dividing by any factor, not only 2, and had expressed their
claim in symbolic notation as A × B = (A × C) × (B/C). Through the fall, they had
encountered this idea a number of times, but always in the context of whole num-
bers. Like many of the teachers whose work is included in this chapter, the teacher
of this class, Ms. McCorkle, frequently asked students to analyze expressions and
equations without doing any computation. On one day in November, she posed a list
of problems that focused on place value with decimals:
25 × 1 = 2.5 × 10
25 × 10 = 2.5 × ____
25 × 100 = .25 × ____
25 × .1 = 2.5 × ____
25 × .01 = .25 × ____
Teacher: So look at this for a minute [the first equation above] and when you have decided
if that is a true equation, without calculating, when you have a strategy for determining
whether that is true, raise your hand and let me know.
Fran: I’m not sure if this is right at all, but if 2.5 is timesed by 10, it means moving the
decimal over one, and that is the same thing as 25 times 1.
Britta: Well, 2.5 is ten times smaller than 25, and 10 is ten times bigger than 1.
Charles: 2.5 times 10, if you multiply it by 10, you move it one to the right, so you’re
looking at 25 and 25.
Mariah: I would think of 2.5 times 10 as two 10s and a half a 10, which is 25, so you have
25 and 25.
Britta: This is like the problems we did before but A is divided by ten and B is multiplied
by ten.
Britta’s statement surprised Ms. McCorkle. She had not thought in advance about
how this work would connect to the generalization they had articulated in earlier
lessons. She had designed the lesson to address difficulties her students had exhib-
ited with multiplication of decimals. As the class solved the rest of the problems,
some students began by using mechanical methods, counting decimal places. For
example, for the last problem in the group, one student explained:
I’m sort of like, 25 times .01 equals .25 times, it has to be 1, because to get .25 you have to
move the decimal over two, so then to get to 1 you have to move it two the other way.
As Ms. McCorkle interacted and questioned students, she urged them to consider
what moving the decimal point means in terms of multiplication and division. After
some time, she pulled the whole class together to talk about this issue with the goal
of returning to Britta’s earlier observation:
Teacher: You’re looking at a number of decimal places relationship, and I want to expand
that and talk about how one factor has been multiplied and another factor has been divided.
The number of decimal places is just one way of talking about how factors have been altered.
Fran: It’s not just about the decimal point, it’s about multiplying and dividing the numbers.
Teacher: Exactly. I want to remind you about the pattern we were looking at last week and
the week before, when Britta suggested that A × B = 2A × ½B, and George suggested that
A × B = AC × B/C. Britta, how was 25 times 1 changed into 2.5 times 10?
Britta: The first factor was divided by 10 and the second factor was multiplied by 10.
Teacher: That’s right. The decimal moved back means divided by 10, so to maintain this
equality, what should happen to this factor?
Most students: Multiply by 10.
Britta: And it’s still A × B = AC × B/C.
Ms. McCorkle wrote: “By following my students’ thinking, I saw how some of them
connected this page of problems directly to our previous work on the doubling and
halving rule, which I did not expect.”
Considering different kinds of numbers—fractions, decimals, negative numbers
—is an opportunity for students to revisit the generalizations they have worked on
with whole numbers. Through reconsidering these general claims, they identify the
consistencies in the behavior of the operations as the number system to which they
are applying those operations expands. Instead of operating with a new class of
numbers as if they require a new set of rules (e.g., rules about counting or moving
decimal points), they can extend and apply the foundational properties they have
already encountered in operations with whole numbers.
Using Notation with Meaning
In the previous episode, as well as in Trisha and Emily’s proof (episode E), students
expressed general claims in symbolic notation. In the student curriculum we have
developed, we introduce some use of algebraic notation in the elementary grades.
However, we have been careful not to move too quickly. In order to support students’
use of algebraic notation with meaning, they first need to spend a good deal of time
articulating general claims clearly in words and then connecting those statements
to arguments based on representations. The use of phrases that refer to a class of
numbers, such as those used by Carlson in Episode D (“you can take some away
from an amount and the same plussed to another amount”) or second grader Kamika
as she justifies the commutative property of addition in Episode A (“You’re not
adding any more and you’re not taking away any numbers. You’re just changing
them around”) are an important link to meaningful use of symbolic notation.
Using representations and story contexts to model general claims helps students
develop meaning for the symbols of arithmetic. In particular, students’ study of
equivalent expressions, such as 16 + 16 = 15 + 17 (Episode D) or 26 − 1 = 27 − 2
(Episode C) provides the opportunity for meaningful use of the equal sign, signi-
fying equivalence of expressions, rather than “now write down the answer” (Behr
et al. 1980; Carpenter et al. 2003; Kieran 1981). Once students have considerable
experience stating generalizations in words and justifying these general claims by
using representations of the operations, they have images and explanations to which
they can connect algebraic symbols. In our final episode, we see a group of students
making this connection.
Episode H: Using Algebraic Notation to Represent Equivalent Addition

Expressions (Grade 5)
This class of 5th graders has been investigating equivalent addition expressions as
they looked at this sequence:
30 + 2
29 + 3
28 + 4
27 + 5
.
.
.
Students considered a general claim based on this sequence—that if 1 is sub-

tracted from one of the addends and added to the other addend, the sum is
maintained—and developed some arguments to justify the claim. The teacher,
Alina Martinez, introduced a cube representation (similar to what Laura uses in
Episode D) to model addition as joining two quantities. Students talked about how
they could move one cube from one quantity to the other quantity, maintaining the
same sum because “you aren’t adding any or taking any away . . . and since all the
numbers are made up of ones, we can just move all those ones around.” At this
point in the discussion, Ms. Martinez judged that the students’ ideas and images
were quite clear and that symbolic notation would provide another representation
with which they could continue to think about this idea.
Teacher: You all are thinking about lots of numbers and trying to make sense of what is
happening. It seems that you all are thinking that this is true about all numbers and you are
trying to make convincing arguments. I wonder if we could write a sentence that wouldn’t
use numbers to show what is happening. Could we call these numbers up here on the chart
just some numbers?
Will: We could write letters for them. Like n for number, like n one and n two.
Teacher: That’s a great idea. One thing that mathematicians do sometimes is use different
letters so they don’t get confused. How about if we use a and b?
Jonah: We could write a plus b equals a number.
Ms. Martinez then asked students to look at the cube representation of joining two
quantities.
Teacher: So let’s use Jonah’s idea and try to write down what we did to the two quantities.
What are we doing to the a and the b in this pattern?
Ms. Martinez recorded a above the first addend and b above the second addend in
the list of expressions:
a b
30 + 2
29 + 3
28 + 4
27 + 5
.
.
.
Kathryn: We can write a plus b is the same as a take one away and b add one to it.
Teacher [recording a + b = (a − 1) + (b + 1)]: How does this match what Kathryn said and
what we did with the cubes?
Reynold: We take one away like here in the pattern . . . one goes up and one goes down.
Amelia: Oh, look the minus one and the plus one is like a zero! That is why we don’t change
it. It is like staying on zero on the number line.
At this point, many of the children began talking in their groups excitedly. Ms. Mar-
tinez asked the small working groups to consider what would happen if more than
one cube was moved from one quantity to the other. Several students then shared
that they could move two cubes, three cubes, or lots of cubes and still maintain the
sum. Ms. Martinez then asked the class if they thought the notation could be revised
to accommodate this idea.
Adena: We could write lots of them and change the numbers. [Adena is suggesting they
write a series of number sentences, a + b = (a − 1) + (b + 1); a + b = (a − 2) + (b + 2),
and so forth.]
Will: Or we could write add any and take any away.
Jonah: We could use another letter.
Teacher: What do you all think?
Adena: Put a c. Put a plus b equals c.
Jonah: But put the c where the 1 is.
Teacher [recording: a + b = (a − c) + (b + c)]: Do you mean like this, put the c where the
1 is? What does this mean now?
Reynold: See, the c is the cubes you move around to the other side.
In this example, students move among their words, a representation, and the sym-
bols, so that the words and representation are a referent for their thinking about
notation: Kathryn’s words, “a take one away and b add one to it,” becomes
(a − 1) + (b + 1); similarly, “add any” becomes b + c and “take any away” be-
comes (a − c). The teacher asks students to consider how this notation matches
their sequence of expressions and the cube model, and students are able to articulate
these connections, for example, “the c is the cubes you move around to the other
side.”
Introducing this notation at a point at which students have already articulated
their ideas in words and images allows them to maintain meaning for the symbols.
But something else happens as well. Any representation can provide a different
view, a new insight into the mathematical relationships that are represented. The
symbolic representation in this case may make the +1 and −1 even more prominent.
Even though students had noticed that “one goes up and one goes down” as they
considered the sequence of expressions, Amelia now sees something new about this
relationship: “the minus one and the plus one is like a zero! That is why we don’t
change it.” In fact, she has come up with a new argument that involves the fact that
the result of adding 1 and subtracting 1 is 0. Thus, the introduction of symbols in
this case not only provides a concise expression of students’ ideas but offers new
ways of seeing the mathematical relationships.
We don’t want to underestimate the complex issues students encounter as they
begin to work with symbols. The error in which students simply substitute letters
for words in an English sentence is well known, as in writing 6S = P to represent
“there are 6 times as many students as professors” (Clement et al. 1981; Kaput and
Sims-Knight 1983). This incorrect notation stems from using a letter as if it is an
abbreviation for a word, standing for the thing itself rather than the quantity of that
thing, and also perhaps from misinterpretation of the equal sign.
Students need time and experience to develop an understanding of the conven-
tions for using algebraic notation and how the use of letters to represent variables
differs from the use of multidigit numbers. Later in the lesson on multiplying dec-
imals (Episode G) a student tries to rewrite the notation they have been using,
A × B = AC × B/C, to accommodate decimals as: A.B × C = AB × .C. Grounded
in the experience of multidigit numbers and the emphasis in these grades on decimal
computation, it is not surprising that students might think there is a need to have a
letter for each digit, and that the decimal point must be explicitly shown.
Furthermore, when negative numbers or fractions are introduced, students don’t
automatically realize they can use the same letters to represent them. In the equation,
a + b = c, if a represents a negative number, many students think it now must be
written as −a. Because there is no negative sign, a somehow looks positive. It takes
experience to accept that a single symbol might represent a positive or negative
value, a whole number or a fraction.
In all of these cases, students are making sensible choices, based on their ex-
perience with numbers. The transition to use of algebraic notation requires both
connecting these new symbols to what they represent and also learning new conven-
tions. For this reason, even though it may appear easy to make a transition to use of
symbols in particular cases in the elementary grades, teachers’ and students’ expe-
riences indicate that it makes sense to proceed cautiously with early introduction of
algebraic notation.
Connecting Arithmetic and Algebra

The four aspects of early algebra discussed in this chapter have the potential to pro-
vide students with a strong foundation in whole number computation, which they
can extend to their study of fractions, decimals, negative numbers, and algebraic
symbols. Some might ask: How can work in algebra fit into an already crowded cur-
riculum? We would argue that early algebra, defined in this way, not only provides
crucial links between arithmetic and algebra, but also is an essential part of good
arithmetic instruction.
As seen in the classroom episodes, investigation into these aspects of arith-
metic—understanding the behavior of the operations, generalizing and justifying,
extending the number system, and using notation with meaning—provides a means
for students to re-examine and strengthen foundational understandings about the
meaning of the operations and ways of thinking in mathematics.
Further, we are intrigued by the level of student engagement with investigation
of general claims that teachers are seeing in their classrooms. Although it might be
thought that this kind of reasoning is accessible only to “top” students, several of
these examples come from schools in which there is a history of poor performance
on standardized tests. We are accumulating documentation of how both students
who have been relatively successful and relatively unsuccessful in grade-level com-
putation as measured by school and district assessments are engaged by such inves-
tigations (Russell and Vaisenstein 2008; Schifter et al. 2009). Our hypothesis is that
mathematical activities that connect arithmetic and algebra have the potential both
to strengthen the foundations of computation for all students, perhaps especially for
those who have relied on poorly understood procedures, and to intrigue many stu-
dents, including those who excel in mathematics, with challenging questions about
mathematical relationships.
Finally, many questions remain about what teachers need to know and under-
stand in order to carry out this kind of instruction that links arithmetic to algebra.
The isolation in which teachers often work, and the concomitant lack of communi-
cation between elementary and middle grades teachers, is one barrier to the kind of
continuity that might be built in mathematics instruction from arithmetic to algebra.
Elementary teachers need a better grasp of how their curriculum can embody ideas
that are foundational to algebra and how these ideas might be made more explicit
objects of study. Similarly, middle grades teachers need to know more about how
to build on the work of the elementary grades and how to assess the ways students’
conceptions of arithmetic may inform or undermine their understanding of algebra.
Acknowledgement This work was supported in part by the National Science Foundation
through Grant No. ESI-0550176 to Susan Jo Russell at TERC. This work is being undertaken
jointly by Susan Jo Russell at TERC, Deborah Schifter (Education Development Center), Virginia
Bastable (SummerMath for Teachers, Mt. Holyoke College), and a group of elementary and middle
grades teachers. Pseudonyms are used for teachers and students in this article except for those who
have been identified by name in another citation. Any opinions, findings, conclusions, or recom-
mendations expressed here are those of the authors and do not necessarily reflect the views of the
National Science Foundation.
References
Behr, M. J., Erlwanger, S., & Nichols, E. (1980). How children view the equals sign. Mathematics
Teaching, 92, 13–15.
Blanton, M. (2008). Algebra and the Elementary Classroom. Portsmouth, NH: Heinemann.
and Algebra in Elementary School. Portsmouth, NH: Heinemann.
Clement, J., Lochhead, J., & Monk, G. (1981). Translation difficulties in learning mathematics.
American Mathematical Monthly, 88, 286–290.
Harel, G., & Sowder, L. (1998). Students’ proof schemes. In E. Dubinsky, A. Schoenfeld, & J. Ka-
put (Eds.), Research on Collegiate Mathematics Education (Vol. III, pp. 234–283). Providence:
AMS.
Harel, G., & Sowder, L. (2007). Toward a comprehensive perspective on proof. In F. Lester (Ed.),
Second Handbook of Research on Mathematics Teaching and Learning. Reston, VA: National
Kaput, J., & Sims-Knight, J. (1983). Errors in translations to algebraic equations: Roots and impli-
cations. Focus on Learning Problems in Mathematics, 5(3–4), 63–78.
Kaput, J., Carraher, D., & Blanton, M. (Eds.) (2008). Algebra in the Early Grades. New York:
Lawrence Erlbaum Associates.
Kieran, C. (1981). Concepts associated with the equality symbol. Educational Studies in Mathe-
matics, 12, 317–326.
Kieran, E. J., Slaughter, Ml., Choppin, J., & Sutherland, J. (2002). Mapping the conceptual terrain
of middle school students’ competencies in justifying and proving. In D. Mewborn, P. Sztajn,
D. Y. White, H. G. Wiegel, R. L. Bryant, & K. Noney (Eds.), Proceedings of the 24th Meeting
for PME-NA, Athens, GA (Vol. 4, pp. 1693–1700).
Martin, G., & Harel, G. (1989). Proof frames of preservice elementary teachers. Journal for Re-
search in Mathematics Education, 12(1), 41–51.
Recio, A. M., & Godino, J. D. (2001). Institutional and personal meanings of mathematics proof.
Educational Studies in Mathematics, 48, 83–99.
Russell, S. J., & Vaisenstein, A. (2008). Computational fluency: working with a struggling student.
Connect, 22(1), 8–12.
Russell, S. J., Schifter, D., & Bastable, V. (2006). Is it 2 more or less? Algebra in the elementary
classroom. Connect, 19(3), 1–3. http://www.terc.edu/newsroom/776.html.
Russell, S. J., Economopoulos, K., Wittenberg, L., et al. (2008). Investigations in Number, Data,
and Space (2nd ed.). Glenview, IL: Scott Foresman.
Schifter, D. (2009). Representation-based proof in the elementary grades. In D. A. Stylianou, M.
Blanton, & E. Kieran (Eds.), Teaching and Learning Proof Across the Grades: A K-16 Perspec-
tive. Oxford: Routledge—Taylor Francis and National Council of Teachers of Mathematics.
Schifter, D., Bastable, V., & Russell, S. J. (1999). Developing Mathematical Ideas. Number and
Operations, Part 1: Building a System of Tens Video. Parsippany, NJ: Dale Seymour Publica-
tions.
Schifter, D. V., Bastable, V., & Russell, S. J. (2008a). Developing Mathematical Ideas. Number and
Operations, Part 3: Reasoning Algebraically about Operations. Parsippany, NJ: Dale Seymour
Publications.
Schifter, D., Bastable, V., & Russell, S. J. (2008b). Developing Mathematical Ideas. Patterns,
Functions, and Change. Parsippany, NJ: Dale Seymour Publications.
Schifter, D., Monk, S., Russell, S. J., & Bastable, V. (2008c). Early algebra: What does under-
standing the laws of arithmetic mean in the elementary grades. In J. Kaput, D. Carraher, &
M. Blanton (Eds.), Algebra in the Early Grades (pp. 413–447). New York: Lawrence Erlbaum
Associates.
Schifter, D., Russell, S. J., & Bastable, V. (2009). Early algebra to reach the range of learners.
Teaching Children Mathematics, 16(4), 230–237.
The Role of Theoretical Analysis in Developing
Algebraic Thinking: A Vygotskian Perspective
Jean Schmittau
Abstract Vygotsky asserted that the student who has mastered algebra attains “a
new higher plane of thought” (Vygotsky 1986, p. 202), a level of abstraction and
generalization that transforms the meaning of the lower (arithmetic) level. The de-
velopment of this higher (algebraic) plane of thought not only precedes the develop-
ment of arithmetic but becomes a major focus of the child’s elementary education.
It is characterized by orienting children to the most abstract and general level of
mathematical understanding from the beginning of their formal schooling. This ori-
entation to theoretical structure is mediated by the mastery of psychological tools
which are not encountered as incidental to the solution of particular problem types
but are instead the focus of explicit instruction. It is further characterized by the
development of an adequate conceptual base, the incorporation of principles of di-
alectical logic, and the ascent from the abstract to the concrete in the development
of conceptual content.
Introduction
Midway through the first grade curriculum of V.V. Davydov (Davydov et al. 1999)
a child who is using the curriculum in a US school setting solves the following
problem on a teacher made test: “Steven has a bag of jelly beans that weighs d kgs
less than Jennifer’s bag of candy bars. Jennifer has 10 kgs of candy bars. How
much does Steven’s bag weigh?” The child’s answer: “10 kgs − d kgs.” Children
studying Davydov’s second grade curriculum enter their second day of debate on
the problem: “T − 4 − 4 =?” Two opinions have been expressed and neither side
This chapter is a revised version of an article published in ZDM—International Reviews on

Mathematical Education, 37(1), 16–22. DOI 10.1007/BF02655893.
J. Schmittau ()
School of Education, State University of New York at Binghamton, Binghamton, USA

72 J. Schmittau
has been able to convince the other. After an hour both opinions, T and T − 8,
remain as options. Early in the class the next day the argument for T − 8 emerges as
convincing. A seventh grade teacher comments that her students would have no idea
how to solve this problem, and compares the first grade problem above to a multiple
choice question on the state’s eighth grade mathematics test, in which 1 inch of grass
is cut from a lawn k inches tall, with students asked to identify which of four given
expressions represent the remaining height of the grass. Finally, during the second
semester of the Davydov third grade curriculum children solve rate-time-distance
problems while a group of high school teachers observe the lesson, commenting
that “our eleventh graders can’t do these problems”.
What is it in the Davydov curriculum that enables elementary school children
to work with abstractions—so necessary to algebraic understanding—at a level that
proves challenging for many US secondary students? The answer to this question
begins with Vygotsky’s assertion that the student who has mastered algebra attains
“a new higher plane of thought” (Vygotsky 1986, p. 202), a level of abstraction and
generalization that transforms the meaning of the lower (arithmetic) level. In the
early decades of the twentieth century the Russian elementary curriculum focused
on arithmetic, with algebra taught at the secondary level. Vygotsky saw this pro-
gression as a “rise from preconcepts (which the child’s arithmetic concepts usually
are) to true concepts, such as the algebraic concepts of adolescents” (1986, p. 202).
Vygotsky held that only theoretical (also designated “scientific”) concepts were real
concepts; empirical concepts were not true concepts and hence, were designated as
preconcepts above.
Empirical concepts are spontaneously derived from everyday experience, often
by comparing and contrasting the empirical features of objects or phenomena. This
occurs frequently in school settings as well, as for example, when polygons are
compared and classified according to the number of their sides. A theoretical un-
derstanding of polygons, on the contrary, would be oriented to the central role of
the triangle in their genesis, an observation made as far back in history as Aristo-
tle. It is a matter of more than passing significance that mathematics concepts are
quintessentially theoretical (or scientific) in nature (Schmittau 1993).
Scholars who had either influenced (as did the philosopher Hegel) or studied with
Vygotsky, emphasized the role of theoretical thinking early in the child’s develop-
ment. Hegel held that a child should not be kept for too long in an empirical mode of
thinking. And D. B. Elkonin found that the years best suited for theoretical learning
were, in fact, the elementary school years, before interest shifted to peer relations
and a focus on future careers in adolescence (Elkonin 1975). Consequently, in or-
der to achieve the higher plane of thought envisioned by Vygotsky, Davydov saw
clearly the disadvantages of a curriculum in which the order of the development of
concepts was from arithmetic to algebra—an elementary curriculum comprised of
empirical concepts (or preconcepts as Vygotsky called them in the above quote) fol-
lowed by a secondary curriculum where students finally gained access to the realm
of theoretical thought in mathematics.
Davydov and his colleagues sought therefore, to introduce theoretical or alge-
braic thinking earlier in the school experience. At a time when set theory was being
The Role of Theoretical Analysis in Developing Algebraic Thinking 73
adopted as a foundation for the “new mathematics” reform in the United States,
Davydov (1975) observed that set theory was not as fully general as it appeared,
and that this constituted a major drawback to building mathematics on sets as a
foundation. He subscribed to the position held by Bourbaki, who also rejected the
set theoretic foundation, asserting that it is not sets, but rather mathematical struc-
tures that constitute the essential content of mathematics (Bourbaki 1963; cited in
Davydov 1975).
Thus although historically in mathematics and traditionally in education, algebra
followed arithmetic, Vygotskian theory with its emphasis on scientific concepts and
theoretical understanding, supports the reversal of this sequence in the service of
orienting children to the most abstract and general level of understanding from the
beginning of their formal schooling. However, given that elementary school children
do not possess the sophisticated understandings of mathematicians or even the arith-
metical knowledge of secondary students, it was by no means obvious how instruc-
tion might be designed to render algebraic structure preeminent at the elementary
level, without imposing a conceptually sterile and largely unlearnable formalism.
The fact that the algebraic structure of positive scalar quantities is shared by the
real numbers became the key to maintaining a theoretical focus while at the same
time allowing for the accommodation of children’s learning needs. Children could
study scalar quantities such as the length, area, volume, and weight of real objects,
which they can access visually and tactilely, discern their properties, and in this
way equivalently access the algebraic structure of the real number system. This is
the approach taken by Davydov’s elementary curriculum, which stands as a major
departure from conventional programs. By developing number from the measure-
ment of quantities, Davydov’s curriculum also breaks with the common practice of
beginning formal mathematical study with number. Observing that culturally and
in individual development, the concept of quantity is prior to that of number, he
indicted the rush to number as a manifestation of ignorance of the real origins of
concepts (Davydov 1990). In his first grade curriculum, so extensive is the founda-
tion of investigation of (mostly continuous) quantities that number does not appear
until the second semester.
The Davydov curriculum represents a departure from Piaget’s proposed stages
of development as well. Vygotsky, however, accepted neither Piaget’s separation of
instruction and development, nor the assumption that the latter had to precede the
former. Rather, Vygotsky contended that learning, and in particular, the mastery of
scientific concepts, leads development. He wrote, “The formal discipline of scien-
tific concepts gradually transforms the structure of the child’s spontaneous concepts
and helps organize them into a system: this furthers the child’s ascent to higher
developmental levels” (Vygotsky 1986, p. 206).
An orientation to algebraic structure requires a focus on the most general and ab-
stract characteristics of real phenomena, beginning with children’s initial classroom
encounters with such phenomena. This, in turn, requires the development of volun-
tary attention. And as is the case with scientific concepts and theoretical learning in
general, pedagogical mediation is necessary.
This is accomplished in Davydov’s curriculum by focusing children on the the-
oretical characteristics of real objects, objects with which they are familiar, asking
74 J. Schmittau
them to compare such objects with respect to their length, area, volume, or weight,
and to progressively refine such comparisons until they culminate in measurement
itself, from which number is then defined. Children are confronted with tasks in
which they determine that when adding a volume A to a volume B, for example,
the result is the same as when the order of addition is reversed. Repeating this task
in combining two different lengths of wood, or adding the weight of a pinecone and
a pattern block in either order, children are developing the commutative property
of addition of positive scalar quantities. Since they do not know the actual length
of the wood or weight of the objects (that would require measurement, which will
come later), the children label their result with letters, such as T + C = C + T ,
with the understanding that such a result is generalizable to any two quantities. It
is noteworthy that initially the mere prizing out of the quantity from the empirical
features, such as the shape and color of the objects in question, is the beginning of
a theoretical orientation to the task.
Once children’s comparisons have progressed to the actual measurement of quan-
tities (through the laying off of a part of the quantity arbitrarily designated as a unit),
a number is generated that is the measure of the quantity by the unit employed. The
properties of quantities apply also to their numerical designations, which vary with
the unit of measurement. Children have already established these generalized prop-
erties for any quantities, hence their extension to specific numerical designations
of quantities is a concrete application of a general result previously obtained and
expressed in the symbolism of algebra. Now algebra is no longer initially learned
as a generalization of arithmetic, but rather as a generalization of the relationships
between quantities and the properties of actions on quantities. With the introduction
of measurement and the definition of number emanating there from, the application
of the properties of quantities to their numerical designations represents the ascent
from the abstract to the concrete, whereby the abstract essence of a concept is dis-
cerned in its concrete (here numerical) embodiments, which “fill in” and enrich, as
it were, the conceptual content.
In the Davydov curriculum the development of the higher plane of thought al-
luded to by Vygotsky not only precedes the development of arithmetic but becomes
a major focus of the child’s elementary education. Further, throughout their three
years of elementary schooling children are oriented to the highest levels of abstrac-
tion in their understanding of mathematical concepts. How this is accomplished and
the role of psychological tools in this process is the subject of the next sections.
Orienting Children to Theoretical Concepts

The first grade curriculum introduces number after a semester of work on compar-
ison of quantities and measurement. Considerable time during the second semester
is spent on developing the concept of positional system through the development
of many different number bases. This serves to establish a sufficiently generalized
conceptual base for operations with whole numbers, and has foundational applica-
bility to decimals and polynomials as well. It is difficult not to notice the theoretical
level at which children are challenged to understand the material. Consider, for ex-
ample, the following problems which reflect the level of abstraction and theoretical
thinking expected of children studying the first grade curriculum.
A G
(1) =4 =6 A?G
C C
B K
(2) =7 =? B <K
T T
(3) Andrew had c stamps. After he was given some as a gift, he had g stamps.
How many did he receive as a gift?
(4) There is a vase of flowers. r flowers died. t flowers are still fresh. How many
flowers were in the vase originally?
(5) Write the number for each expanded form below:
0 + = 00 + D = 00 + D0 + =
(6) Locate the numbers m, n, and v on the number lines below.
___.___.___.___.___.___.___.___.___.___
m+2
___.___.___.___.___.___.___.___.___.___
n−3
___.___.___.___.___.___.___.___.___.___
v+1
(7) Compare the following numbers (by writing <, = or > in the blank):
D___D 0, 0___00, 0 __ 0
(8) Subtract the following numbers:
− 0 = ? 0 − = ? DDD − D00 = ?
(9) Daniel had some pencils. His friend gave him another 6 pencils. Now Daniel
has 15 pencils. How many pencils did Daniel have originally? Which of the
formulas below fits this problem?
a +b=c a+b=c a+b=c
15 6 x x 6 15 6 x 15
(10) Make up two problems about pencils to fit the two formulas given in problem
#9 other than the one that matches Daniel’s situation.
The reader is invited to actually solve the above problems in order to experience
the level of abstraction and mathematical reasoning required and to notice one other
very important element, viz. the manner in which they require the student to nego-
tiate cognitive reversals and to consider a concept from virtually every angle. Note
the different ways in which children are required to think and analyze in #1 and #2
(more than one solution is possible for #2, of course), and in #9 and #10. Note also
the differing word orders in #3, #4, and #9. So-called “key word” approaches would
be of little use with word problems such as these, nor can students rely on context
cues gleaned from numbers. Problem #10 asks children to make up word problems
to fit the equations, which requires them to reverse the perspective from which they
analyze the problem situation. Students are not given collections of routine prob-
lems all of which can be worked in the same way; they are challenged to think and
analyze, and to look at problem situations from different perspectives. In this sense,
76 J. Schmittau
none are routine. And children must understand the structure of the number line in
order to be successful with problem #6, since no beginning of the number line is
designated, nor are there any numerical designations marking the points. Similarly,
they must understand the structure of a positional system in order to correctly solve
#5, #7, and #8, since there are only place holder (zero) and literal designations of
positions. When the author confronted a group of forty middle school teachers with
just a few of the above problems, they found especially those similar to #5, #7, and
#8 so difficult that many of the teachers gave up trying to solve them. Yet these are
representative of problems found in the first grade curriculum of Davydov.
Role of Psychological Tools

There is, however, another characteristic of the Vygotskian approach to the develop-
ment of algebraic thinking that is also of critical importance for student mastery. It
is the mediational role of psychological tools in the appropriation of theoretical un-
derstandings. This is exemplified by the representational “schematic”, the mastery
of which is not merely incidental to but the explicit focus of instruction. According
to Vygotsky, psychological tools, unlike their material counterparts that are directed
toward action on the outer environment, are directed inward toward the control of
one’s own behavior. What is the direction that the schematic gives to the problem
solver? It orients the child’s attention to the internal relations among quantities,
rather than to the empirical features of the problem or familiar aspects of numbers.
By the time children reach the second grade, they generally have little need of ac-
tual objects (or “manipulatives”), preferring to work instead with the “schematics’
that focus them on, isolate, and express the mathematical actions in which they are
engaged. These schematics are representations, but are not pictorial in nature as is
often the case when US children model problem situations.
The Part-Whole Relation
An important schematic introduced in the first grade curriculum of Davydov is the

“/\” representation for a part-whole relationship. Early in the first grade, for exam-
ple, children determine that they can make two unequal volumes equal by adding to
the smaller or subtracting from the larger the difference between the original quan-
tities. If A > B, and the difference between volumes A and B is volume C, then the
relationship among the three quantities can be modeled by the part-whole schematic
in which A is the whole and B and C are the parts:
A
/ \
/ \
B C
The children then determine that if volume A is greater than volume B, A =

B + C, where C is the difference between A and B. They may schematize their
result with a “length” model, and symbolize it with three equations:
A>B A > B by C
A=B +C
B =A−C
C =A−B
In solving problem #3 that appears in the above list of problems, children would
first analyze the problem structure, identify it as a part-whole structure, then de-
termine that c, the number of stamps Andrew had originally, is a part and that g,
the number of stamps he had after receiving more, is the whole. They would then
model the quantitative relations expressed in the problem using the “/\” schematic,
placing g at the top of the “/\” and c and ? at the bottom. Since the unknown part
can be found by subtracting the known part from the whole, they set up the equation
x = g − c. It is immediately obvious that this is a missing addend problem whose
equation g = c + x, set up directly from the problem sequence, suggests addition.
However, it is the schematic and its relationship to the equation that enables children
using Davydov’s curriculum to avoid the error suggested by the sequence of (addi-
tive) actions performed in the problem situation and to correctly subtract the part
from the whole. Note that “counting on”, a common strategy employed by children
who learn arithmetic before algebra, is not an option where the numerical values of
quantities are unknown, as is the case in an algebraic formulation wherein quantities
have only literal designations.
In the Davydov curriculum, the situation that occurs in problem #3 above is ex-
plicitly addressed within the context of cutting a piece of wire from a coil. Here
the action is always the same. But three separate problems can be formed from this
action, each of which is simply delineated by an equation below:
B−T =R D − A = N and G − C = M
x 9 6 32 x 17 21 9 x
In each case what is asked for is the number of meters of wire, which requires
finding the value of x in each of the three problems. After completion of the third
and final problem, it is noted that in all three of the exercises above the same action
78 J. Schmittau
was performed, i.e., some wire was cut from a coil. Children must then determine
why they sometimes added and sometimes subtracted in finding the amount of wire.
The action on the wire in the coil was always to cut off a piece of the wire as
each of the three literal equations indicates. However, the children must reason that
in the case of the first problem (B − T = R), it is the amount of wire originally
in the coil that is unknown, while the measures of the part removed and the part
remaining are known. Consequently, they must add the parts to obtain the original
length of wire in the coil prior to cutting off a piece. In the two subsequent cases,
the part removed and the part remaining are to be found. Since the whole is known,
subtraction is called for. And it is, of course, precisely this distinction that is relevant
to problems #3, #4, #9, and #10 in the list of problems appearing above. The wire
cutting problem renders explicit the difference between the action performed within
a problem context and the action required to obtain the solution to a problem arising
from within that context.
In addition, a fundamental dictum of dialectical logic is reflected throughout the
curriculum, viz. that “the essence of a thing can be revealed only by considering the
process of its development” (Davydov 1990, p. 288). Hence, the origin of number
is traced back through its development from the progressive refinement of compari-
son of quantities to their measurement. Number does not appear “fully formed” as it
were, as is typical in conventional textbook treatments. Nor do addition and subtrac-
tion appear as separate operations, but rather as dialectically interrelated actions that
arise from the part-whole relation between quantities. The “/\” schematic serves to
both focus children’s attention on this relation and provide a means to represent it at
an intermediate stage between its direct observation and its transformation into an
inequality or equation.
Not infrequently, children are confronted with problems that are unsolvable. In
our implementation of the Davydov curriculum (to our knowledge a first in the
United States), children were asked to solve individually the following problem:
“There were seven books on the bookshelf. Nine children entered the library and
each child took a book from the shelf. How many books are now on the shelf?” None
of the children answered “16” or “2”. Instead they recognized that a part cannot be
larger than the whole and all identified the problem as a “trap” (i.e., unsolvable).
When asked to change the problem into one that had a solution, the following types
of changes were made by the children.
(A) There were seven books on the shelf. Five children entered the library and each
removed a book from the shelf. How many books are now on the shelf?
(B) There were 27 books on the shelf. Nine children entered the library and each
took a book from the shelf. How many books are now on the shelf?
(C) There were seven books on the shelf. Nine children entered the library and each
placed a book on the shelf. How many books are now on the shelf?
The children recognized that either the whole must be increased (B) or the part
decreased (A) (or both), or the action on the objects must be changed. In (C), for
example, both the books initially on the shelf and those added by the children, are
now parts; whereas in the other formulations the first was the whole and the second
was a part.
One may again discern in these problems an ascent from the abstract to the con-
crete. From the three equations generated from problem situations such as that aris-
ing from the comparison of two volumes (described above), children discern their
application to the part-whole situation not only in problems such as #3, #4, #9,
and #10 in the list provided above, but also to the specific numerical designations of
quantities in problems such as the book problem, where they must reason in reverse,
as it were, that the whole must be greater than the parts in order for the problem to
have a solution. Because they have a theoretical orientation to the problem structure,
they can analyze the relationships between quantities in problems without any nu-
merical designations (such as problems #3 and #4 in the above list) to provide them
with cues. And in the book problem they are not led by the numerical aspects of the
problem to simply add the two numbers or subtract the smaller from the larger.
Another example of orientation to theoretical understanding is exemplified by
a problem encountered early in the second grade curriculum, and mentioned in the
opening paragraph of this chapter, viz.: “What is T − 4 − 4?” After much discussion,
the classroom debate centered on two possible answers: T − 0 (= T ) or T − 8.
Neither side in the discussion could convince the other of their point of view. Finally,
as the debate entered its second day, the teacher asked a child who argued that the
answer was T − 0 to mark the parts and the whole in the statement: 14 − 4 − 4 = 14
(since the child had argued that T − 4 − 4 = T ). The child marked the 14 as the
whole in both places where it appeared in the statement of equality, and marked
each 4 as a part. The teacher then asked the child to mark the whole and the parts
in the statement: P − A − B = C. Here the child correctly marked P as the whole,
and A, B, and C as the parts, and then drew the following schematic:
P
/|\
A B C
When the child realized that a similar schematic of the earlier answer would have
to show 14 as the whole and 4, 4, and 14, as the parts, the child realized that in the
original answer 14 had been treated simultaneously as the whole and also as one of
the parts, and at the suggestion of classmates, changed the parts to 4, 4, and x (Lee
2002). What is striking here is that the child did not see the error until the relation-
ship between the quantities was expressed algebraically in the abstract. The numbers
in the problem had been the source of the error; the numbers had misled the child
and the abstract algebraic representation of the mathematical structure corrected the
error.
A negative illustration of the role of such psychological tools occurred in the
case of a US child who was very quick with numerical computation, and refused to
use schematics once numbers were introduced, protesting that they were unneces-
sary and cumbersome. This child was able to solve one-step word problems fairly
well, but when two-step problems were introduced, the child “just picked numbers
and calculated” (Lee 2002), ignoring the internal relationships among the quantities,
and making many errors. And when no numbers were present, as in the following
problem, the child’s resulting reduction to “A + B” was predictable: Tanya picked
80 J. Schmittau
A mushrooms. Her grandmother picked B mushrooms more than Tanya picked.

How many mushrooms did Tanya and her grandmother pick? (Davydov et al. 2000,
p. 54). What is significant is that this child showed considerable ability to reason
mathematically with simpler first grade tasks, but as a result of failure to master the
important semiotic tools of analysis, was prone to error on tasks of greater com-
plexity and consequently this capability did not develop along the lines of its initial
promise.
By directing the focus to the underlying theoretical structure of a mathematical
action or problem situation, the psychological tool serves to unite empirically dis-
parate actions or phenomena. Addition and subtraction constitute a case in point,
since the “/\” schematic reveals the part-whole structure common to both. This
same psychological tool functions in the analysis of the underlying part-whole struc-
ture of so-called “fact families”, the addends and sums that primary school children
learn, and the part-whole structure of numbers represented in positional systems.
Consequently, it figures prominently in children’s ability to correctly answer ques-
tions #5, #7, and #8 in the list of problems above, without any need for recognizable
numerals in the numbers they are asked to compare. The numbers can be compared
simply by discernment of the theoretical (part-whole) structure of a number and its
designation within a positional system. And later when children encounter equations
such as 17 + 5x = 42, they are by third grade generally able to work them without
the use of the “/\”, and simply write 5x = 42 − 17, then 5x = 25 and x = 5. And
an equation such as 324 − x = 13 is quickly transformed into x + 13 = 324 and
x = 311. It is the child’s orientation to theoretical structure, which has now largely
become internalized through the mastery of the requisite psychological tools, that
figures prominently in enabling the algebraic concepts to transform the meaning of
the lower (numerical and empirical) concepts as Vygotsky indicated.
The Table as Psychological Tool
Rather than being directed to “make a table” the first time the need for one arises
in the process of solving a particular type of problem, the construction of a table
evolves through considerable stages beginning early in the third grade curriculum.
Here again, as in the case of the “/\” schematic, a table is a psychological tool,
and its construction is the subject of explicit exploration and instruction. And as is
typical of the Davydov curriculum, every aspect of its construction is explored.
While doing research in Russia, the author watched beginning third graders solve
problems such as the following:
Group the quantities below in several ways.
A = the height of the tree
B = 16 kg
C = the weight of the board
D = the age of the tree
E = 20 cm
M = the area of the board
H = 16 seconds
K = 24 meters
T = the length of the rope
Some children would group B, E, H, and K together under the designation
“Known” and place all the other quantities under the designation “Unknown”. Oth-
ers would choose to group A and D under the designation “Tree”, C and M under
“Board”, and T under “Rope”, then group the other data under “Numbers”. A third
grouping might be comprised by A, E, K, and T grouped under “Length”; M under
“Area”; B and C under “Weight”; and D and H under “Time”. The children might
then set up tables with the designated headings in a row across the top and the ap-
propriate quantities listed under each heading.
As is typical of the Davydov curriculum, not only are children asked to solve
problems such as the one above, but they are also given tables in which one or more
numbers or letters appear under the designations “Time”, “Weight”, and “Length”,
and they are asked to generate a list containing a description of quantities that fit the
categories in the table. A similar table might have categories such as “Elephant”,
“Recess”, “Lesson”, and “Log”, and a third might contain only the two categories
of “Known” and “Unknown”. As with the first table, children are asked to supply
a list of quantities that fit the categories in the last two tables. Thus they are re-
quired to think flexibly and to reverse their perspectives in solving these two types
of problems.
From here children progress to more open ended problems that ask that they
make up a list of quantities and group them according to property, object, and
whether the quantities are known or unknown. It will be immediately noted that
when working problems such as those above, these (i.e. property, object, and
known/unknown) are the general categories. In the first problem, for example,
“Length”, “Area”, “Weight”, and “Time”, are properties, whereas “Tree”, “Board”,
and “Rope” are objects. Eventually it will be seen that in mathematics we are in-
terested in the properties of quantities, and there will be tabular groupings such as
are found in rate, time, and distance problems, wherein properties such as “Time”
and “Distance” are listed horizontally in a row across the top of a table while the
objects to which they pertain—Train #1 and Train #2, for example, or Bicycle and
Runner—will appear in an adjacent vertical column. What is important to notice
here is the thoroughness with which the children explore initially the element of ar-
bitrariness in the designations and arrangements of a table, so that they can flexibly
make use of this knowledge whenever they choose to construct a table to use as a
psychological tool to organize the information given in a problem in a way that will
assist them in its solution. At this point, however, they are asked only to generate
the table, and this is illustrative of the manner in which the psychological tool itself
becomes a matter of intentional and explicit instruction rather than being introduced
as simply incidental to obtaining the solution to some problem.
The use of a table as a psychological tool to solve a problem will come later. In
the interim children will be asked to make tables simply from narratives with no
problem formulated and hence, no solution called for. For example, they may be
asked to construct a table to organize information such as the following: “There are
82 J. Schmittau
20 oranges in a box, and there are 35 oranges in another box.” Such a problem intro-
duces the need to designate both rows and columns, and they may place “First Box”
and “Second Box” in a column under what is obviously an object designation, and
list “20 pieces” and “35 pieces” respectively in a second column under the heading
“Quantity”. This may be followed by a similar but slightly different narrative, viz.
“There are 20 kg of oranges in one box and 35 kg of oranges in another one.” Now
they may use the same first column in their new table as in the previous one, but list
“20 kg” and “35 kg” respectively in the second column under the category “Weight”.
Finally, the problem may be changed as follows: “There are 20 kg of oranges in one
box and 35 oranges in another.” Now while retaining the list of boxes in the first
column, the children will have to discern the need to create two additional columns,
one containing “20 kg” under the heading “Weight” and the other containing “35
pieces” under the heading “Quantity”. These are three simple narratives, but they
require children to pay close attention to the properties involved and how they need
to be dealt with so that a table will accurately reflect the categories of its members.
Eventually a question will be attached to a simple narrative in a problem such
as the following: “A small melon weighs m lbs. A watermelon is heavier than the
melon and weighs n lbs. By how many pounds is the watermelon heavier than the
melon?” The accompanying instruction directs the students to build a table for the
problem. It will be immediately obvious to the reader that this is a problem that chil-
dren could readily solve in the first grade using the “/\” schematic to discern that the
weight of the melon was only a part of the weight of the watermelon, but the prob-
lem directs the children to construct and use a different psychological tool, namely
a table. Requiring the use of a table for the solution to such a simple problem paves
the way for its use in the solution of more complex problems that will be introduced
later, such as those involving proportional reasoning. It allows the child to continue
to focus mainly on the building and role of the newly introduced psychological tool
(since s/he can by now easily solve the problem without it and is therefore not con-
sumed with the need to focus on the solution), while at the same time introducing
the possibility of applying the tool to the solution of a mathematical problem.
As they progress, children may be asked to build tables for two statements such
as the following: “(A) A worker made 28 widgets before lunch and 42 widgets after
lunch. (B) A worker made 28 widgets in 3 hours and 42 widgets during the next 3
hours.” Now they must deal with variable quantities in the form of the number of
widgets and in statement (B) also with time. They must identify not only the vari-
able(s), but the process and the separate events occurring during the process. Here,
of course, the process is the making of widgets, and there are two events—before
lunch and after lunch in the case of statement (A), and the first event (consisting
of 2 hours) and the second event (consisting of 3 hours) in the case of statement
(B). Children note that during the process of work, time and the amount of work
changed. Consequently, these are the variables.
The children may build a table for statement (A) consisting of just two columns,
viz. “Event” and “Quantity”. Under “Event” they might list “I” and “II” referring to
the first and second time periods, respectively, and “I + II” or “Total” for the third
and final entry. And under “Quantity”, entries 28, 42, and __, would appear opposite
Events I, II, and Total, respectively. In the case of statement (B) above, the students
would need to add a third column, viz. “Time (hours)”, under which they would list
2, 3, and 5. Now the children might be asked to formulate a question for each of the
above statements. Clearly, the question “How many widgets did the worker make
altogether?” would be appropriate for both statements (A) and (B).
In a problem requiring a reversal in their perspective, children might be given
a table with the Process of Buying, and Events listed as First Purchase, Second
Purchase, and Total Purchase. The corresponding amounts (in dollars) might be
listed as c, t, and c + t. Another column of amounts (in numbers) might list 3, 1,
and 4, respectively. The children would then be asked to make up a text to fit the
table.
Eventually, a table such as the following will be encountered. Two columns,
one representing the variable “Weight (lbs)”, and another representing the variable
“Number of Parts”, will be presented. The following entries appear under “Weight
(lbs)”: 12, 24, 72, and 60. The corresponding entries under “Number of Parts” are
4, 8, 24, and 20. In a second table the entries under “Weight (lbs)” are the same as
in the first table, but now the corresponding entries under “Number of Parts” are
4, 16, 32, and 24. Children are to name the process and both observe and discuss
what is different concerning the variable listings in the two tables. Eventually they
discern that in the first table as one of the variables increases several times, the other
increases the same number of times. Such a process is called a uniform process and
the variables are called direct proportional variables. The process described by the
second table is not uniform.
The children have now been formally introduced to direct variation and later in
the third grade they will use the knowledge they have acquired of building and using
a table as a psychological tool to solve problems such as the following. “The speed
of a car is 60 mph and the speed of a bicyclist is 4 times less. The bicyclist can travel
from his house to the nearest city in 2 hours. How long will it take to go the same
distance in the car?” The children also are able to construct a table to assist them in
determining how many runs must be made by a truck and a trailer in order to trans-
port 1080 tons of coal, if the truck can haul 30 tons per load and the trailer can carry
twice as less coal per load (cf. Schmittau 2004). These are but two examples of the
variety and types of problems for which children might choose to construct a table
as an aid to finding a solution. Their extensive experience with situations in which
tables may serve to organize information and to focus and direct attention to the
problem structure and the relevant processes, events, and variables involved, render
the table a flexible tool that can be of use in a wide variety of problem situations.
This includes but is not limited to problems such as those involving rate, time, and
distance, differential rates of work, and other sorts of problems involving propor-
tional reasoning. Children may even construct a table to find missing dimensions of
geometric figures for which only literal data are provided (Davydov et al. 2001).
In the manufacture of widgets problem it will be noted that the two events listed
are parts of the whole event which is the total time worked. Consequently, the table
might be thought of as embodying the part-whole structure of the “/\” schematic.
This is the case as well in the melon problem preceding it, and some of the more
84 J. Schmittau
difficult problems that follow. The table embeds this structure within itself, as a
psychological tool capable of accommodating the greater conceptual complexity
that occurs in problem situations involving processes with multiple variables.
Concluding Remarks
In our investigation of Davydov’s elementary curriculum in what is, to the best of

our knowledge, its first implementation in the US, we found that the US children
evinced abstract and generalized understandings not unlike those the author ob-
served among their Russian counterparts. (For information additional to that pre-
sented in this chapter cf. Schmittau 2004.) They were aided in developing their
knowledge of algebraic structure by the schematic models, which functioned as
“helpers” in the children’s language and as psychological tools within the frame-
work of Vygotskian theory (Vygotsky 1986).
By the end of the third grade curriculum, the children studying Davydov’s pro-
gram were solving applied problems involving proportional reasoning, differential
rates of work, and rate, time and distance, that continued to challenge eleventh
graders in their second year of formal algebra study in their regional high schools.
Moreover, although Davydov’s third grade curriculum (the final year of Russian el-
ementary school) consisted of 969 problems (each year of the curriculum was com-
posed entirely of very deliberately designed and sequenced problems), the class’s
teacher reported that students had no difficulties with the last 400 problems, despite
the fact that objectively the level of difficulty of the problems themselves had in-
creased considerably as indicated above. In fact, the more complex and difficult the
problems became, the less difficulty the children had in solving them, and the more
rapidly they progressed through them. This attests to the value of laying a broad in-
depth conceptual foundation (extensive work with quantity as a concept antecedent
to number) with a focus on theoretical (i.e., algebraic) structure, and to the mastery
of psychological tools—two of which are discussed above—that function to orient
students to that structure. Their appropriation of these psychological tools which
are not encountered merely incidental to but are the explicit focus of instruction as-
sist children in their engagement with “the formal discipline of scientific concepts”
(Vygotsky 1986, p. 206).
It is worth noting also that the approach taken by the Davydov curriculum, fol-
lowing as it does Vygotskian theory and dialectical logic, is far removed from, for
example, the simple introduction of equations into the elementary school mathemat-
ics curriculum. Teaching a method of solving equations by, for instance, “balancing”
both sides of the equation, while it provides a means of obtaining a solution to the
equation, does not ensure understanding of the theoretical essence of the relation-
ship between the quantities represented, both known and unknown, and how that
relationship arises from the progressive comparisons of quantities found in objects,
through measurement and the origins of number, and subsequently finds expression
in a variety of mathematical actions and situations. Similarly, the introduction of
a table as a means to solve a particular type of problem fails to exploit its effec-
tiveness as a psychological tool that can be flexibly applied wherever it is useful to
organize information in the service of directing attention to the theoretical structure
of a problem situation and its mathematical interrelationships. Such episodic intro-
ductions into the curriculum also fail to comply with the central tenet of dialectical
logic concerning the necessity of tracing a concept through its entire developmental
path as a necessary means to ensure its understanding.
In summary, the Vygotskian imperative concerning theoretical understanding as
essential for the development of algebraic thinking is reflected throughout the el-
ementary curriculum of Davydov. It is the mastery of theoretical concepts which
“gradually transforms the structure of the child’s spontaneous concepts and . . . fur-
thers the child’s ascent to higher developmental levels” (Vygotsky 1986, p. 206).
References
Bourbaki, N. (1963). Essays on the History of Mathematics. Moscow: IL, trans. from the French.
Davydov, V. V. (1975). Logical and psychological problems of elementary mathematics as an aca-
demic subject. In L. P. Steffe (Ed.), Soviet Studies in the Psychology of Learning and Teaching
Mathematics (Vol. VII, pp. 55–107). Chicago, IL: University of Chicago Press.
Davydov, V. V. (1990). Types of Generalization in Instruction: Logical and Psychological Prob-
lems in the Structuring of School Curricula. Reston, VA: National Council of Teachers of Math-
ematics (Original published in 1972).
Davydov, V. V., Gorbov, S. F., Mikulina, G. G., & Saveleva, O. V. (1999). Mathematics: Class 1.
J. Schmittau (Ed.). Binghamton, NY: State University of New York.
Davydov, V. V., Gorbov, S. F., Mikulina, G. G., & Saveleva, O. V. (2000). Mathematics: Class 2.
J. Schmittau (Ed.). Binghamton, NY: State University of New York.
Davydov, V. V., Gorbov, S. F., Mikulina, G. G., Savyelyeva, O. V., & Tabachnikova, N. L. (2001).
Mathematics: 3rd Grade. J. Schmittau (Ed.). Binghamton, NY: State University of New York.
Elkonin, D. B. (1975). Primary schoolchildren’s intellectual capabilities and the content of in-
struction. In L. P. Steffe (Ed.), Soviet Studies in the Psychology of Learning and Teaching
Mathematics (Vol. VII, pp. 13–54). Chicago, IL: University of Chicago Press.
Lee, J. (2002). An analysis of difficulties encountered in teaching Davydov’s mathematics cur-
riculum to students in a U.S. setting and measures found to be effective in addressing them.
Binghamton, NY: State University of New York at Binghamton, Doctoral Dissertation.
Schmittau, J. (1993). Vygotskian scientific concepts: Implications for mathematics education. Fo-
cus on Learning Problems in Mathematics, 15(2–3), 29–39.
Schmittau, J. (2004). Vygotskian theory and mathematics education: Resolving the conceptual-
procedural dichotomy. European Journal of Psychology of Education, XIX(1), 19–43.
Vygotsky, L. S. (1986). Thought and Language. Cambridge, MA: MIT Press (Original published
in 1934).
The Arithmetic-Algebra Connection:
A Historical-Pedagogical Perspective
K. Subramaniam and Rakhi Banerjee
Abstract The problem of designing a teaching learning approach to symbolic al-

gebra in the middle school that uses students’ knowledge of arithmetic as a starting
point has not been adequately addressed in the recent revisions of the mathematics
curriculum in India. India has a long historical tradition of mathematics with strong
achievements in arithmetic and algebra. We review an explicit discussion of the re-
lation between arithmetic and algebra in a historical text from the twelfth century,
emphasizing that algebra is more a matter of insight and understanding than of using
symbols. Algebra is seen as foundational to arithmetic rather than as a generalization
of arithmetic. We draw implications from these remarks and present a framework
that illuminates the arithmetic-algebra connection from a teaching-learning point of
view. Finally, we offer brief sketches of an instructional approach developed through
a design experiment with students of grade 6 that is informed by this framework, and
discuss some student responses.
Introduction
Mathematics is widely believed in India to be the most difficult subject in the cur-
riculum and is the major reason for failure to complete the school year in secondary
school (National Centre for Educational Research and Training 2006). The edu-
cation minister of a western Indian state recently complained that students spend
vast amounts of time studying mathematics, with limited success and at the cost
K. Subramaniam ()
Homi Bhabha Centre for Science Education, Tata Institute of Fundamental Research, Mumbai,
India
R. Banerjee
School of Social Science, Tata Institute of Social Sciences, Mumbai, India

88 K. Subramaniam and R. Banerjee
of neglecting other subjects and extracurricular activities. Similar complaints pres-

sured the state government into removing the mandatory pass requirement in math-
ematics for the school exit examination in the year 2010. This is reflective of a
trend among some school systems in India to make mathematics an optional subject
in the school exit examination. Students’ difficulties in mathematics may however
have deeper causes located in the education system as a whole, which need to be
addressed on multiple fronts. The nation wide annual ASER surveys, based on rep-
resentative samples of rural schools, found very low levels of learning of mathe-
matics in the primary grades (ASER report 2010). A survey of the most preferred,
“top” schools in leading Indian metro cities found surprisingly low levels of concep-
tual understanding in science and mathematics (Educational Initiatives and Wipro
2006).
Efforts to address the issue of failure and low learning levels include an im-
portant reform of the school curriculum following the 2005 National Curriculum
Framework (NCF 2005), which emphasized child-centered learning (National Cen-
tre for Educational Research and Training 2005). New textbooks for grades 1–12
following the NCF 2005 were brought out by the National Council of Educational
Research and Training (NCERT) through a collaborative process involving educa-
tors and teachers. We shall refer to these as the “NCERT textbooks”. The NCERT
textbooks in mathematics have introduced significant changes in the instructional
approach, especially in the primary grades. However, one of the issues that remain
inadequately addressed in the new textbooks is the introduction to symbolic algebra
in the middle grades, which follows a largely traditional approach focused on sym-
bol manipulation. Since algebra is an important part of the secondary curriculum,
bringing mathematics to wider sections of the student population, entails that more
thought be given to how algebra can be introduced in a manner that uses students’
prior knowledge. Our aim in this chapter is to articulate a framework that addresses
the issue of transition from arithmetic to symbolic algebra, and to outline an instruc-
tional approach based on this framework that was developed by the research group
at the Homi Bhabha Centre through a design experiment. In this section of the pa-
per, we shall briefly sketch the background of the reform efforts, insofar as they are
relevant to the teaching and learning of algebra.
In India, school education includes the following levels of schooling: primary:
grades 1–5, middle or upper primary: grades 6–8, secondary: grades 9–10 and higher
secondary: grades 11–12. The provision of school education is largely the domain of
the state government, subject to broad regulations laid down by the central govern-
ment. The vast majority of students learn from textbooks published and prescribed
by the state or the central government. Following the reform process initiated by the
central government through NCF 2005, many state governments have revised or are
in the process of revising their own curricula and textbooks to align them with the
new curriculum framework. In comparison to the earlier years, the mathematics cur-
riculum and the NCERT textbooks at the primary level have changed significantly,
while the middle school curriculum, where algebra is introduced continues largely
unchanged (Tripathi 2007).
Algebra, as a separate topic, forms a large chunk of the middle and secondary
school syllabus in mathematics and also underlies other topics such as geometry or
trigonometry. Students’ facility with algebra is hence an important determinant of

success in school mathematics. Thus, as elsewhere, algebra is a gateway to higher
learning for some pupils and a barrier for others. In the new NCERT textbooks, for-
mal algebra begins in grade 6 (age 11+) with integer operations, the introduction
of variables in the context of generalization, and the solution of simple equations in
one unknown. Over the five years until they complete grade 10, pupils learn about
integers, rational numbers and real numbers, algebraic expressions and identities,
exponents, polynomials and their factorization, coordinate geometry, linear equa-
tions in one and two variables, quadratic equations, and arithmetic progressions.
The grade 6 NCERT mathematics textbook introduces algebra as a branch of
mathematics whose main feature is the use of letters “to write formulas and rules in
a general way” (Mathematics: Text book for class VI 2006, p. 221). It then provides
a gentle introduction to the use of letters as variables, and shows how expressions
containing variables can be used to represent formulas, rules for a growing pattern,
relations between quantities, general properties of number operations, and equa-
tions. However, this easy-paced approach gives way to a traditional approach to the
manipulation of algebraic expressions in grade 7, based on the addition and subtrac-
tion of like terms. The approach in the higher grades is largely formal, with real life
applications appearing as word problems in the exercises. Thus, although an effort
has been made in the new middle school textbooks to simplify the language, the
approach is not significantly different from the earlier approach and does not take
into account the large body of literature published internationally on the difficulty
students face in making the transition from arithmetic to algebra and the preparation
needed for it. (For details and examples, see Banerjee 2008b.)
The NCERT mathematics textbooks for the primary grades, have attempted to
integrate strands of algebraic thinking. In a study of the primary mathematics cur-
ricula in five countries, Cai et al. (2005), have applied a framework that identifies
the algebra strand in terms of algebra relevant goals, algebraic ideas and algebraic
processes. Some of the elements identified by Cai et al. are also found in the NCERT
primary mathematics textbooks. There is a consistent emphasis on identifying, ex-
tending, and describing patterns through all the primary grades from 1 to 5. “Pat-
terns” have been identified as a separate strand in the primary mathematics syllabus,
and separate chapters appear in the textbooks for all the primary grades with the
title “Patterns.” Children work on repeating as well as growing patterns in grade 2
and grade 3. Many other kinds of patterns involving numbers appear in these books:
addition patterns in a 3 × 3 cell on a calendar, magic squares, etc. A variety of num-
ber puzzles are also presented at appropriate grade levels; some of the puzzles are
drawn from traditional or folk sources (for an example, see Math-magic: Book 3
2006, pp. 92–94).
Simple equations with the unknown represented as an empty box or a blank ap-
pear in grade 2 and later. The inverse relation between addition and subtraction is
highlighted by relating corresponding number sentences and is also used in check-
ing column subtraction (Math-magic: Book 3 2006). Change also appears as an
important theme in these textbooks. The quantitative relation between two varying
quantities is discussed at several places: the weight of a growing child which, ac-
cording to a traditional custom, determines the weight of sweets distributed on her
birthday (Grade 3), the number of elders in each generation of a family tree, the
annual growth of a rabbit population, the growth chart of a plant over a number of
days (all in Grade 5, Math-magic: Book 5 2008). No letter symbols are used in these
examples, and relationships are expressed in terms of numerical tables, diagrams or
charts.
These strands in the NCERT primary school textbooks are not taken up and
developed further in the NCERT middle school textbooks, which appear to begin
afresh by introducing a symbolic approach to algebra. A part of the reason lies in
the fact that the two sets of textbooks are produced by different teams, and the
schedule of publication does not always allow for smooth co-ordination. (The grade
6 textbook, for example, was published two years before the grade 5 textbook.)
Another reason, we hypothesize, is the pressure to build students’ capabilities with
symbolic algebra, which is needed for secondary school mathematics. Curriculum
design involves striking a balance between different imperatives. The balance real-
ized in the primary mathematics textbooks emphasizes immersion in realistic con-
texts, concrete activities, and communicating the view that mathematics is not a
finished product (Mukherjee 2010, p. 14). The middle school curriculum is more
responsive to the features of mathematics as a discipline and emphasizes the ab-
stract nature of the subject. In the words of the coordinator for the middle school
textbooks, “learners have to move away from these concrete scaffolds and be able
to deal with mathematical entities as abstract ideas that do not lend themselves to
concrete representations” (Dewan 2010, p. 19f).
Besides finding ways of building on the strands of algebraic thinking that are
present in the primary curriculum and textbooks, a concern, perhaps even more
pressing in the curriculum design context in India, is to find more effective ways
for the majority of children to make the transition to the symbolic mathematics of
secondary school. Algebra underlies a large part of secondary mathematics, and
many students face difficulties of the kind that are identified in studies conducted
elsewhere (Kieran 2006). A compilation of common student errors from discussions
with teachers includes well-known errors in simplifying algebraic expressions and
operating with negative numbers (Pradhan and Mavlankar 1994). Errors involving
misinterpretation of algebraic notation and of the “=” sign are common and per-
sistent (Rajagopalan 2010). Building on students’ prior knowledge and intuition to
introduce symbolic algebra remains one of the challenges facing mathematics cur-
riculum designers, and it is yet to be adequately addressed.
In this chapter, we offer a perspective on the relationship between arithmetic and
algebra and an example of a teaching approach developed by a research group at the
Homi Bhabha Centre led by the authors to manage the transition from arithmetic
to symbolic algebra. The key aspect of this approach is focusing on symbolic arith-
metic as a preparation for algebra. Students work with numerical expressions, that
is, expressions without letter variables, with the goal of building on the operational
sense acquired through the experience of arithmetic. This, however, requires a shift
in the way expressions are interpreted. The aim is not just to compute the value of
an expression, but to understand the structure of the expressions. Numerical expres-
sions offer a way of expressing the intuitions that children have about arithmetic and
have the potential to strengthen this intuition and enhance computational efficiency.
To enable this transition, numerical expressions must be viewed not merely as en-
coding instructions to carry out a sequence of binary operations, but as revealing a
particular operational composition of a number, which is its “value.” Thus facility
with symbolic expressions is more than facility with syntactic transformations of ex-
pressions; it includes a grasp of how quantities or numbers combine to produce the
resultant quantity. This view of expressions leads to flexibility in evaluating expres-
sions and to developing a feel for how transforming an expression affects its value.
We argue that understanding and learning to “see” the operational composition en-
coded by numerical expressions is important for algebraic insight. We elaborate on
the notion of operational composition in a later section and discuss how this per-
spective informs a teaching approach developed through trials with several batches
of students.
The idea that numerical expressions can capture students’ operational sense or re-
lational thinking has been explored in other studies (for example, Fujii and Stephens
2001). In appropriate contexts, students show a generalized interpretation of num-
bers in a numerical expression, treating them as quasi-variables. We will review
these findings briefly in a later section. The idea that algebra can enhance arith-
metical insight is a view that finds support in the Indian historical tradition of math-
ematics. Algebra is viewed not so much as a generalization of arithmetic, but rather
as providing a foundation for arithmetic. An implication is that building on the arith-
metic understanding of students is, at the same time, looking at symbols with new
“algebra eyes.” It is not widely known that Indian mathematicians achieved impres-
sive results in algebra from the early centuries CE to almost modern times. The fact
that Indian numerals and arithmetic were recognized as being superior and adopted
first by the Islamic cultures and later by Europe is more widely known. The ad-
vances in arithmetic and algebra are possibly not unconnected, since arithmetic may
be viewed as choosing a representation of the operational composition of a number
in a way that makes calculation easy and convenient. In the next section, we shall
briefly review some of the achievements in Indian algebra and discuss how the rela-
tion between arithmetic and algebra was viewed in the Indian historical tradition.
Arithmetic and Algebra in the Indian Mathematical Tradition

India had a long standing indigenous mathematical tradition that was active from
at least the first millennium BCE till roughly the eighteenth century CE, when it
was displaced by Western mathematics (Plofker 2009, p. viii). Some of the achieve-
ments of Indian mathematics worth highlighting are the appearance, in a text from
800 BCE, of geometrical constructions and statements found in Euclid’s Elements,
including the earliest explicit statement of the “Pythagoras theorem,” discussion of
the binomial coefficients and the Fibonacci series in a work dated to between 500
and 800 CE, the solution of linear and quadratic indeterminate equations in integers,
a complete integer solution of indeterminate equations of the form x 2 − Ny 2 = 1
(“Pell’s” equation) in a twelfth century text, the finite difference equation for the
sine function in the fifth century CE and power series expansion for the inverse
tangent function in the fourteenth century CE (Mumford 2010; Plofker 2009).
It is well attested that Indian numerals and arithmetic were adopted first by the
Islamic civilization following exchanges between the two cultures around the eighth
century CE, and later by Europe (Plofker 2009, p. 255). Indian algebra was also de-
veloped by this time as seen in the seventh century work of Brahmagupta, which
we shall discuss below. However there are important differences between Arabic
algebra (as found in al-Khwarizmi’s work al-jabr, for example) and the algebra
in Indian mathematical texts. Two of the main differences are that Arabic algebra
avoided negative quantities, while Indian texts routinely used them, and Indian al-
gebra used notational features such as tabular proto-equations and syllabic abbrevi-
ations for unknown quantities, while Arabic algebra was purely rhetorical (Plofker
2009, p. 258f).
We will first give an overview of how topics in arithmetic and algebra are orga-
nized in the central texts of Indian mathematics, and then turn to explicit statements
about the relation between arithmetic and algebra. Indian texts containing mathe-
matics from the first millennium CE are typically one or more chapters of a work
dealing with astronomy. Purely mathematical texts appear only later, as for example,
in the work of Bhaskara II in the twelfth century CE. The Aryabhateeyam, from the
5th century CE, one of the oldest and most influential astronomical-mathematical
texts, contains a single chapter on mathematics that includes arithmetic and the so-
lution of equations.
The Brahmasphuta Siddhanta (c. 628 CE) by Brahmagupta, considered to be one
of the greatest Indian mathematicians of the classical period, has two separate chap-
ters dealing respectively with what we might classify as arithmetic and algebra. The
word that Brahmagupta uses for the second of these chapters (algebra) is kuttaka
ganita or “computation using kuttaka.” Kuttaka (frequently translated as “pulver-
izer”) is an algorithm for reducing the terms of an indeterminate equation, which is
essentially a recasting of the Euclidean algorithm for obtaining the greatest common
divisor of two natural numbers (Katz 1998). Interestingly, puzzles called kuttaka are
found even now in folklore in India and require one to find positive integer solutions
of indeterminate equations. (For an example, see Bose 2009.)
The “arithmetic” chapter in the Brahmasphuta siddhanta deals with topics such
as the manipulation of fractions, the algorithm for cube roots, proportion problems
of different kinds and the “rule of three” (a representation of four quantities in pro-
portion with one of them unknown), the summation of arithmetic progressions and
other kinds of series, miscellaneous computational tips, and problems dealing with
geometry and geometrical measurement (Colebrooke 1817). The kuttaka or algebra
chapter deals with techniques for solving a variety of equations. In the initial verses
of this chapter, we find the oldest extant systematic description in the Indian tradi-
tion of rules of operating with various kinds of quantities: rules for operations with
positive and negative quantities, zero, surds (irrational square roots of natural num-
bers), and unknown quantities. The approach of beginning the discussion of algebra
by presenting the rules of operations with different kinds of numbers or quantities
became a model for later texts. Laying out these rules at the beginning prepared the
way for demonstrating results and justifying the procedures used to solve equations.
Mathematicians who came after Brahmagupta referred to algebra as avyakta

ganita or arithmetic of unknown quantities, as opposed to vyakta ganita (arithmetic
of known quantities). Others, starting from around the 9th century CE, have used
the word bijaganita for algebra. Bija means “seed” or “element,” and bijaganita has
been translated as “computation with the seed or unknown quantity, which yields
the fruit or phala, the known quantity (Plofker 2007, p. 467). The word bija has also
been translated as “analysis” and bijaganita as “calculation on the basis of analy-
sis” (Datta and Singh 1938/2001). “Bijaganita” is the word currently used in many
Indian languages for school algebra.
Bhaskara II from the 12th century CE (the numeral “II” is used to distinguish
him from Bhaskara I of an earlier period) devoted two separate works to arithmetic
and algebra—the Lilavati and the Bijaganita, respectively, both of which became
canonical mathematical texts in the Indian tradition. Through several remarks spread
through the text, Bhaskara emphasizes that bijaganita, or analysis, consists of math-
ematical insight and not merely computation with symbols. Bhaskara appears to
have thought of bijaganita as insightful analysis aided by symbols.
Analysis (bija) is certainly the innate intellect assisted by the various symbols [varna or
colors, which are the usual symbols for unknowns], which, for the instruction of duller
intellects, has been expounded by the ancient sages. . . (Colebrooke 1817, verse 174)
At various points in his work, Bhaskara discourages his readers from using symbols
for unknowns when the problem can be solved by arithmetic reasoning such as using
proportionality. Thus after using such arithmetic reasoning to solve a problem in-
volving a sum loaned in two parts at two different interest rates, he comments, “This
is rightly solved by the understanding alone; what occasion was there for putting a
sign of an unknown quantity? . . . Neither does analysis consist of symbols, nor are
the several sorts of it analysis. Sagacity alone is the chief analysis . . . ” (Colebrooke
1817, verse 110)
In response to a question that he himself raises, “if (unknown quantities) are to
be discovered by intelligence alone what then is the need of analysis?”, he says,
“Because intelligence alone is the real analysis; symbols are its help” and goes on
to repeat the idea that symbols are helpful to less agile intellects (ibid.).
Bhaskara is speaking here of intelligence or a kind of insight that underlies the
procedures used to solve equations. Although he does not explicitly describe what
the insight is about, we may assume that what are relevant in the context are the
relationships among quantities that are represented verbally and through symbols.
We shall later try to flesh out what one may mean by an understanding of quantitative
relationships in the context of symbols.
The word “symbol” here is a translation for the sanskrit word varna, meaning
color. This is a standard way of representing an unknown quantity in the Indian
tradition—different unknowns are represented by different colors (Plofker 2009,
p. 230). Bhaskara’s and Brahmagupta’s texts are in verse form with prose com-
mentary interspersed and do not contain symbols as used in modern mathematics.
This does not imply, however, that a symbolic form of writing mathematics was not
present. Indeed, in the Bakshaali manuscript, which is dated to between the eighth
and the twelfth centuries CE, one finds symbols for numerals, operation signs, frac-
tions, negative quantities and equations laid out in tabular formats, and their form
is closer to the symbolic language familiar to us. For examples of the fairly com-
plex expressions that were represented in this way, see Datta and Singh (1938/2001,
p. 13).
Bhaskara II also explicitly comments about the relation between algebra and
arithmetic at different places in both the Lilavati and the Bijaganita. At the be-
ginning of the Bijaganita, he says, “The science of calculation with unknowns is
the source of the science of calculation with knowns.” This may seem to be the
opposite of what we commonly understand: that the rules of algebra are a general-
ization of the rules of arithmetic. However, Bhaskara clearly thought of algebra as
providing the basis and the foundation for arithmetic, or calculation with “determi-
nate” symbols. This may explain why algebra texts begin by laying down the rules
for operations with various quantities, erecting a foundation for the ensuing analy-
sis required for the solution of equations as well as for computation in arithmetic.
Algebra possibly provides a foundation for arithmetic in an additional sense. The
decimal positional value representation is only one of the many possible represen-
tations of numbers, chosen for computational efficiency. Algebra may be viewed as
a tool to explore the potential of this form of representation and hence as a means
to discover more efficient algorithms in arithmetic, as well as to explore other con-
venient representations for more complex problems.
At another point in the Bijaganita, Bhaskara says, “Mathematicians have de-
clared algebra to be computation attended with demonstration: else there would be
no distinction between arithmetic and algebra” (Colebrooke 1817, verse 214). This
statement appears following a twofold demonstration, using first geometry and then
symbols, of the rule to obtain integer solutions to the equation axy = bx + cy + d.
Demonstration of mathematical results in Indian works often took geometric or al-
gebraic form (Srinivas 2008), with both the forms sometimes presented one after
the other. The role of algebra in demonstration also emerges when we compare the
discussion of quadratic equations in the arithmetic text Lilavati and the algebra text
Bijaganita. In the Lilavati, the rule is simply stated and applied to different types of
problems, while in the Bijaganita, we find a rationale including a reference to the
method of completing the square.
Algebra in earlier historical periods has often been characterized as dealing with
“the solution of equations” (Katz 2001). While this view is undoubtedly correct in
a broad sense, it is partial and misses out on important aspects of how Indian math-
ematicians in the past thought about algebra. Most importantly, they laid stress on
understanding and insight into quantitative relationships. The symbols of algebra
are an aid to such understanding. Algebra is the foundation for arithmetic and not
just the generalization of arithmetic, implying that arithmetic itself must be viewed
with “algebra eyes.” Further, algebra involves taking a different attitude or stance
with respect to computation and the solution of problems; it is not mere description
of solution, but demonstration and justification. Mathematical insight into quantita-
tive relationships, combined with an attitude of justification or demonstration, leads
to the uncovering of powerful ways of solving complex problems and equations.
Making procedures of calculation more efficient and more accurate was often
one of the goals of mathematics in the Indian tradition, and the discovery of efficient
formulas for complex and difficult computations in astronomy was a praiseworthy
achievement that enhanced the reputation of mathematicians. Thus not only do we
find a great variety of procedures for simple arithmetic computations, but also for
interpolation of data and approximations of series (Datta and Singh 1938/2001).
The karana texts contain many examples of efficient algorithms (Plofker 2009,
pp. 105ff). In the “Kerala school” of mathematics, which flourished in Southern
India between the 14th and the 18th centuries CE, we find, amongst many remark-
able advancements including elements of calculus, a rich variety of results in find-
ing rational approximations to infinite series. Thus algebra was related as much
to strengthening and enriching arithmetic and the simplification of complex com-
putation as to the solution of equations. It was viewed both as a domain where the
rationales for computations were grasped and as a furnace where new computational
techniques were forged.
From a pedagogical point of view, understanding and explaining why an inter-
esting computational procedure works is a potential entry point into algebra. Since
arithmetic is a part of universal education, a perspective that views algebra as deep-
ening the understanding of arithmetic has social validity. Thus, while algebra builds
on students’ understanding of arithmetic, in turn, it reinterprets and strengthens this
understanding. In the remaining sections, we explore what this might mean for a
teaching learning approach that emphasizes the arithmetic-algebra connection.
Building on Students’ Understanding of Arithmetic

Modern school algebra relies on a more extensive and technical symbolic appara-
tus than the algebra of the Bijaganita. As students learn to manipulate variables,
terms, and expressions as if they were objects, it is easy for them to lose sight of
the fact that the symbols are about quantities. In the context of arithmetic, students
have only learned to use symbols to notate numbers and to encode binary opera-
tions, usually carried out one at a time. Algebra not only introduces new symbols
such as letters and expressions, but also new ways of dealing with symbols. Without
guidance from intuition, students face great difficulty in adjusting to the new sym-
bolism. So Bhaskara’s precept that algebra is about insight into quantities and their
relationships and not just the use of symbols is perhaps even more relevant to the
learning of modern school algebra.
What do students carry over from their experience of arithmetic that can be useful
in the learning of algebra? Do students obtain insight into quantitative relationships
of the kind that Bhaskara is possibly referring to through their experience of arith-
metic, which can be used as a starting point for an entry into symbolic algebra? Of
course, one cannot expect such insight to be sophisticated. We should also expect
that students may not be able to symbolize their insight about quantitative relation-
ships because of their limited experience of symbols in the context of arithmetic.
Fujii and Stephens (2001) found evidence of what they call students’ relational
understanding of numbers and operations in the context of arithmetic tasks. In a
missing number sentence like 746 + __ − 262 = 747, students could find the number
in the blank without calculation. They were able to anticipate the results of operating
with numbers by finding relations among the operands. Similar tasks have also been
used by others in the primary grades (Van den Heuvel-Panhuizen 1996). Missing
number sentences of this kind are different from those of the kind 13 + 5 = ___ + 8,
where the algebraic element is limited to the meaning of the “=” sign as a relation
that “balances” both sides. Relational understanding as revealed in the responses
to the former kind of sentence lies in anticipating the result of operations without
actual calculation. Fujii and Stephens argue that in these tasks although students are
working with specific numbers, they are attending to general aspects by treating the
numbers as “quasi-variables.”
Students’ relational understanding, as described by Fujii and Stephens, is a form
of operational sense (Slavit 1999), limited perhaps to specific combinations of num-
bers. The students’ performance on these tasks needs to be contrasted with the
findings of other studies. For example, Chaiklin and Lesgold (1984) found that
without recourse to computation, students were unable to judge whether or not
685 − 492 + 947 and 947 + 492 − 685 are equivalent. Students are not consis-
tent in the way they parse expressions containing multiple operation signs. It is
possible that they are not even aware of the requirement that every numerical ex-
pression must have a unique value. It is likely, therefore, that students’ relational
understanding are elicited in certain contexts, while difficulties with the symbolism
overpowers such understanding in other contexts. Can their incipient relational un-
derstanding develop into a more powerful and general understanding of quantitative
relationships that can form the basis for algebraic understanding, as suggested by
Bhaskara? For this to be possible, one needs to build an idea of how symbolization
can be guided by such understanding, and can in turn develop it into a more pow-
erful form of understanding. In a later study, Fujii and Stephens (2008) explored
students’ abilities to generalize and symbolize relational understanding. They used
students’ awareness of computational shortcuts (to take away six, take away ten and
add four) and developed tasks that involved generalizing such procedures and using
symbolic expressions to represent them.
Other efforts to build students’ understanding of symbolism on the basis of
their knowledge of arithmetic have taken what one may describe as an inductive
approach, with the actual process of calculation supported by using a calculator
(Liebenberg et al. 1999; Malara and Iaderosa 1999). In these studies, students
worked with numerical expressions with the aim of developing an understanding
of the structure by applying operation precedence rules and using the calculator
to check their computation. These efforts were not successful in leading to an un-
derstanding of structure that could then be used to deal with algebraic expressions
because of over-reliance on computation (Liebenberg et al.) or because of interpret-
ing numerical and algebraic expressions in different ways (Malara and Ioderosa).
The findings suggest that an approach where structure is focused more centrally and
is used to support a range of tasks including evaluation of expressions, as well as
comparison and transformation of expressions, may be more effective in building a
more robust understanding of symbolic expressions.
We attempted to develop such an approach in a study conducted as a design

experiment with grade 6 students during the two-year period 2003–2005. The
teaching-learning approach evolved over five trials, with modifications made at
the end of each trial based on students’ understanding as revealed through a va-
riety of tasks and our own understanding of the phenomena. The first year of the
study consisted of two pilot trials. In the second year, we followed 31 students over
three teaching trials. These students were from low and medium socio-economic
backgrounds, one group studying in the vernacular language and one in the En-
glish language. Each trial consisted of 1 12 hours of instruction each day for 11–15
days. These three trials, which comprised the main study were held at the beginning
(MST-I), middle (MST-II), and end of the year (MST-III) during vacation periods.
The schools in which the students were studying followed the syllabus and text-
books prescribed by the State government, which prescribe the teaching of evalu-
ation and simplification of arithmetic and algebraic expressions in school in grade
6 in a traditional fashion—using precedence rules for arithmetic expressions and
the distributive property for algebraic expressions. Discussion with students and a
review of their notebooks showed that only the vernacular language school actually
taught simplification of algebraic expressions in class 6; the English school omitted
the chapter.
These students joined the program at the end of their grade 5 examinations and
were followed till they completed grade 6. They were randomly selected for the
first main study trial from a list of volunteers who had responded to our invitation to
participate in the program. The students were taught in two groups, in the vernacular
and the English language respectively by members of the research team.
Data was collected through pre- and post-tests in each trial, interviews conducted
eight weeks after MST-II (14 students) and 16 weeks after MST-III (17 students),
video recording of the classes and interviews, teachers’ log and coding of daily
worksheets. The pre- and post-tests contained tasks requiring students to evaluate
numerical expressions and simplify algebraic expressions, to compare expressions
without recourse to calculation and to judge which transformed expressions were
equivalent to a given expression. There were also tasks where they could use al-
gebra to represent and draw inferences about a given problem context, such as a
pattern or a puzzle. In the written tests, the students were requested to show their
work for the tasks. The students chosen for the interview after MST-II had scores
in the tests which were below the group average, at the average, and above the
group average, and who had contributed actively to the classroom discussions. The
same students also participated in the interviews after MST-III, and a few additional
students were also interviewed. The interviews probed their understanding more
deeply, using tasks similar to the post test. In particular, they probed whether student
responses were mechanical and procedure-based or were based on understanding.
The overall goal of the design experiment was to evolve an approach to learn-
ing beginning algebra that used students’ arithmetic intuition as a starting point.
The specific goal was to develop an understanding of symbolic expressions together
with the understanding of quantitative relationships embodied in the expressions.
Although this was done with both numerical and algebraic expressions, the approach
entailed more elaborate work with numerical expressions by students compared to

the approach in their textbooks. Students worked on tasks of evaluating expressions,
of comparing expressions without calculation, and of transforming expressions in
addition to a number of context-based problems where they had to generalize or ex-
plain a pattern. A framework was developed that allowed students to use a common
set of concepts and procedures for both numerical and algebraic expressions. Details
of the study are available in Banerjee (2008a). Here we shall briefly indicate how
the teaching approach evolved, describe the framework informing the approach, and
present some instances of students’ responses to the tasks.
In the pilot study, students worked on tasks adapted from Van den Heuvel-
Panhuizen (1996), and that were similar to the tasks used by Fujii and Stephens
(2001). We found several instances of relational thinking similar to those reported
by Fujii and Stephens. For example, students could judge whether expressions like
27 + 32 and 29 + 30 were equivalent and also give verbal explanations. One of the
explanations used a compensation strategy: “the two expressions are equal because
we have [in the first expression] taken 2 from 32 and given it to 27 [to obtain the
second expression].” Students worked with a variety of such expressions, contain-
ing both addition and subtraction operations, with one number remaining the same
or both numbers changed in a compensating or non-compensating manner (Subra-
maniam 2004). Some pairs were equivalent, and some pairs were not. For the pairs
which were not equivalent, they had to judge which was greater and by how much.
As seen in the explanation just cited, students used interesting strategies including
some ad-hoc symbolism, but this did not always work. In general, when they at-
tempted to compare the expressions by merely looking at their structure and not
by computing, students made accurate judgements for expressions containing the
addition operation but not for those containing the subtraction operation. Similarly
they were not always successful in judging which expression was greater in a pair
of expressions when the compensation strategy showed that they were unequal.
We noticed that students were separating out and comparing the additive units in
the pairs of expressions but were comparing numbers and operation signs in incon-
sistent ways. This led to an approach that called attention more clearly to additive
units in comparison tasks. However, an important moment in the evolution of the ap-
proach was the decision to use a structure-based approach for comparing as well as
for evaluating numerical expressions. Other important aspects of the approach were
dealing with arithmetic and algebraic expressions in a similar manner in the different
tasks and relating these processes to algebraic contexts of generalizing and justifi-
cation of patterns. We have described the evolution of the approach in greater detail
elsewhere (Banerjee and Subramaniam submitted). Here we describe a framework
that supports a structure based approach to working with numerical expressions on
a range of tasks including evaluation, comparison and transformation.
The Arithmetic-Algebra Connection—A Framework

As we remarked earlier, learning algebra involves learning to read and use symbols
in new ways. These new ways of interpreting symbols need to build on and am-
plify students’ intuition about quantitative relationships. The view that algebra is
the foundation of arithmetic, held by Indian mathematicians, entails that students

need to interpret the familiar symbols of arithmetic also in new ways. The literature
on the transition from arithmetic to algebra has identified some of the differences
in the way symbols are used in arithmetic and algebra: the use of letter symbols,
the changed interpretation of key symbols such as the “=” sign, and the acceptance
of unclosed expressions as appropriate representations not only for operations but
also for the result of operations (Kieran 2006). An aspect related to the last of the
changes mentioned that we wish to emphasize is the interpretation of numerical and
algebraic expressions as encoding the operational composition of a number.
The use of expressions to stand for quantities is related to the fact that, while in
arithmetic one represents and thinks about one binary operation, in algebra we need
to represent and think about more than one binary operation taken together. As stu-
dents learn computation with numbers in arithmetic, they typically carry out a single
binary operation at a time. Even if a problem requires multiple operations, these are
carried out singly in a sequence. Consequently, the symbolic representations that
students typically use in arithmetic problem-solving contexts are expressions en-
coding a single binary operation. In the case of formulas, the representation may
involve more than one binary operation, but they are still interpreted as recipes for
carrying out single binary operations one at a time. They do not involve attending
to the structure of expressions or manipulating the expressions. Indeed, one of the
key differences of the arithmetic approach to solving problems, as opposed to the
algebraic, is that students compute intermediate quantities in closed numerical form
rather than leaving them as symbols that can be operated upon. And these inter-
mediate quantities need to be thought about explicitly and must be meaningful in
themselves (Stacey and Macgregor 2000).
The representational capabilities of students need to be expanded beyond the
ability to represent single binary operations before they move on to algebra. In
the traditional curriculum, this is sought to be achieved by including a topic on
arithmetic or numerical expressions, where students learn to evaluate expressions
encoding multiple binary operations. However, students’ work on this topic in the
traditional curriculum is largely procedural, and students fail to develop a sense of
the structure of expressions. As discussed earlier, students show relational under-
standing in certain contexts, but in general have difficulty in interpreting symbolic
expressions.
One problem that arises when numerical expressions encode multiple binary op-
erations is that such expressions are ambiguous with respect to operation prece-
dence when brackets are not used. At the same time, one cannot fully disambiguate
the expression using brackets since the excessive use of brackets distracts from the
structure of the expression and is hence counter-productive. Students are, therefore,
taught to disambiguate the expression by using the operation precedence rules. The
rationale for this, namely, that numerical expressions have a unique value is often
left implicit and not fully grasped by many students. Even if the requirement is made
explicit, students are unlikely to appreciate why such a requirement is necessary. The
transformation rules of algebra are possible only when algebraic expressions yield
numerical expressions with a unique value when variables are appropriately sub-
stituted. Thus disambiguating numerical expressions is a pre-condition for the use
of rules of transformations that preserve the unique value of the expression. Since
students are yet to work with transformations of expressions, they cannot appreci-
ate the requirement that numerical expressions must be unambiguous with regard to
value.
In the traditional curriculum, students’ work with numerical expressions is lim-
ited and is seen merely as preparatory to work with algebraic expressions. How
does one motivate a context for work with numerical expressions encoding multiple
binary operations? Student tasks with such expressions need to include three inter-
related aspects—representational, procedural (evaluation of expressions), and trans-
formational. To fully elaborate these aspects, we need to interpret expressions in a
way different from the usual interpretation of an expression as encoding a sequence
of such operations to be carried out one after another, a sequence determined by the
visual layout in combination with the precedence rules. The alternative interpreta-
tion that students need to internalize is that such expressions express or represent
the operational composition of a quantity or number. In other words, the expres-
sion reveals how the number or quantity that is represented is built up from other
numbers and quantities using the familiar operations on numbers. This interpreta-
tion embodies a more explicit reification of operations and has a greater potential
to make connections between symbols and their semantic referents. The idea of the
operational composition of a number, we suggest, is one of the key ideas marking
the transition from arithmetic to algebra.
Let us illustrate this idea with a few examples: (i) the expression 500 − 500 ×
20/100 may indicate that the net price is equal to the marked price less the
discount, which in turn is a fraction of the marked price, (ii) the expression
5 × 100 + 3 × 10 + 6 shows the operational composition expressed by the canoni-
cal representation of a number (536) as composed of multiunits which are different
powers of ten, (iii) the expression 300 + 0.6t may indicate cell phone charges as
including a fixed rent and airtime charges at a fixed rate per unit of airtime. In ex-
amples (i) and (iii), the operational composition refers back to quantities identifiable
in particular situations, while in example (ii) abstract quantities are put together or
“operationally composed” to yield the number 536. It is important to preserve both
these senses in unpacking the notion of operational composition.
By operational composition of a quantity, we mean information contained in the
expression such as the following: what are the additive part quantities that a quantity
is composed of? Are any of these parts scaled up or down? By how much? Are
they obtained as a product or quotient of other quantities? The symbolic expression
that denotes the quantity simultaneously reveals its operational composition, and in
particular, the additive part quantities are indicated by the terms of the expression.
A refined understanding of operational composition includes accurate judgments
about relational and transformational aspects. What is the relative contribution of
each part quantity (each term) as indicated by the expression? Do they increase or
decrease the target quantity? Which contributions are large, which small? How will
these contributions change if the numbers involved change? How does the target
quantity change when the additive terms are inverted, that is, replaced by the additive
inverse of the given term? What changes invert the quantity as a whole? What are
the transformations that keep the target quantity unchanged? If additive parts are
themselves composed from other quantities, how do we represent and understand
this?
The idea of the operational composition encoded by an expression is similar to
the idea of a function but is more general and less precise. Looking at an expression
as a function has a more narrow focus: how does the target quantity vary when one
or more specific part quantities are varied in a systematic manner while retaining the
form of the operational composition? When expressions are compared and judged
to be equivalent, we judge that different operational compositions yield the same
value. However, the idea of operational composition may play a role in developing
the understanding of functions.
When we interpret expressions as encoding operational composition, we are
not restricted to algebraic expressions. In fact, numerical expressions emerge as
an important domain for reasoning about quantity, about relations and transforma-
tions, and for developing a structure based understanding of symbolic representation
through the notion of operational composition. The pedagogical work possible in the
domain of numerical expressions as a preparation for algebra expands beyond what
is conceived in the traditional curriculum. Numerical expressions emerge as a do-
main for reasoning and for developing an understanding of the structure of symbolic
representation.
When students’ tasks focus on numerical expressions as encoding operational
composition, attention is drawn to the relations encoded by the expression. Students
are freed from the need to unpack the expression as a sequence of operations, fixed
by a set of operation precedence rules. In the teaching approach that we developed,
we emphasized ways of working with expressions that attend to the structure of
expressions and are broadly aimed at developing an insight into quantitative rela-
tionships that must accompany working with symbols.
A simple numerical expression like 5 + 3 is usually interpreted as encoding an
instruction to carry out the addition operation on the numbers 5 and 3. In changing
the focus to operational composition, the first transition that students make is to see
the expression as “expressing” some information about the number 8. This infor-
mation can be expressed verbally in various ways: 8 is the sum of 5 and 3, 8 is 3
more than 5, etc. Other expressions such as 6 + 2 or 2 × 4 contain other information
about the number 8, i.e., they encode different operational compositions of the num-
ber 8. Starting from this point, students move on to expressions with two or more
operations of addition and subtraction. Each expression gives information about the
number which is the “value” of the expression, and reveals a particular operational
composition of the number.
What grounding concepts can scaffold students’ attempts to study and understand
the operational composition revealed in an expression? The basic level of informa-
tion is contained in the terms or the additive units of the expression. Simple terms are
just numbers together with the preceding “+” or “−” sign. Positive terms increase
the value of the number denoted by the expression and negative terms decrease the
value. Additive units are dimensionally “homogenous,” and can be combined in any
order.
Fig. 1 Evaluation of expressions containing only simple terms by students using flexible ways in
the three trials of the study (MST I, II and III)
This shift in perspective subtly turns attention away from procedure towards
structure. In order to evaluate an expression, students do not need to work out
and implement a sequence of binary operations in the correct order. Rather, to
determine the value of the expression, they may combine simple terms in any
order, keeping in view the compensating contributions of positive and negative
terms. The concept of negative terms provides an entry point into signed num-
bers as encoding increase or decrease, which is one of the three interpetations of
integers proposed by Vergnaud cited in Fuson (1992, p. 247). The approach of
combining simple terms in any order, affords flexibility in evaluating an expres-
sion or in comparing expressions that is critical to uncovering structure. Thus stu-
dents may cancel out terms that are additive inverses of one another; they may
gather together some or all of the positive terms or the negative terms and find
easy ways to compute the value of the expression by combining terms. Figure 1
shows students combining terms in flexible ways while evaluating expressions rather
than proceeding according to operation precedence rules. Since the identification
of additive units namely, terms, is the starting point of this approach, we have
described this approach elsewhere as the “terms approach” (Subramaniam 2004;
Banerjee and Subramaniam 2008).
Identifying the additive units correctly is one of the major hurdles that some stu-
dents face. This is indicated by the frequency of such errors as “detachment of the
minus sign” (50 − 10 + 10 = 30), and “jumping off with the posterior operation”
(115 − n + 9 = 106 − n or 106 + n) (Linchevski and Livneh 1999). Although these
errors are often not taken to be serious, they are widespread among students and
impede progress in algebra. Not having a secure idea about the units in an expres-
sion and not knowing how they combine to produce the value may enhance the
experience of algebra as consisting of arbitrary rules.
In working with transformations of expressions, some studies indicate that vi-
sual patterns are often more salient to students than the rules that the students may
know for transforming expressions (Kirshner and Awtry 2004), suggesting that vi-
sual routines are easier to learn and implement than verbal rules. One advantage
with the “terms approach” is the emphasis on visual routines rather than on verbal
rules in parsing and evaluating an expression. Terms were identified in our teaching
approach by enclosing them in boxes. In fact, the rule that multiplication precedes
addition can be recast to be consistent with visual routines. This is done by mov-
ing beyond simple terms, which are pure numbers with the attached + or − sign,
to product terms. In expressions containing “+,” “−,” and the “×” operation signs,
students learn to distinguish the product terms from the simple terms: the product
terms contain the “×” sign. Thus in the expression 5 + 3 × 2 the terms are +5 and
+3 × 2. In analyzing the operational composition encoded by the expression, or in
combining terms to find the value of the expression, students first identify the simple
and the product terms by enclosing them in boxes. The convention followed is that
product terms must be converted to simple terms before they can be combined with
other simple terms. Thus the conventional rule that in the absence of brackets multi-
plication precedes addition or subtraction is recast in terms of the visual layout and
operational composition. Product terms are the first of the complex terms that stu-
dents learn. Complex terms include product terms, bracket terms (e.g., +(8 − 2 × 3))
and variable terms (e.g., −3 × x).
The approach included both procedurally oriented tasks such as evaluation of
expressions and more structurally oriented tasks, such as identifying equivalent ex-
pressions and comparing expressions. As remarked earlier, one of the main features
of the approach evolved only after the initial trials—the use of the idea of terms in
the context of both procedurally and structurally oriented tasks. In the earlier trials,
the use of the idea was restricted to structurally oriented tasks involving comparison
of expressions, and the operation precedence rules were used for the more proce-
durally oriented tasks of evaluating expressions. By using the “terms idea” in both
kinds of tasks, students began to attend to operational composition for both evaluat-
ing and comparing expressions, which allowed them to develop a more robust un-
derstanding of the structure of expressions. By supporting the use of structure for the
range of tasks, this approach actually blurred the distinction between structural and
procedural tasks. Students’ written as well as interview responses revealed that they
were relatively consistent in parsing an expression and that they appreciated the fact
that evaluation of a numerical expression leads to a unique value (Banerjee 2008a;
Banerjee and Subramaniam submitted).
In the students’ written responses, we found a reduction as they moved from the
first trial (MST-I) to the last (MST-III) in the common syntactic errors in evaluating
numerical expressions or in simplifying algebraic expressions such as the conjoining
error (5 + x = 5x), the detachment error described above, and the LR error (evalu-
ating an expression from left to right and ignoring multiplication precedence). More
importantly, students who were interviewed showed a reliance on identifying sim-
ple and complex terms to assess whether a particular way of combining terms was
correct. Their understanding of procedural aspects was robust in the sense that they
were able to identify and correct errors in a confident manner, when probed with
alternative ways of computing expressions.
The interviews also revealed how some students were able to use their under-
standing of terms to judge whether two expressions were equal. One of the ques-
tions required students to identify which expression was numerically greater, when
two expressions were judged to be unequal. Although this was not a question fa-
miliar to the students from classroom work, they were able to interpret the units or
terms in the expression to make correct judgments. The following interview excerpt
post-MST III from one of the better performing students illustrates how the idea of
operational composition could be put to use in making comparisons:
Interviewer: Ok. If I put m = 2 in this first expression [13 × m − 7 − 8 × 4 + m] and I put
m = 2 in the original expression [13 × m − 7 − 8 × m + 4], would I get the same value?
BK: No.
Interviewer: It will not be. Why?
BK: Because it is 8 × 4 [in the first expression], if it [the value of m] is 4 here, then it would
be the same value for both.
[The student is comparing the terms which are close but not equal: −8 × 4 and −8 × m.
She says that if m were equal to 4, the expressions will be equal, but not otherwise.]
Interviewer: . . . If I put m = 2 in (this) expression [−7 + 4 + 13 × m − m × 8] and m = 2
in the original expression [13 × m − 7 − 8 × m + 4], then would they be the same?
BK: Yes.
Interviewer: Why?
BK: Because, m is any number, if we put any number for that then they would be the same.
[Comparing the two expressions the student judges correctly that they are equal.]
Our study focused largely on expressions that encoded additive composition and,
to a limited extent, combined it with multiplicative composition. Learning to parse
the additive units in an expression is an initial tool in understanding the operational
composition encoded by the expression. Multiplicative composition as encoded in
a numerical expression is conceptually and notationally more difficult and requires
that students understand the fraction notation for division and its use in representing
multiplication and division together. In our study, multiplicative composition was
not explored beyond the representation of the multiplication of two integers since
students’ understanding of the fraction notation was thought to be inadequate.
Even with this restriction, the study revealed much about students’ ability to
grasp operational composition and showed how this can lead to meaningful work
with expressions as we have tried to indicate in our brief descriptions above. It is
generally recognized that working with expressions containing brackets is harder for
students. While this was not again explored in great detail in the study, we could find
instances where students could use and interpret brackets in a meaningful way. In an
open-ended classroom task where students had to find as many expressions as they
could that were equivalent to a given expression, a common strategy was to replace
one of the terms in the given expression, by an expression that revealed it as a sum or
a difference. For example, for the expression, 8 × x + 12 + 6 × x, students wrote the
equivalent expression (10 − 2) × x + 12 + (7 − 1) × x, using brackets to show which
numbers were substituted. This was a notation followed commonly by students for
several such examples. Besides the use of brackets, this illustrates students using the
idea that equals can be substituted one for the other, and that “unclosed” expressions
could be substituted for “closed” ones. In the same task, students also used brackets
to indicate use of the distributive property as for example, when they wrote for the
given expression 11 × 4 − 21 + 7 × 4 the equivalent expression 4 × (7 + 11) − 21.
The study also included work with variable terms and explored how students
were able to carry over their understanding of numerical expressions to algebraic ex-
pressions. We found that students were capable of making judgments about equiva-
lent expressions or of simplifying expressions containing letter symbols just as they
were in working with numerical expressions. This did not, however, necessarily
mean that they appreciated the use of algebraic symbols in contexts of generaliza-
tion and justification (Banerjee 2008a). The culture of generalization that algebra
signals probably develops over a long period as students use algebraic methods for
increasingly complex problems.
We have attempted here to develop a framework to understand the arithmetic-
algebra connection from a pedagogical point of view and to sketch briefly how a
teaching approach informed by this framework might begin work with symbolic al-
gebra by using students’ arithmetic intuition as a starting point. Although the design
experiment through which the teaching approach was developed was not directly
inspired by the historical tradition of Indian mathematics, we have found there a
source for clarifying the ideas and the framework that underlie the teaching ap-
proach. The view that understanding quantitative relationships is more important
than just using symbols and the idea that algebra provides the foundation for arith-
metic are powerful ideas whose implications we have tried to spell out. We have
argued that symbolic expressions, in the first instance, numerical expressions, need
to be seen as encoding operational composition of a number or quantity rather than
as a set of instructions to carry out operations. We have also pointed to the im-
portance, from a perspective that emphasizes structure, of working with numerical
expressions as a preparation for beginning symbolic algebra.
References
ASER report (2010). Annual Status of Education Report (Rural) 2009 (Provisional). New Delhi:
ASER Centre. Retrieved from http://asercentre.org/asersurvey/aser09/pdfdata/aser09.pdf.
Banerjee, R. (2008a). Developing a learning sequence for transiting from arithmetic to elemen-
tary algebra. Unpublished doctoral dissertation. Mumbai: Homi Bhabha Centre for Science
Education, Tata Institute of Fundamental Research.
Banerjee, R. (2008b). Assessing the curriculum reforms in India: the case of integers and algebra
for beginning middle school students. In International Conference on Mathematics Education
(ICME-11), Mexico. Retrieved from http://dg.icme11.org/document/get/99.
Banerjee, R., & Subramaniam, K. (2008). Bridging arithmetic and algebra: Evolution of a teaching
sequence. In O. Figueras et al. (Eds.), International Group of the Psychology of Mathematics
Education: Proceedings of the Joint Meeting of PME 32 and PME-NA XXX (PME29) (Vol. 2,
pp. 121–128). Morelia, Mexico.
Banerjee, R., & Subramaniam, K. (submitted). Evolution of a teaching approach for beginning
algebra.
Bose, A. (2009). Mathematical riddles among the Mushars: Linked to a historical tradition. In
M. Tzekaki, M. Kaldrimidou, & H. Sakonidis (Eds.), Proceedings of the 33rd of the Inter-
national Group for the Psychology of Mathematics Education (Vol. 5, p. 439). Thessaloniki,
Greece.
Cai, J., Lew, H. C., Morris, A., Moyer, J. C., Ng, S. F., & Schmittau, J. (2005). The develop-
ment of students’ algebraic thinking in earlier grades: A cross-cultural comparative perspective.
Zentralblatt fuer Didaktik der Mathematik (International Review on Mathematics Education),
37(1), 5–15.
Chaiklin, S., & Lesgold, S. (1984). Prealgebra students’ knowledge of algebraic tasks with arith-
metic expressions. Paper presented at the annual meeting of the American Research Associa-
tion.
Colebrooke, H. T. (1817). Algebra, with Arithmetic and Mensuration from the Sanscrit of Brah-
megupta and Bhascara. London: John Murray.
Datta, B., & Singh, A. N. (1938/2001). History of Hindu Mathematics, Vol. II, Edition 2001. Delhi:
Bharatiya Kala Prakashan.
Dewan, H. K. (2010). Pedagogy of mathematics. Learning Curve, XIV, 16–22.
Educational Initiatives and Wipro (2006). Student learning in the Metros 2006: How well are
our students learning? Report available online from http://www.ei-india.com/full-report.pdf.
Accessed 1st Nov 2009.
Fujii, T., & Stephens, M. (2001). Fostering an understanding of algebraic generalization through
numerical expressions: The role of quasi-variables. In H. Chick, K. Stacey, & J. Vincent (Eds.),
Proceedings of the 12th ICMI Study Conference: The Future of the Teaching and Learning of
Algebra (Vol. 1, pp. 258–264). Melbourne, Australia: The University of Melbourne.
Fujii, T., & Stephens, M. (2008). Using number sentences to introduce the idea of a variable. In
C. E. Greenes & R. Rubenstein (Eds.), Algebra and Algebraic Thinking in School Mathematics,
Seventieth Yearbook (pp. 127–140). Reston, VA: NCTM.
Fuson, K. C. (1992). Research on whole number addition and subtraction. In D. A. Grouws (Ed.),
Handbook of Research in Mathematics Teaching and Learning. New York: MacMillan.
Katz, V. (1998). A History of Mathematics: An Introduction (2nd ed.). Reading, Massachusetts:
Addison Wesley.
Katz, V. J. (2001). Using the history of algebra in teaching algebra. In H. Chick, K. Stacey, J.
Vincent, & J. Vincent (Eds.), Proceedings of the 12th ICMI Study Conference: The Future of the
Teaching and Learning of Algebra (Vol. 2, pp. 353–359). Melbourne: University of Melbourne.
Kieran, C. (2006). Research on the learning and teaching of algebra: A broad source of meaning.
In A. Gutierrez & P. Boero (Eds.), Handbook of Research on the Psychology of Mathematics
Education: Past, Present and Future (pp. 11–49). Rotterdam, The Netherlands: Sense Publish-
ers.
Kirshner, D., & Awtry, T. (2004). Visual salience of algebraic transformations. Journal for Re-
search in Mathematics Education, 35(4), 224–257.
Liebenberg, R., Linchevski, L., Sasman, M. C., & Olivier, A. (1999). Focusing on the structural
aspects of numerical expressions. In J. Kuiper (Ed.), Proceedings of the 7th Annual Confer-
ence of the Southern African Association for Research in Mathematics and Science Education
(SAARMSE) (pp. 249–256). Harare, Zimbabwe.
Linchevski, L., & Livneh, D. (1999). Structure sense: The relationship between algebraic and nu-
merical contexts. Educational Studies in Mathematics, 40(2), 173–196.
Malara, N., & Iaderosa, R. (1999). The interweaving of arithmetic and algebra: Some questions
about syntactic and structural aspects and their teaching and learning. In I. Schwank (Ed.),
Proceedings of the First Conference of the European Society for Research in Mathematics
Education (Vol. 2, pp. 159–171). Osnabrueck: Forschungsinstitut fuer Mathematikdidaktik.
Math-magic: Book 3 (2006). New Delhi: National Council of Educational Research and Training.
Math-magic: Book 5 (2008). New Delhi: National Council of Educational Research and Training.
Mathematics: Text book for class VI (2006). New Delhi: National Council of Educational Research
and Training.
Mukherjee, A. (2010). The nature of mathematics and its relation to school education. Learning
curve, XIV, 16–22.
Mumford, D. (2010). Review of Mathematics in India by Kim Plofker. Notices of the American
Mathematical Society, 27(3), 385–390.
National Centre for Educational Research and Training (2005). National Curriculum Framework.
Retrieved from http://www.ncert.nic.in/html/pdf/schoolcurriculum/framework05/nf2005.pdf.
National Centre for Educational Research and Training (2006). Position paper of the National
Focus Group on the Teaching of Mathematics. Retrieved from www.ncert.nic.in/html/pdf/
schoolcurriculum/position_papers/math.pdf.
Plofker, K. (2007). Mathematics in India. In V. Katz (Ed.), The Mathematics of Egypt,
Mesopotamia, China, India, and Islam: A Sourcebook (pp. 385–513). Princeton, New Jersey:
Princeton University Press.
Plofker, K. (2009). Mathematics in India. Princeton, New Jersey: Princeton University Press.
Pradhan, H. C., & Mavlankar, A. T. (1994). Compendium of Errors in Middle School Mathematics.
Mumbai: Homi Bhabha Centre for Science Education, Tata Institute of Fundamental Research.
Rajagopalan, S. (2010). Insights about student learning from an adaptive learning math program.
Learning Curve, XIV, 93–99.
Slavit, D. (1999). The role of operation sense in transitions from arithmetic to algebraic thought.
Educational Studies in Mathematics, 37(3), 251–274.
Srinivas, M. D. (2008). Proofs in Indian mathematics, epilogue. In K. V. Sarma (Ed.), Ganita-
Yukti-Bhasa of Jyesthadeva, Vol. 1—Mathematics (pp. 267–293). New Delhi: Hindustan Book
Agency.
Stacey, K., & Macgregor, M. (2000). Learning the algebraic method of solving problem. Journal
of Mathematical Behavior, 18(2), 149–167.
Subramaniam, K. (2004). Naming practices that support reasoning about and with expres-
sions. In Proceedings of the International Congress on Mathematics Education (ICME 10),
Denmark. Available online at http://www.icme10.dk/proceedings/pages/regular~pdf/RL~K
~Subramanian.pdf.
Tripathi, P. (2007). Review of mathematics textbooks. Contemporary Education Dialogue, 4(1),
142–151.
Van den Heuvel-Panhuizen, M. (1996). Assessment and Realistic Mathematics Education. Utrecht,
The Netherlands: CD-ß Press Utrecht University.
Shiki: A Critical Foundation for School Algebra
in Japanese Elementary School Mathematics
Tad Watanabe
Abstract An analysis of the Japanese elementary school (Grades 1 through 6)

mathematics curriculum materials reveals that the study of functional relationships
(patterns) is a major emphasis in Japan, as is the case in curricula from other coun-
tries. However, the Japanese curriculum considers the ideas related to mathematical
expressions, called “shiki” in Japanese, as the second pillar of elementary school
algebra. This chapter elaborates how a Japanese textbook series attempts to realize
this emphasis on writing and interpreting mathematical expressions.
The Final Report of the National Mathematics Advisory Panel (2008) identifies the
major topics of school algebra. The Panel argues that mathematical experiences stu-
dents encounter in elementary and middle schools must therefore address what the
Panel labeled the Critical Foundations for school algebra (see Table 1). The items on
this list come from three major components: fluency with whole numbers; fluency
with fractions; and particular aspects of geometry and measurement. Although these
topics identified by the Panel are certainly foundations for school algebra, there may
be other elementary and middle school mathematics topics that are equally impor-
tant for, and perhaps more directly related to, school algebra. The purpose of this
chapter is to present the findings from an analysis of the treatment of algebra in the
Japanese elementary school (Grades 1 through 6) mathematics curriculum materials
and to examine what might be foundational for school algebra. The purpose of this
study is not to make an evaluative judgment about the Japanese, nor the US, treat-
ment of algebra in elementary schools but rather to help us more clearly understand
our own practices.
T. Watanabe ()
Department of Mathematics & Statistics, Kennesaw State University, Kennesaw, GA, USA

110 T. Watanabe
Table 1 Critical foundations for school algebra by the National Mathematics Advisory Panel
(2008)
Fluency With Whole Numbers
1. By the end of Grade 3, students should be proficient with the addition and subtraction of
whole numbers.
2. By the end of Grade 5, students should be proficient with multiplication and division of whole
numbers.
Fluency With Fractions
1. By the end of Grade 4, students should be able to identify and represent fractions and decimals,
and compare them on a number line or with other common representations of fractions and
decimals.
2. By the end of Grade 5, students should be proficient with comparing fractions and decimals
and common percents, and with the addition and subtraction of fractions and decimals.
3. By the end of Grade 6, students should be proficient with multiplication and division of frac-
tions and decimals.
4. By the end of Grade 6, students should be proficient with all operations involving positive and
negative integers.
5. By the end of Grade 7, students should be proficient with all operations involving positive and
negative fractions.
6. By the end of Grade 7, students should be able to solve problems involving percent, ratio, and
rate and extend this work to proportionality.
Geometry and Measurement
1. By the end of Grade 5, students should be able to solve problems involving perimeter and area
of triangles and all quadrilaterals having at least one pair of parallel sides (i.e., trapezoids).
2. By the end of Grade 6, students should be able to analyze the properties of two-dimensional
shapes and solve problems involving perimeter and area, and analyze the properties of three-
dimensional shapes and solve problems involving surface area and volume.
3. By the end of Grade 7, students should be familiar with the relationship between similar
triangles and the concept of the slope of a line.
School Algebra and Algebra in Early Grades

Kilpatrick and Izsak (2008) discusses how the role of “algebra” in the US school
curricula has changed from its complete absence to a requirement for college bound
students and then to a high school graduation requirement. In general, the movement
in the United States has been to push algebra for more students, i.e., “algebra for
all,” and earlier, i.e., “algebra in Grade 8.” Principles and Standards for School
Mathematics by the National Council of Teachers of Mathematics (2000) includes
“algebra” as one of the five common strands for pre-K through Grade 12 school
mathematics.
As the target audience and the timing of school algebra was changing, the nature
of school algebra itself was evolving as well. A traditional image of school algebra
often centers on solving various types of equations and inequalities through symbol
manipulation. Although solving equations and inequalities are still an important
Shiki: A Critical Foundation for School Algebra in Japanese Elementary School 111
component of school algebra, in the last few decades, school algebra has become
more broadly conceptualized. Although there are some differences, researchers and
curriculum developers (e.g., Blanton and Kaput 2005; Kilpatrick and Izsak 2008;
Usiskin 1988) seem to agree that school algebra includes the following aspects:
• Algebra as generalized arithmetic
• Algebra as a study of functions, patterns, and relationships
• Algebra as a tool for problem solving
• Algebra as a study of structures
Along with this broader conceptualization of school algebra, a consensus seems
to have developed that students’ experiences in elementary schools is a key factor
in their success with school algebra. However, as Kieran (2004) noted, there does
not appear to be a general consensus on what algebraic thinking in the early grades
should look like. In fact, analyses of the elementary school curricula from China
(Cai 2004), Korea (Lew 2004), Singapore (Ng 2004), the United States (Moyer et
al. 2004), and Russia (Schmittau and Morris 2004) reveal that the approaches to al-
gebra in the early grades vary significantly. However, these analyses serve as useful
reference points to design, examine, and refine our own practices. The present study
will expand the existing knowledge base by adding a case study of the treatment of
algebra in the Japanese elementary school curriculum. In this chapter, I will report
briefly the findings from the case study, and then provide a more in-depth analysis
and discussion of one specific aspect in a textbook series.
Methodology
The present study tried to address the following research questions:
1. What are the big ideas of algebra in the Japanese elementary school mathematics
curriculum?
2. What ideas related to algebra are included in the Japanese elementary school
mathematics curriculum and when are they discussed?
3. How are those ideas developed in textbooks?
Materials In the present study, two documents published by the Japanese Min-
istry of Education, Culture, Sports, Science and Technology (hereafter the Ministry
of Education) were examined. The first document is the national Course of Study
(COS) for elementary school mathematics. This document specifies what topics
are to be taught at what grade levels. The second document, often called Teach-
ing Guide, elaborates and explains the COS in more details. In addition, two most
widely used textbook series, including their teachers’ manuals, were examined. Both
series followed the 1989 COS. The content of the 1989 COS is very similar to the
2008 COS, which will be implemented beginning in the 2011–12 school year.
Analysis Framework The two documents published by the Ministry of Education

were first analyzed to identify those topics that are related to algebra. This analysis
112 T. Watanabe
initially used the four aspects of school algebra discussed earlier as a framework to
specify the topics. These documents, but particularly Teaching Guide, were further
analyzed to reveal any other topics related to the initially specified topics. Through
these analyses, the big ideas of algebra in the Japanese elementary school mathemat-
ics curriculum were identified. Finally, the ways the identified topics were treated
in the two textbook series were analyzed. The textbook analysis focused on how
textbook authors intended to realize what was described in the Ministry documents.
The analysis of the teachers’ manuals played a significant role.
The materials analyzed for the study were all in Japanese. However, Teaching
Guide for the 1989 COS was translated into English. In addition, a revised version
of one of the series was also translated into English. The content of this version
still followes the 1989 COS and is very similar, and often identical, to the version
that was examined in the study. Therefore, when quoting from those documents,
corresponding segments from the translated documents are used in this chapter.
Algebra in Japanese Curriculum

In the elementary school mathematics COS, the content is divided into four do-
mains: Numbers and Calculations, Geometric Figures, Quantities and Measure-
ments, and Quantitative Relations. In the Lower Secondary School (Grade 7 through
9) COS, there are three content domains: Number and Mathematical Expressions,
Geometric Figures, and Quantitative Relations. Some may be surprised that there
is no domain called algebra in these COS, even at the middle school level. How-
ever, further reading of the COS and Teaching Guide suggests that those topics that
are often considered as “algebra” are found mostly in the Quantitative Relations
domain. Teaching Guide for the 1989 COS states,
The contents of this domain include. . . items which are useful in examining or manipulating
contents in other domains. . . The objectives and contents of this domain cover a wide range,
but can be divided into three categories: idea of functions, writing and interpreting math-
ematical expressions, and statistical manipulation. (Takahashi et al. 2004, p. 36, emphasis
added)
Table 2 shows the grade level goal statements related to the Quantitative Rela-
tions domain from the 1989 COS and the focal points related to algebra from the
National Council of Teachers of Mathematics’ (NCTM) Curriculum Focal Points
(2006). From this table, we can easily see that these focal points align well with the
Math Panel’s Critical Foundations.
However, both the Japanese curriculum and the NCTM emphasize other aspects
of Elementary school mathematics that are related to algebra. Since the Quantita-
tive Relations strand does not start until Grade 3, there was no explicit goal related
to the domain in Grades 1 and 2. However, the Japanese COS certainly empha-
sizes the mastery of addition and subtraction operations in those grades. What the
Japanese COS calls “idea of functions” is comparable to what Focal Points calls
the examination of patterns in Grades pre-K through 5. However, in Japanese text-
books, there is almost no formal discussion of patterns such as repeating patterns
Table 2 Grade level goal statements in the quantitative relations domain of the Japanese Course
of Study and algebra focal points in NCTM’s Focal Points
1989 Japanese COS NCTM. Focal Points
Grade 1
Number and Operations and Algebra: De-
veloping understandings of addition and
subtraction and strategies for basic addition
Grades 1 & 2
facts and related subtraction facts.
Grade 2
Number and Operations and Algebra: De-
veloping quick recall of addition facts and
related subtraction facts and fluency with
multidigit addition and subtraction.
To help children become able to arrange Number and Operations and Algebra: De-
data, and to use mathematical expressions veloping understandings of multiplication
Grade 3
and graphs, and to help children appreciate and division and strategies for basic multi-
their meaning and become gradually able o plication facts and related division facts.
represent or to investigate sizes and quanti-
ties and their mathematical relations.
To help children become able to represent or Number and Operations and Algebra: De-
Grade 4
consider quantities and their mathematical veloping quick recall of multiplication facts
relations by using mathematical expressions and related division facts and fluency with
or graphs, and further, to help them become whole number multiplication.
able to investigate dependence relations be-
tween them.
To help children become able to concisely Number and Operations and Algebra: De-
Grade 5
represent mathematical expressions by us- veloping an understanding of and fluency

ing letters, and to investigate mathematical with division of whole numbers.
relations represented by them.
To help children deepen their idea of func- Algebra: Writing, interpreting, and using
Grade 6
tion through their understanding of propor- mathematical expressions and equations.

tion and become able to efficiently use it in
considering quantitative relations.
and growing patterns. Rather, the emphasis in the Japanese textbooks is on two co-
varying quantities. For example, in Grade 1, teachers are encouraged to organize the
results of decomposing 10 into two numbers so that children might notice how the
two numbers change, that is, as one number increases by 1, the other decreases by 1
(Watanabe 2008).
However, the Japanese curriculum also includes the study of mathematical ex-
pressions as an important component of the domain, which is not emphasized in
Focal Points. The Japanese word translated to “mathematical expressions” is shiki,
. This word cannot be translated into one English term as shiki includes expres-
sions such as 3 + 5, x − 4, and ÷ 3, as well as equations, 3 + 5 = 8, x − 4 = 7,
and ÷ 3 = 7. Moreover, even inequalities such as x + 5 > 2 are also considered as
shiki. In the Japanese mathematics curriculum, writing and interpreting shiki is con-
114 T. Watanabe
sidered an important component of algebra. In fact, one of the content domains of

Lower Secondary School Mathematics is titled “numbers and shiki.” A former Min-
istry of Education official even translated it as “Numbers and Algebra” (Yoshikawa
2008, p. 14).
Mathematical Expressions in Japanese Curriculum

The emphasis on mathematical expressions in the Japanese curriculum is clearly
evident in the following paragraph found in Teaching Guide for the 1989 COS:
Tables, diagrams, graphs, and mathematical expressions are used to represent numbers and
quantities and their relations in our daily life. Especially, mathematical expressions can be
said to be a good way to express relations among numbers and quantities accurately, simply,
and generally. By the way, mathematical expression is something in which certain symbols
are arranged according to specific rules. If mathematical expression is taught to be some-
thing that only expresses how to do calculations, or as a mere convention, the understanding
of the meaning that mathematical expression represents or how the expression works can
be insufficient. So, when teaching mathematical expressions, it is important to interpret ex-
pressions, to manipulate expressions, and to be able to explain the process of transformation
of expressions, as well as to become able to express phenomena and relations in concrete
situations. Especially, it is important to focus on understanding the meaning that the math-
ematical expressions represent. (Takahashi et al. 2004, p. 38)
It is clear that mathematical expressions are not simply indicating what arithmetic
operation is to be executed. Rather, mathematical expressions are used to represent
phenomena and relationships, to promote and facilitate mathematical thinking, and
to communicate reasoning processes.
In the following section, we will examine how Japanese textbooks deal with the
ideas related to mathematical expressions, giving us more concrete examples. The
textbook examples used in this section are from Hironaka and Sugiyama (2000), the
English translation of a revised version of one of the series analyzed in the study.
Mathematical Expressions in Japanese Textbooks

In Japanese textbooks, writing mathematical expressions is emphasized as soon as
children are introduced to the addition operation. Figure 1 shows the first 2 pages of
the Grade 1 unit that introduces the addition operation.
Notice that they introduce the “math sentence”1 as a way to represent what is
happening in the problem situation. Furthermore, in Problems 2 and 3, students
are gradually guided to write math sentences independently. Similar emphasis is
observed each time a new operation is introduced and when the range of numbers
with a specific arithmetic operation is expanded. Figure 2 shows two examples of
such occasions from the textbook series.
1 The original Japanese word here is shiki, but the translators decided to use the phrase “math
sentence” instead of “mathematical expression.”

Fig. 1 The first 2 pages of the Grade 1 unit on addition. Students are gradually guided to write
complete math sentences independently (Book 1, pp. 28 & 29)
Fig. 2 Students are asked to write a math sentence first when the range of number is expand-
ed—(a) Grade 1 sums greater than 10 (Book 1, p. 62), and (b) Grade 5 multiplication by decimal
numbers (Book 5A, p. 26)
The first three examples showed how this textbook series emphasizes the writing
of mathematical expressions starting from Grade 1. The emphasis on interpreting
mathematical expressions also begins in Grade 1. When a new arithmetic opera-
tion is introduced, the textbook often asks students to write stories/problems that go
with a specific mathematical expression. Figure 3(a) is found at the end of the first
unit on subtraction in Grade 1. Students have studied two meanings of subtraction,
take-away and comparison, in this unit. The illustration includes several instances
of both situations. Figure 3(b) comes from the multiplication unit in Grade 2. It is
worth noting here that this problem is given to students in the first sub-unit where
116 T. Watanabe
Fig. 3 Students are often asked to write stories/problems for the given math sentences (a, Book 1,
p. 43; b, Book 2B, p. 17)
the meaning and representation of multiplication are introduced. In other words, stu-
dents have yet to study the actual multiplication facts, 6 × 4. Therefore, the emphasis
is to interpret the mathematical expression.
The idea of using mathematical sentences to represent students’ own thinking
process is often developed through a problem like the one shown in Fig. 4.
This problem is found at the end of the two multiplication units in Grade 2.
The problem not only asks students to find different ways to find the total number
of chocolates but also to describe their thinking processes. Although the pupil’s
page does not explicitly state it, the intent is that students will use mathematical
expressions to represent their (or the ones presented in the book) thinking processes.
Thus, Mami’s thinking process may be represented by the following set of math
sentences:
6 × 3 = 18
3×2=6
18 + 6 = 24
In Grade 2, students write a set of math sentences, but in Grade 4, students learn
to write compound math sentences by using parentheses. Figure 5 shows the illustra-
tion found on the opening pages of the unit in which students learn to write, interpret
and calculate mathematical expressions involving parentheses.
Fig. 4 Students are asked to

represent thinking processes
using mathematical
expressions (Book 2B, p. 44)
Fig. 5 Grade 4 unit on compound math sentences—mathematical expressions with parentheses

(Book 4A, pp. 70 & 71)
It is interesting to note that students learn about the order of operations in the
context of writing compound math sentences. Naoko’s math sentences can be simply
combined into 1000 − 140 − 460. However, Makoto’s thinking cannot be written as
1000 − 140 + 460. From such situations, students realize the need for an agreement
on the order of operations and new notations for grouping of quantities.
118 T. Watanabe
Fig. 6 Grade 3 unit on writing mathematical expressions with (Book 3B, pp. 65 & 66)
In Grade 3, students also learn to write math sentences with missing quantities.
Figure 6 shows the first 2 pages of the unit in which students will learn to write math
sentences using the symbol, .
Prior to this unit, students have learned missing number type problems. For ex-
ample, students previously learned a missing addend subtraction problem similar to
the one shown in Fig. 6 in Grade 2. Therefore, solving these problems is not nec-
essarily the main focus of the unit. Rather, the emphasis is to help students write
mathematical expressions using for an unknown quantity. The textbooks use
to indicate unknown quantities in diagrams and even in some blank math sentences,
like 5 + = 8, but this unit is the first time where students are actually asked to
write math sentences with .
Symbols such as , , and are initially used as unknowns, and students are
often asked to find what number can replace the symbol. However, starting in Grade
4, students are exposed to these symbols used as variables. For example, in the unit
on writing mathematical expressions with parentheses, students study the distribu-
tive property formally. In that unit, the property is expressed using symbols as shown
in Fig. 7.
As you can see, students are encouraged to substitute different numbers to verify
the property. Through such experiences, students are expected to develop an under-
standing that multiple values can be substituted in a single symbol.
In Grade 4, students also learn to write mathematical expressions using sym-
bols such as, , , and , as variables. This is done in the context of exploring
relationships between two quantities. Figure 8 shows an example of such problems.
Fig. 7 Symbols are used as

variables in Grade 4 (Book
4A, p. 76)
Fig. 8 Students express the

relationship between two
quantities using mathematical
expressions with symbols
(Book 4B, p. 58)
Based on these experiences with mathematical expressions, students are intro-

duced to the use of letters as variables in Grade 5 in the 1989 COS and in Grade
6 in the 2008 COS (see Fig. 9). However, as you can see from Fig. 9, letters are
introduced as simply replacing familiar symbols such as , , and . Students
have already begun to learn about the concept of variables, and they have also stud-
ied how symbols can be used in mathematical expressions to represent relationships
among quantities. Furthermore, they are familiar with problems requiring finding
the missing number represented by those symbols (see Fig. 6). Thus, their focus in
this unit is really much more notational than conceptual.
In Fig. 6, you notice that the textbook includes two strategies to determine the
missing number, guess-then-adjust and using a diagram. Students learned how to
represent problems using the diagram like the one shown in Fig. 6 at the end of
Grade 2. The Japanese curriculum materials emphasize linear models in general,
and in Grade 4, the textbook introduces another type of linear model. Consider the
following problem:
Teams A and B are making 40 posters for the sports festival. Team B will make
8 more than Team A. How many posters will they each make?
Although it may be possible to represent this problem in a single tape/segment,
it may be difficult to use it if there are more than 2 quantities involved. Thus, the
textbook suggests a diagram like the one shown in Fig. 10. This diagram can easily
120 T. Watanabe
Fig. 9 Letters as variables are introduced after students have learned to work with mathematical
expressions with symbols like , , and (Book 5A, pp. 85 & 86)
Fig. 10 A new type of linear

model is introduced in Grade
4 (Book 4A, p. 40)
be adjusted even if there are more than 2 quantities. Furthermore, from the diagram,
students may be able to come up with different ways to calculate the answer. For
example, if Team A were to make 8 more posters, then the two teams would make
48, i.e., 40 + 8, posters altogether, each making the same number. So, dividing it by
2 will give you the number of posters Team B will make. On the other hand, if Team
B were to make 8 fewer posters, then the two teams would make 32, i.e., 40 − 8,
posters. Since each team would make the same number of posters, you can divide
32 by 2 to determine the number of posters Team A will make.
The textbook continues to use a linear model to represent various problems in
Grades 5 and 6. In Grade 5, students learn to use a linear model to solve problems
shown in Fig. 11. These problems can be solved using linear equations, or systems
of linear equations, but the focus here is for students to learn to represent problem
situations using linear diagrams. From the diagrams, students can identify different
solution methods, using only elementary school arithmetic. Clearly, each step of
the arithmetic solution processes correspond to the solution processes with linear
Fig. 11 Grade 5 textbook uses linear models to solve more complex problems (Book 5A, pp. 48
& 49)
equations. (See Watanabe et al. 2010 for a more detailed discussion of the way
Japanese textbooks introduce and develop diagrams to support students’ thinking.)
Discussion
As it was noted earlier, pre-K through Grade 5 focal points (NCTM 2006) are well
aligned with the content of the Math Panel’s Critical Foundations for School Mathe-
matics.2 In addition, NCTM recommends the importance of the study of patterns in
various “connections” in pre-K through Grade 5. Starting in Grade 6, Focal Points
begin emphasizing the understanding of mathematical (symbolic) Representations.
It is rather surprising that both the Math Panel and NCTM completely omit the
reference to written/symbolic representations in elementary grades. The Math Panel
seems to focus primarily on computational fluency. Even though they use the term
“proficiency,” they state that by the term, “the Panel means that students should
understand key concepts, achieve automaticity as appropriate (e.g., with addition
and related subtraction facts), develop flexible, accurate, and automatic execution
of the standard algorithms, and use these competencies to solve problems” (p. xvii).
This emphasis seems to be consistent with the Panel’s major topics of algebra, which
seems to take on the perspective of algebra as generalized arithmetic.
NCTM’s Focal Points, in alignment with their Standards (NCTM 2000), take a
much broader view of algebra, which incorporates all four conceptualizations of al-
gebra discussed earlier. Thus, they emphasize the importance of the study of patterns
and functions in pre-K through Grade 5. NCTM (2000) also discusses how learning
about various properties of operations is the beginning of algebraic thinking. Focal
Points also makes references to representations several times, but they seem to focus
on representations using concrete materials or visual representations such as graphs
and diagrams.
In contrast, the Japanese curriculum documents consider the study of functional
relationships (patterns) and the ideas related to mathematical expressions (shiki) as
2 Note that neither Focal Points nor the Math Panel report is a mathematics curriculum. They are
mentioned in this discussion because both documents offer frameworks for a school mathematics
curriculum.
122 T. Watanabe
the two pillars of elementary school algebra. Moreover, the Japanese curriculum
materials emphasize the writing and interpreting of mathematical expressions as a
major focus in the domain of quantitative relations. In addition, the Japanese cur-
riculum considers that a goal of this domain “is to understand the contents of other
domains using the ideas and methods discussed in this domain” (Takahashi et al.
2004, p. 36).
This emphasis on mathematical expressions makes sense mathematically. For ex-
ample, a Grade 4 focal point discusses the importance of students’ understanding of
the distributive property. However, it seems like a true appreciation and understand-
ing of this property requires students to write composite mathematical expressions.
Consider the following problem:
A fruit basket contains 5 apples, 8 oranges, and 4 bananas. If you buy 3 baskets,
how many fruits are there all together?
Students can write their solution steps using sets of math sentences:
3 × 5 = 15
5 + 8 + 4 = 17 3 × 8 = 24
, or
17 × 3 = 51 3 × 4 = 12
15 + 24 + 12 = 51
Clearly, it is important for students to understand that both of these thinking pro-
cesses are mathematically valid. However, to understand the distributive prop-
erty they must understand that each set of math sentences can be written as a
composite math sentence and that those two math sentences are equal; that is,
3 × (5 + 8 + 4) = 3 × 5 + 3 × 8 + 3 × 4. Similarly, the need for understanding
of the order of operations seems to arise when students write compound mathemat-
ical expressions.
Perhaps mathematical/symbolic representations are not receiving much atten-
tion because there is a perception that formal algebra instruction in the past over-
emphasized the symbolic manipulation of mathematical expressions (Kilpatrick and
Izsak 2008). Whether or not that perception is correct, the fact remains that symbolic
manipulation is a component of school algebra. With the development of computing
technologies, actual manipulation of mathematical expressions themselves may not
be as critical as it used to be. However, the availability of computing technologies
increases the importance of students’ ability to write and interpret mathematical ex-
pressions. Moreover, it is widely known that many children consider the equal sign
as simply a do-something symbol, and such an understanding can hinder their un-
derstanding of algebra (for example, Knuth et al. 2006). Kieran (2004) lists “a refo-
cusing on the meaning of equal sign” (p. 141) as one of the adjustments from arith-
metical thinking to algebraic thinking. The Japanese curriculum materials, however,
seem to take the position that mathematical expressions, not just the equal sign, are
a central feature of mathematics.
Overall, the treatment of algebra in the Japanese elementary school mathematics
curriculum has many similarities to those of other countries. Kieran (2004) noted
that the emphasis on quantitative relationships is one commonality among the el-
ementary school curricula from China, Korea, Russia, Singapore, and the United
States. The Japanese curriculum also shares the same emphasis. The Japanese cur-
riculum introduces literal symbols in upper elementary grades like the other Asian
curricula. The explicit emphasis on helping students develop diagrams to support
students problem solving is similar to the Singaporean approach. However, the em-
phasis on writing and interpreting mathematical expressions seems to be a unique
feature of the Japanese approach. Fujii (2003) noted “the importance of recogniz-
ing the potentially algebraic nature of arithmetic, as distinct from trying to move
children from arithmetic to algebra” (p. 62). The Japanese elementary mathematics
curriculum seems to embody this perspective.
References
Blanton, M. L., & Kaput, J. J. (2005). Characterizing a classroom practice that promotes algebraic
reasoning. Journal for Research in Mathematics Education, 36, 412–446.
Cai, J. (2004). Developing algebraic thinking in the early grades: A case study of the Chinese
elementary school curriculum. The Mathematics Educator (Singapore), 8(1), 107–130.
Fujii, T. (2003). Probing students’ understanding of variables through cognitive conflict problems:
Is the concept of a variable so difficult for students to understand? In N. A. Pateman, B. J.
Dougherty, & J. T. Zilliox (Eds.), Proceedings of the 27th Conference of the International
Group for the Psychology of Mathematics Education (PME) (Vol. 1, pp. 49–65).
Hironaka, H., & Sugiyama, Y. (2000). New Elementary Mathematics, Books 1 Through 6B. Tokyo:
Tokyo Shoseki.
(Singapore), 8(1), 139–151.
Kilpatrick, J., & Izsak, A. (2008). A history of algebra in the school curriculum. In C. E. Greenes
& R. Rubenstein (Eds.), Algebra and Algebraic Thinking in School Mathematics: Seventieth
Yearbook (2008 Yearbook of the National Council of Teachers of Mathematics) (pp. 3–18).
Reston, VA: NCTM.
Knuth, E., Stephens, A. C., McNeil, N. M., & Alibali, M. W. (2006). Does understanding the equal
sign matter? Evidence from solving equations. Journal for Research in Mathematics Education,
37, 297–312.
Lew, H. (2004). Developing algebraic thinking in the early grades: Case study of Korean elemen-
tary school mathematics. The Mathematics Educator (Singapore), 8(1), 88–106.
Moyer, J., Huinker, D., & Cai, J. (2004). Developing algebraic thinking in the early grades: A case
study of the US Investigations curriculum. The Mathematics Educator (Singapore), 8(1), 6–38.
matics. Reston, VA: NCTM.
National Council of Teachers of Mathematics (2006). Curriculum Focal Points for Prekindergarten
Through Grade 8 Mathematics: A Quest for Coherence. Reston, VA: The Council.
Ng, S. (2004). Developing algebraic thinking in early grades: Case study of the Singapore primary
mathematics curriculum. The Mathematics Educator (Singapore), 8(1), 39–59.
Schmittau, J., & Morris, A. (2004). The developing algebraic thinking in the elementary mathe-
matics curriculum of V. V. Davydov. The Mathematics Educator (Singapore), 8(1), 60–87.
Takahashi, A., Watanabe, T., & Yoshida, M. (2004). Elementary School Teaching Guide for the
Japanese Course of Study: Arithmetic (Grades 1–6). Madison, NJ: Global Education Re-
sources.
124 T. Watanabe
Usiskin, Z. (1988). Conceptions of school algebra and uses of variable. In A. F. Coxford & A. P.
Shulte (Eds.), The Ideas of Algebra, K-12, 1988 Yearbook of the National Council of Teachers
of Mathematics (pp. 8–12). Reston, VA: NCTM.
Watanabe, T., Takahashi, A., & Yoshida, M. (2010). Supporting focused and cohesive curricula
through visual representations: An example from Japanese textbooks. In B. J. Reys & R. E.
Reys (Eds.), Mathematics Curriculum: Issues, Trends, and Future Directions, 2010 Yearbook
of the National Council of Teachers of Mathematics (pp. 131–143).
Yoshikawa, S. (2008). Education ministry perspectives on mathematics curriculum in Japan. In
Z. Usiskin & E. Wilmore (Eds.), Mathematics Curriculum in Pacific Rim Countries—China,
Japan, Korea, and Singapore (pp. 9–22). Charlotte, NC: Information Age Publishing.
Watanabe, T. (2008). Algebra in elementary school: A Japanese perspective. In C. Greenes (Ed.),
Algebra and Algebraic Thinking in School Math: 70th Yearbook (pp. 183–194). Reston, VA:
NCTM.
Commentary on Part I
Jeremy Kilpatrick
During the nineteenth century, the study of algebra moved into the secondary school
curriculum as colleges and universities began to require it for admission (Kilpatrick
and Izsák 2008). Coming after an extensive treatment of arithmetic in the elemen-
tary grades, school algebra was commonly introduced formally as a generalization
of that arithmetic, with an emphasis on symbol manipulation and equation solving.
Given the well-established status of algebra in the secondary curriculum, mathe-
matics educators today confront the question of, in the words of Subramaniam and
Banerjee, how “to manage the transition from arithmetic to symbolic algebra.”
The U.S. National Mathematics Advisory Panel (2008, pp. 17–18) addressed the
transition question by identifying what the panel called Critical Foundations for
School Algebra, which comprise three clusters of concepts and skills: (1) fluency
with whole numbers, (2) fluency with fractions, and (3) particular aspects of geom-
etry and measurement. Watanabe observes that although those items are certainly
foundational, they are neither exhaustive nor even the components of the elemen-
tary school grades mathematics curriculum that are, perhaps, or ought to be, most
directly related to school algebra. As Blanton and Kaput note, although the National
Mathematics Advisory Panel focused on getting learners ready for the study of al-
gebra,
experiences in building, expressing, and justifying mathematical generalizations . . . should
be a seamless process that begins at the start of formal schooling, not content for later grades
for which elementary school children are “made ready” through a singular, myopic focus
on arithmetic.
Russell, Schifter, and Bastable pose the curriculum question differently: “How
can work in [early] algebra fit into an already crowded curriculum?” Their response
is echoed in the other chapters in this part: early algebra “not only provides crucial
links between arithmetic and algebra, but also is an essential part of good arithmetic
instruction.”
J. Kilpatrick ()
Department of Mathematics and Science Education, University of Georgia, Athens, USA

126 J. Kilpatrick
Algebra First
The dominant theme of the six chapters offering a curricular perspective on early
algebra learning is that algebra is not something to be postponed until arithmetic
has been mastered but instead ought to be present in the curriculum from the begin-
ning. This approach has been taken in those school mathematics curricula in which
arithmetic is handled in general terms and functional thinking is encouraged. As
Izaak Wirszup noted, such an approach characterizes school mathematics in some
countries but not others:
In Europe, in other countries [including Russia], they start arithmetic combined with ge-
ometry, and arithmetic is not calculations alone. They introduce algebraic concepts and
algebraic thinking from the first grade, from kindergarten. Variables, place holders, empty
squares, question marks, these are the way variables are smuggled in without explaining,
only through examples. And in the elementary grades you learn about equations, inequali-
ties, and systems of equations. So all this experience results in the fact that the transition in
Europe from arithmetic to algebra is almost invisible, while in the United States after some
eight years of endless, meaningless calculations, you are given a one-year algebra course,
the first-year algebra course, with variables and polynomials and exponents and equations.
(quoted in Roberts 2010, p. 56)
The most radical argument for incorporating algebra into the curriculum of the
early grades, as explained by Schmittau, was given by Vygotsky, who argued not
only that thinking moves from the abstract to the concrete rather than vice versa
but also that consequently the algebraic “plane of thought” needs to be developed
before arithmetic is developed. Vygotsky’s follower Davydov said that rather than
viewing the process of generalization as an induction from examples,
the specific examples should be seen as carrying the generalization within them; the gen-
eralization process ought to be one of enrichment rather than impoverishment. Instead of
thinking of generalization as moving from the concrete to the abstract, we should think of
it as beginning with the abstract and moving to the “intellectually concrete” and then on to
an enriched abstraction. (Kilpatrick 1990, pp. xv–xvi)
Implementing the elementary curriculum developed by Davydov, Schmittau demon-

strates the value of a systematic, theory-based approach to algebraic structure in
helping children learn to solve challenging mathematics problems.
Less radical but certainly equally challenging approaches to early algebra are out-
lined by the other authors in the section. Blanton and Kaput emphasize functional
thinking and the need to consider both covariation and correspondence between
variables; Cai, Ng, and Moyer discuss some curriculum practices in China and Sin-
gapore that promote algebraic thinking; Russell, Schifter, and Bastable elaborate
mathematical activities that can provide a bridge between arithmetic and algebra;
Subramaniam and Banerjee show how children’s arithmetic intuition can be used to
help them interpret and evaluate numerical and algebraic expressions; and Watan-
abe demonstrates how Japanese textbook materials help children write and interpret
symbolic mathematical expressions. In every case, arithmetic is treated as not sim-
ply a venue for learning and doing calculations but rather as an arena for developing
children’s ideas about quantities and their interrelationships, representations, and
use.
Commentary on Part I 127
A Curriculum Topic
Several of the authors in the section point out that introducing algebraic thinking
into the elementary school grades entails not the introduction of new topics but
rather new approaches to existing topics. The question of how the curriculum is
affected, therefore, depends on one’s concept of the curriculum. If curriculum is a
topic list, nothing changes. But if curriculum is the set of experiences that learners
have, then the change can be profound. To illustrate, consider a common topic from
the elementary school curriculum: the area of a rectangle as the product of its length
and width.
This topic commonly appears about the time that children are learning the oper-
ation of multiplication of whole numbers. In the new Portuguese program of math-
ematics for basic education (Ponte et al. 2007), for example, students in the third
and fourth grades are expected to understand and use formulas to calculate the area
of the square and the rectangle (p. 25). In the United States, several of the Grade
3 measurement and data standards in the Common Core State Standards Initiative
(see http://www.corestandards.org/the-standards/mathematics) address the area of a
rectangle. For example, one standard reads as follows: “Multiply side lengths to
find areas of rectangles with whole-number side lengths in the context of solving
real world and mathematical problems, and represent whole-number products as
rectangular areas in mathematical reasoning.”
A third-grade teacher who wanted to follow the approach advocated by Blanton
and Kaput would hardly stop at having the children find the areas of some rectangles
given their sides. The children would also look at patterns formed as sides and areas
vary together. They might, for example, explore how, for a fixed width, changes in
the length are reflected in changes in the area or how, for a length 2 units longer
than a varying width, the width and area covary. Similarly, a teacher who wanted
to pursue the Chinese “one problem multiple solutions” approach outlined by Cai,
Ng, and Moyer might offer the children a problem like the following: The area of
a rectangle is 120 square centimeters, and one side is 16 centimeters. How long
is the other side? The children would be asked to represent the quantitative rela-
tionship in several different ways both arithmetic and algebraic. A teacher wanting
to develop algebraic thinking by engaging in the approach promoted by Subrama-
niam and Banerjee might ask students to judge whether rectangles with dimensions
33 × 14 and 11 × 42 have the same area and to explain why or why not. In ev-
ery case, the children would have opportunities to investigate and use the aspects of
arithmetic identified by Russell, Schifter, and Bastable: “understanding the behavior
of the operations [in this case multiplication and division], generalizing and justify-
ing, extending the number system [in this case to fractions or decimals], and using
notation with meaning.” Few teachers would want to deny children opportunities to
develop such understanding and ways of thinking simply because the elementary
school curriculum is “already crowded.”
128 J. Kilpatrick
Numerical Patterns
The National Mathematics Advisory Panel (NMAP 2008) made a strong recom-
mendation “that ‘algebra’ problems involving patterns should be greatly reduced
in the NAEP [National Assessment of Educational Progress]” (p. 59), arguing that
“at Grade 4, most of the NAEP algebra items relate to patterns or sequences (Daro
et al. 2007)” (p. 59) and that neither mathematical considerations nor comparative
analyses of curricula support “the prominence given to patterns in PreK-8” (p. 59).
Regarding comparative analyses, the panel cites a paper by Schmidt and Houang
(2007) in making its claim that “patterns are not emphasized in high-achieving
countries” (NMAP 2008, p. 59). According to the Schmidt and Huang analysis,
a majority of the countries whose eighth graders were high achieving in the 1995
Trends in International Mathematics and Science Study (TIMSS) did not address
the topic of “patterns, relations, and functions” until Grade 8.
The NMAP claim, however, is not supported by the chapters at hand. Cai, Ng,
and Moyer point out that in Singapore, already in second grade, children are looking
at “sets of ordered pairs representing the input and output of problem situations” and
are being asked to find rules for the patterns they are observing in those situations.
Watanabe argues that the “idea of functions” in the Japanese course of study is
essentially the same thing as the American “examination of patterns in Grades Pre-
K through 5.” And one might note that even though China did not participate in
the 1995 TIMSS, in the words of Cai, Ng, and Moyer, “function ideas permeate the
[elementary school] curriculum.”
In the Daro et al. (2007) study of the validity of the NAEP mathematics assess-
ment, mathematicians who reviewed the items complained that there were not only
too many items dealing with patterns but also too many that were flawed because
they asked that a sequence be extended without specifying the rule for pattern gen-
eration. Those complaints are apparently the source of the NMAP’s argument that
“mathematical considerations” do not support prominent attention being given to
numerical patterns in the first eight grades. Which is worse: asking children, given a
sequence of numbers, to conjecture what the rule might be that would give the next
one, or denying them an opportunity to address such problems because, of course,
neither the rule nor the next term is unique mathematically? Fortunately, the chap-
ters at hand support efforts to give children chances to explore numerical patterns.
Word Problems
Consider the following word problem:

Some pupils gathered 1800 kilograms of maple and acacia seeds. There were five times as
many acacia seeds gathered as maple seeds. How many kilograms of seeds of each type
were gathered? (Yaroshchuk 1969, p. 79)
A problem of this sort might easily appear in a first-year algebra textbook in the
United States, with the expectation that students would solve it using two equations
Commentary on Part I 129
and two unknowns—the number of kilograms of maple seeds and the number of
kilograms of acacia seeds. In the Soviet Union a half century ago, however, it was
given to fourth-grade students of average ability, and they were expected to solve
it by recognizing that it is a problem “in parts” and that they needed to solve it by
finding the total number of parts (six) and then finding the size of one part.
One need not have children recognize the problem as being of a certain “type” to
appreciate that they ought to be solving such problems well before they have learned
to solve systems of linear equations. By reasoning through the arithmetic operations
in the problem along the lines promoted in the various chapters in this part, children
can learn how such problems are structured, learn how to represent them in various
ways, and begin to develop their algebraic thinking skills.
Although the chapters in this monograph are divided among three perspectives, one
can see in the six chapters on the curricular perspective that other perspectives are
present as well. For example, Blanton and Kaput claim that their work shows that
children are “capable of deeper functional analysis than previously thought” and
that those ideas “appear at grades earlier than typically expected.” Further, they ar-
gue that curriculum change alone, without attention to developing teachers’ knowl-
edge of both instruction and mathematics, is “not sufficient to produce real change
in children’s mathematical thinking.” Thus, cognitive and instructional perspectives
play a prominent role in their chapter. Similarly, the other chapters offer ample evi-
dence that young children are capable of impressive feats of algebraic reasoning and
that proposed changes in the arithmetic curriculum ought to be accompanied by in-
structional changes as well. As Subramaniam and Banerjee conclude, children need
to learn arithmetic as more than “a set of instructions to carry out operations”; the
expressions of arithmetic and algebra encode quantitative relationships that children
can and should learn to reason with.
References
Daro, P., Stancavage, F., Ortega, M., DeStefano, L., & Linn, R. (2007, September). In Validity
Study of the NAEP Mathematics Assessment: Grades 4 and 8. Palo Alto, CA: NAEP Validity
Studies, American Institutes for Research. Available from http://www.eric.ed.gov/.
Kilpatrick, J. (1990). Introduction to the English language edition. In V. V. Davydov (Ed.), Types
of Generalization in Instruction: Logical and Psychological Problems in the Structuring of
School Curricula, Soviet Studies in Mathematics Education (Vol. 2, pp. xv–xviii). Reston, VA:
Kilpatrick, J., & Izsák, A. (2008). A history of algebra in the school curriculum. In C. E. Greenes &
R. Rubenstein (Eds.), Algebra and Algebraic Thinking in School Mathematics, 70th Yearbook
of the National Council of Teachers of Mathematics (pp. 3–18). Reston, VA: NCTM.
National Mathematics Advisory Panel. Washington, DC: U.S. Department of Education. Avail-
able from http://www2.ed.gov/about/bdscomm/list/mathpanel/index.html.
130 J. Kilpatrick
Ponte, J. P., Serrazina, L., Guimarães, H. M., Breda, A., Guimarães, F., Sousa, H., Graça
Martins, M. E., & Oliveira, P. A. (2007, December). Programa de matemática do ensino
básico. Lisboa: Ministério da Educação. Available: http://www.dgidc.min-edu.pt/matematica/
Documents/ProgramaMatematica.pdf.
Roberts, D. L. (2010). Interview with Izaak Wirszup. International Journal for the History of
Mathematics Education, 5(1), 53–74.
Schmidt, W. H., & Houang, R. T. (2007). Lack of focus in the mathematics curriculum: Symptom
or cause. In T. Loveless (Ed.), Lessons Learned: What International Assessments Tell Us About
Math Achievement (pp. 65–84). Washington, DC: Brookings Institution Press.
Yaroshchuk, V. L. (1969). A psychological analysis of the processes involved in solving model
arithmetic problems. In J. Kilpatrick & I. Wirszup (Eds.), Problem Solving in Arithmetic and
Algebra, Soviet Studies in the Psychology of Learning and Teaching Mathematics (Vol. 3, pp.
53–96). Stanford, CA, & Chicago, IL: School Mathematics Study Group & Survey of Recent
East European Mathematical Literature.
Part II
Cognitive Perspective
Preface to Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Eric Knuth
Department of Curriculum & Instruction, University of Wisconsin-
Madison, Madison, USA
Jinfa Cai
Newark, USA
Algebraic Thinking with and without Algebraic Representation:

A Pathway for Learning . . . . . . . . . . . . . . . . . . . . . . . . . 137
Faculty of Education, The University of Auckland, Auckland,
New Zealand
Examining Students’ Algebraic Thinking in a Curricular Context:

A Longitudinal Study . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Jinfa Cai
Newark, USA
John C. Moyer
Department of Mathematics, Statistics and Computer Science,
Marquette University, Milwaukee, USA
Ning Wang
Center for Education, Widener University, Chester, USA
Bikai Nie
Newark, USA
Years 2 to 6 Students’ Ability to Generalise: Models, Representations

and Theory for Teaching and Learning . . . . . . . . . . . . . . . . . 187
Tom J. Cooper
Queensland University of Technology, Brisbane, Australia
132 Cognitive Perspective
Elizabeth Warren
Australian Catholic University, Brisbane, Australia
Algebra in the Middle School: Developing Functional Relationships

Through Quantitative Reasoning . . . . . . . . . . . . . . . . . . . . 215
Amy B. Ellis
School of Education, University of Wisconsin-Madison, Madison, USA
Representational Competence and Algebraic Modeling . . . . . . . . . . 239

Andrew Izsák
Department of Mathematics and Science Education, University of
Georgia, Athens, GA, USA
Middle School Students’ Understanding of Core Algebraic Concepts:

Equivalence & Variable . . . . . . . . . . . . . . . . . . . . . . . . . 259
Eric J. Knuth
Martha W. Alibali
Department of Psychology, University of Wisconsin-Madison, Madison,
USA
Nicole M. McNeil
Department of Psychology, University of Notre Dame, Notre Dame, USA
Aaron Weinberg
Department of Mathematics, Ithaca College, Ithaca, USA
Ana C. Stephens
Wisconsin Center for Education Research, University of Wisconsin-
An Approach to Geometric and Numeric Patterning that Fosters Second

Grade Students’ Reasoning and Generalizing about Functions and
Co-variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Joan Moss
Department of Human Development and Applied Psychology, Ontario
Institute for Studies in Education, University of Toronto, Toronto, Canada
Susan London McNab
Ontario Institute for Studies in Education, University of Toronto,
Toronto, Canada
Grade 2 Students’ Non-Symbolic Algebraic Thinking . . . . . . . . . . . 303

Luis Radford
École des science de l’education, Laurentian University, Sudbury,
Ontario, Canada
Cognitive Perspective 133
Formation of Pattern Generalization Involving Linear Figural Patterns

Among Middle School Students: Results of a Three-Year Study . . . 323
Department of Mathematics, San José State University, San José, USA
Commentary on Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

Bharath Sriraman
Department of Mathematical Sciences, University of Montana,
Missoula, USA
Kyeong-Hwa Lee
Department of Mathematics Education, Seoul National University,
Seoul, Korea
Preface to Part II
Enhancing the nature of algebra in the earlier grades requires substantial effort on
the part of teachers insofar as they are responsible for ensuring students have the
means as well as the opportunities to engage in learning to reason algebraically.
Such efforts, however, are predicated in part on understanding the nature of stu-
dents’ algebraic thinking and ways to foster its development. The chapters in this
part present research that focuses on the development of students’ algebraic thinking
across a range of grade levels (from primary grades to intermediate/middle grades).
Many of the chapters also underscore different schools of thought regarding early
algebra; in particular, these chapters illustrate the nature of students’ algebraic think-
ing as it develops within primarily arithmetic contexts (e.g., generalizing about arith-
metic computations) as well as within primarily algebraic contexts (e.g., functions
and functional thinking). As a collection, the chapters in this section illustrate that
regardless of grade level, arithmetic or algebraic context, or even country, young
children are very capable of developing the ability to think algebraically.
The chapter by Britt and Irwin provides compelling evidence that emphasizing
algebraic thinking within arithmetic has a positive influence on students’ algebraic
thinking in later grades. They present results from two studies that suggest that stu-
dents in the intermediate grades who were provided with early algebra experiences
in the primary grades outperform on algebraic tasks their peers who received a more
traditional curriculum and, moreover, such early algebra experience continues to pay
dividends when students are in secondary school. The chapter by Cooper and War-
ren also considers the influence of an algebraic emphasis during the primary grades
on the development of students’ algebraic thinking. In their work, they focus pri-
marily on students’ abilities to generalize from a variety of both arithmetic-based
E. Knuth ()
Department of Curriculum & Instruction, University of Wisconsin-Madison, Madison, USA
J. Cai

136 E. Knuth and J. Cai
situations (e.g., compensation principles generalized from computational problems)

and representational forms. Their results highlight the importance that understand-
ing and communicating the features of representational forms play in students’ abil-
ities to generalize. Moss and London McNab, in their chapter, also provide evidence
that children in the early grades can learn to reason algebraically; in particular, their
work illustrates the development of students’ thinking about linear functions and co-
variation in the context of growing patterns. Interestingly, their work suggests that
the instructional emphasis on linear functions and co-variation also had a positive
effect on aspects of students’ understanding of multiplication—a staple of instruc-
tion in a traditional arithmetic curriculum. Finally, in Radford’s chapter, he discusses
from an epistemological perspective the relationship between arithmetic and alge-
braic thinking, addressing the question of what counts as arithmetic and what counts
as algebra. He then situates this discussion in the context of young children’s first
encounter with algebraic concepts, and also uses this context to discuss the limits
and possibilities of introducing algebra in the early grades.
The remaining chapters in this part continue the focus on students’ algebraic
thinking, however, the focus now shifts to the intermediate/middle grades. Knuth
and colleagues present results from a cross-sectional study that focused on students’
understanding of two fundamental algebraic ideas (equivalence and variable)—
ideas that can be developed in the primary grades—and how their understanding
affected their performance on algebraic tasks that required use of these ideas. In
the chapter by Cai and his colleagues, they compare the effects of curricula on the
longitudinal development of students’ algebraic thinking. In particular, they com-
pare the effects on students’ algebraic thinking of a reform-based curriculum that
takes a functional approach to teaching algebra to a more traditional curriculum that
takes a structural approach to teaching algebra. A functions-based approach also
underscores the chapters by Ellis, Izsák, and Rivera and Becker. In her chapter, El-
lis illustrates how building on students’ capabilities to reason with quantities can
serve as a powerful means of fostering the development of student’s understanding
of linear and quadratic functions. In Izsák’s chapter, he reviews three decades worth
of research on students’ understanding of algebraic and graphical representations
of functions. He then discusses two significant advances in this area of research:
insight into students’ criteria for evaluating representations as well as into students’
coordination of shifts within and between representations and problem situations.
Finally, in the chapter by Rivera and Becker, they detail the results of a longitudi-
nal study that focused on students’ abilities to generalize the underlying functional
relationship for various linear patterns. They provide a detailed account of the de-
velopment of students’ abilities to generalize and the factors that influenced their
development.
Although the chapters in this part focused on students’ algebraic thinking and
its development from a variety of perspectives, collectively, the chapters present a
common message: students are capable of learning to reason algebraically in the
early grades (prior to formal algebra), and curriculum and instruction should build
on such capabilities.
Algebraic Thinking with and without Algebraic
Representation: A Pathway for Learning
Abstract The origins of algebraic thinking precede understanding of arithmetic,

as shown in a study of children aged 4–7. A mathematics curriculum introduced
in some New Zealand schools in 1999, The New Zealand Numeracy Project, now
encourages this algebraic thinking within arithmetic. The underlying framework for
this curriculum is described, with examples of the type of thinking encouraged. The
effect of this emphasis on the algebra underpinning arithmetic operations was ex-
amined in two further studies. One of these involved students in their final year of
elementary and intermediate school, at age 12. This study showed that on a test that
focused on students’ awareness of the underlying algebraic structure of arithmetic,
those students who had been included in the new curriculum in its early stages out-
performed those who had received a traditional curriculum. A later study followed a
cohort of students who received the new curriculum through their two intermediate
school years (aged 11–12) and into their first year of high school at age 13, when
traditional algebra is introduced. The results of this study showed that students who
had developed their understanding of the interrelationship of mathematical relation-
ships for additive, multiplicative and proportional operations could display this un-
derstanding algebraically. The ramifications of these findings for further teaching
algebraic thinking with or without algebraic representation led to a proposal for a
‘pathway for algebraic thinking’ accessible to all students.
We are grateful to the New Zealand Ministry of Education for funding this research and to R. J.
Irwin for help with the statistical analysis. This chapter is a revised version of an article published
in ZDM—International Reviews on Mathematical Education), 40(1), 39–53. DOI
10.1007/s11858-007-0064-x.
M.S. Britt () · K.C. Irwin
Faculty of Education, The University of Auckland, Auckland, New Zealand
K.C. Irwin

138 M.S. Britt and K.C. Irwin
Introduction
In a plenary address to the North American Chapter of the international Group
for the Psychology of Mathematics Education, Radford (2006) presented a com-
pelling argument in support of a generalization approach to algebra in which he
noted that “the algebraic generalization of a pattern rests on the noticing of a lo-
cal commonality that is then generalized to all terms of the sequence and serves
as a warrant to build expressions of elements of the sequence that remains be-
yond the perceptual field” (p. 5). As well, Radford added a third element, us-
ing the commonality to provide a direct expression or rule to specify any term
of the sequence. But while it is customary to require that learners use the sym-
bols of algebra to express such rules, Radford acknowledged that rule-making pro-
ceeds through various layers of awareness articulated through different semiotic
systems; words, gestures, pictures, graphs and symbols. We illustrate some of these
layers of awareness of generality in a study that contributed to an evaluation of
The New Zealand Numeracy Project (New Zealand Ministry of Education 2007a,
http://www.nzmaths.co.nz/teaching-numeracy), a national project in which students
throughout New Zealand are encouraged to devise and experiment with a range of
mental operational strategies in arithmetic (Irwin and Britt 2005a). In that study we
argued that students who could apply a range of mental operational strategies to
solve different numerical problems were disclosing an awareness of the relation-
ships of the numbers involved as well as the underlying structure of the strategy.
We claimed that successful application of such operational strategies demanded an
awareness of the generality of the operational strategy, thereby illustrating algebraic
thinking. Students’ explanation of their thinking revealed that they were treating the
numbers as if they were variables. Fujii and Stephens (2001) refer to numbers used
in this way as quasi-variables. The results of our study led to a view of algebra, par-
ticularly as it pertains to the pedagogy of introductory algebra, in which we do not
see algebra as following arithmetic so that arithmetic has an ending that coincides
with the beginning of algebra. Instead our view is consonant with those of Hewitt
(1998, p. 20) who argued that algebra enables arithmetic to be carried out, and of
Steffe (2001, p. 563), who argued that children’s knowledge of number together
with numerical operational knowledge that is effective and reliable is essentially al-
gebraic in nature. We further argue that the roots of this algebraic, or generalized,
thinking precede the introduction of numbers as unschooled children demonstrate
an understanding of such operations with unnumbered quantities.
While Carraher et al. (2006) and several of the studies included in Kaput et al.
(2008) argue for the early inclusion of algebraic symbols as a valuable tool for
early algebraic thinking (e.g. Brizuela and Earnest 2008; Carraher et al. 2008; and
Dougherty 2008), we have primarily followed the reasoning of Fujii and Stephens
(2001) and Hewitt (1998) that emphasizes algebraic thinking in order to understand
arithmetic. Mason (2008) also pointed to children’s innate power to reason and gen-
eralize in a manner that leads them to understanding the underlying structure of
arithmetic that is central to mathematical development. Our case, which focuses on
this crucial role for generalization, is consistent with those expressed by this sec-
ond group of writers. Prior to their introduction to the semiotics of algebra, young
Algebraic Thinking with and without Algebraic Representation 139
children need to work successfully with several layers of awareness of generality

that involve expression of generality in words, in pictures and graphs as well as with
numerical symbols acting as quasi-variables.
Much has been written about the difficulties encountered during the transition
from arithmetic to algebra (see for example, Herscovics and Linchevski 1994;
Filloy and Rojano 1989). But as Carraher et al. (2006, p. 89) argue, acceptance
of such a transition arises from an impoverished view of elementary school mathe-
matics in which mathematical generalization is postponed until the onset of algebra
instruction. We argue similarly. The notion of algebra in arithmetic, in which gener-
alization provides the basis for successful numerical operational thinking, dismisses
the claim for transition and offers algebra for all through algebraic thinking with
and without the symbols of algebra.
In the discussions that follow we refer to students by year group, which is the
New Zealand nomenclature and because years of schooling do not necessarily match
school grades in the US and elsewhere. All children in New Zealand begin school
on their 5th birthday and that year of entry to compulsory schooling is called Year 0
or Year 1 depending on when the birthday falls. While they may be moved around
between classes in the first two to three years at school, they usually remain in the
same year group from Year 3 onwards. The primary school system goes from Year
0 through Year 8 (roughly age 5 to 12). While some children remain in the same
primary school up to the end of Year 8, most children attend a separate interme-
diate school for Years 7 and 8 (ages 11 to 12). Secondary school begins at Year 9
(typically aged 13) and goes on for five years to Year 13 (about age 17). Algebra is
traditionally introduced formally in Year 9. The curriculum areas algebra, geome-
try, trigonometry, and statistics are not separate subjects as they often are in the US.
They are taught with different emphases within the curriculum heading, Mathemat-
ics and Statistics.
In this chapter, we begin our discussion of algebraic generality by illustrating
how the development of algebraic thinking can evolve, without recourse to the sym-
bols of algebra, from students developing an ongoing awareness of the underlying
structure of operational strategies in arithmetic.
First we describe a study that demonstrated preschool children’s understanding of
generalization in operations. Next we draw on the New Zealand Numeracy Project
to illustrate the development of operational strategies in arithmetic for students at
different ages, and then we focus on two recent studies in which we attempted to
ascertain what level of strategy development was likely to be necessary for stu-
dents to extend their expressions of generality from using numbers themselves as
quasi-variables to a semiotic layer of awareness that embraces the literal symbols of
algebra.
Children’s Understanding of Generalities for Operations Before

Schooling
While the main focus of this chapter is the effect of the New Zealand Numeracy
Project, it is useful to discuss an earlier study (Irwin 1996). This study explored
young children’s understanding of operations before they are introduced to arith-

metic. Children aged 4 through 7 were asked what happened to a total quantity
under three conditions. The total quantity was made up of two parts, in this case,
two small boxes of sweets, called “lollies” in New Zealand. They were asked what
happened to the total quantity if a doll, Ernie©, took a sweet from one of the parts,
if he added or removed a sweet to or from one of the parts, if he moved a sweet from
one of the parts to the other, or if he took away a sweet from one part but the inter-
viewer added a different sweet to the other part. In effect, they were asked for their
understanding of compensation and covariation of the whole with changes to one of
the parts. Children aged 4 were certain that the total quantity would stay the same
if an item was moved or replaced and would increase or decrease if one of the parts
was altered. At age 5 and 6 they could explain these relationships, sometimes with a
principle in their own language that showed that they understood this as a generality.
For example, one child said the total number in a compensatory move would be the
same, “The same, except Ernie put one of the lollies from here to here”. Expressed
algebraically, they understood that if P1 + P2 = W , then (P1 − k) + (P2 + k) = W .
Here, W stands for the total (the whole) uncounted number of sweets, P1 and P2 are
the generalized unknown (uncounted) number of sweets in the separate containers
(the parts comprising the whole), and k stands for the number of lollies transferred
from one container to the other.
However, if they were given a similar task with numbers only, using doubles
facts that they knew such as 5 + 5, and asked whether or not it would be the same
as 4 + 6, most were unsuccessful until age 7. As one child phrased it using a visual
image for the equality of 10 + 10 and 11 + 9, “. . .because if you put one of the
group of 11 over to the 9 group they would both be 10 and that means 20.” Since
young children understand this concept when no numbers are attached, it may be
that the complexity of learning to understand numbers distracts students from the
knowledge that they had in a proto-quantitative form (Resnick 1992) before going
to school.
The New Zealand Numeracy Project built on this understanding of young chil-
dren’s knowledge as well as the studies of Steffe et al. (1988) and Wright (1994).
A major intention of this Numeracy Project was that learners should develop rather
than lose their initial ability to understand the way in which numerical quantities
can be manipulated.
Algebraic Thinking and the New Zealand Numeracy Project

In 1999, the New Zealand Ministry of Education introduced a professional devel-
opment program in mathematics known as the Numeracy Development Project,
motivated by the need to improve students’ number sense and understanding of
operations by introducing a flexible approach to solving problems in numerical sit-
uations (see for example, McIntosh et al. 1992; Slavit 1999; Wright 1994). At first
the project was intended for Years 1–3 students (aged 5–7). In 2001, the project
was extended to Years 4–6 students (aged 8–10). In 2002, following a pilot study,
the project was expanded to include some 13,600 Year 7–10 students (aged 11–
14). By the end of 2008, approximately 95% of New Zealand elementary and
intermediate schools, 40% of secondary schools, and 85% of schools that teach
in Māori, had been involved in two years of numeracy professional development
(New Zealand Ministry of Education 2009b, http://www.nzmaths.co.nz/annual-
evaluation-reports-and-compendium-papers). Nearly all elementary schools were
using at least some aspects of the Numeracy Project and an increasing number of
secondary schools were becoming involved. The major aspects of this project are
now incorporated into the New Zealand Mathematics (New Zealand Ministry of Ed-
ucation 2007b, http://nzcurriculum.tki.org.nz/the_new_zealand_curriculum/learning
_areas/mathematics_and_statistics).
A basis of the Numeracy Project is the Number Framework (New Zealand Minis-
try of Education 2008, http://www.nzmaths.co.nz/sites/default/files/Numeracy/2008
numPDFs/NumBk1.pdf). In this, a distinction is made between strategy and knowl-
edge. The strategy section describes the thinking students use to mentally calculate
answers to numerical problems. The knowledge section describes the key items of
number knowledge that students need to learn and without which they will be unable
to broaden and advance their repertoire of strategies. The two are linked in that oper-
ational strategies create new knowledge through consistent use and knowledge pro-
vides a foundation for the development of new operational strategies. Tables 1 and 2
(see New Zealand Ministry of Education 2008, http://www.nzmaths.co.nz/sites/
default/files/Numeracy/2008numPDFs/NumBk1.pdf, pp. 15–17) show the progres-
sion of operational strategies that we argue forms a strong basis for ongoing oppor-
tunities for students to develop an awareness of generality and hence of algebraic
thinking.
The operational strategies illustrated in Table 2 involve part-whole thinking in
which students recognize that numbers are abstract units that can be partitioned and
then recombined in different ways to facilitate numerical calculation.
An essential part of the Numeracy Project is a diagnostic test administered in
whole or part as an interview with individual students. The intention of the inter-
view, which is used in different forms from ages 5 through 14, and includes ques-
tions of increasing complexity so that students are not asked to make responses
that are clearly beyond them, is to provide information about the knowledge and
mental strategies of the student (see, New Zealand Ministry of Education 2009a,
http://www.nzmaths.co.nz/gloss-forms). Teachers use the interview data to help de-
vise teaching programs to match the learning needs of students assessed as being at
similar stages across the three operational domains: addition and subtraction, mul-
tiplication and division, and proportions and ratio described in the Number Frame-
work (see Tables 1 and 2). Students are encouraged to work mentally as this provides
information that can be analysed more fully than answers from a paper and pencil
test.
A sample of additive, multiplicative and proportional tasks from Form H of the
interview is given below:
Task (9): On a hot day the tomato plants absorbed 1.5 litres of water. On a cold
day they absorbed 0.885 litres (885 mL). How much more water did the
plants absorb on the hot day than the cold day?
Table 1 The number framework for stages 1 to 4 operational strategies that involve counting
Operational domains
Global stage Addition and subtraction Multiplication and division Proportions and ratio
Stage Emergent
Emergent • Unable to count or form a given set of up to ten objects
Stage 1 One-to-one counting One-to-one counting Unequal sharing
One to one • Able to count a set of • Able to count a set of • Unable to divide a re-
counting objects objects gion or set into two or
• Unable to form a set • Unable to form a set of four equal parts
of objects to solve sim- objects to solve simple
ple addition and sub- multiplication and di-
traction problems vision problems
Stage 2 Counting from one Counting from one Equal sharing
Counting • Counts objects to solve • Solves simple multi- • Able to divide a re-
from one simple addition and plication and division gion or set into a given
on subtraction problems problems by counting number of equal parts
materials • Needs to use materials one-to-one with the aid using materials
such as counters or fin- of materials
gers
Stage 3 Counting from one Counting from one Equal sharing
Counting • Counts objects by vi- • Counts all the objects • Able to share a re-
from one sualizing or imaging in simple multiplica- gion or set into a given
by • Unaware of 10 as a tion and division prob- number of equal parts
imaging counting unit lems by imaging the using materials or by
• Solves multi-digit objects imaging
problems by counting • Uses materials to solve
all the objects multiplication and di-
vision problems with
larger numbers
Stage 4 Counting on Skip counting
Advanced • Counts on or back to • Skip counts to solve
counting solve simple addition simple multiplication
and subtraction prob- and division problems
lems using materials or
imaging
Task (10): Bas needs to buy 114 cans of soft drink for the volley ball club party?
How many 6-packs should he get?
Task (11): The dog ate three-eighths of an 800 gram can of jollymeat. The cat ate
three-quarters of a 400 gram can. Which ate more, the dog or the cat?
Since 1999, when it began as a modest experiment, the Numeracy Project has
been subjected to a number of evaluations (see New Zealand Ministry of Education
2009b, http://www.nzmaths.co.nz/annual-evaluation-reports-and-compendium-
papers). These evaluations have provided ongoing national data for the project
over an eight-year period. Evidence including measures of effect sizes related to
Table 2 The number framework for stages 5–8 operational strategies that involve part-whole
thinking
Operational domains
Global Addition and subtraction Multiplication and division Proportions and ratio
stage
Stage 5 Early addition and Multiplication by repeated Fraction of a number by

subtraction addition addition
Early • Uses a limited range of • Uses a combination of • Uses addition facts to
additive mental partitioning & known multiplication find the fraction of a
compensation strategies, facts and repeated ad- number. E.g., 13 of 12 is
to solve addition and sub- dition. E.g., 4 × 6 is 4 since 4 + 4 + 4 = 12
traction problems. E.g., (6 + 6) + (6 + 6) = • Solves division prob-
8 + 7 is 8 + 8 − 1 and 12 + 12 = 24 lems mentally using
39 + 26 = 40 + 25 = 65 • Uses known multiplica- halving or deriving
tion facts with repeated from known addition
addition, to anticipate facts. E.g., when 7 pies
the result of division. are shared among 4
E.g., 20 ÷ 4 = 5 since children each gets 1 pie
5 + 5 = 10 and 10 + plus 12 of a pie plus 14
10 = 20 of a pie
Stage 6 Advanced addition and Derived multiplication Fraction of a number by
subtraction of whole addition and
numbers multiplication
Advanced • Can estimate answers and • Uses a combination of • Uses repeated halving
additive - solve mentally addition known multiplication or known multiplication
early mul- and subtraction problems facts and mental strate- and division facts to
tiplicative that involve whole num- gies to derive answers solve problems that in-
bers by choosing appro- to multiplication and volve finding fractions
priately from a broad division problems. E.g., of a set or region, and
range of advanced men- 4 × 8 = 2 × 16 = 32 division with remain-
tal strategies. E.g., 324 − (doubling and halv- ders. E.g., 13 of 36 = 12
86 = 324 − 100 + 14 and ing) and 9 × 6 is since 3 × 10 = 30, 6 ÷
1242 − 986 = 1242 + (10 × 6) − 6 = 54 3 = 2, and 10 + 2 = 12
14 − (986 + 14)
Stage 7 Addition and subtraction of Advanced multiplication Fractions, ratios, and
decimals and integers and division proportions by
multiplication
Advanced • Can estimate answers and • Chooses appropriately • Uses a range of multi-
multi- solve mentally addition from a broad range of plication and division
plicative - and subtraction problems mental strategies to strategies to estimate
early pro- that involve decimals, in- estimate answers and answers and solve
portional tegers and related frac- solve multiplication problems with frac-
tions by choosing ap- and division prob- tions, proportions, and
propriately from a broad lems. E.g., 24 × 6 is ratios. E.g., 13 ÷ 5 =
range of advanced men- (20 × 6) + (4 × 6) or (10÷ 5)+ (3÷5) = 2 35 ;
tal strategies. E.g., 3.2 + 25 × 6 − 6; 81 ÷ 9 = 9 3 : 5 is equivalent to
1.95 = 3.2 + 2 − 0.05 = so 81 ÷ 3 = 3 × 9; 24:40 since 8 × 3 = 24
5.2 − 0.05 = 5.15 and 4 × 25 = 100, so and 8 × 5 = 40
92 ÷ 4 = 25 − 2 = 23
Table 2 (Continued)
Operational domains
Global Addition and subtraction Multiplication and division Proportions and ratio
stage
Stage 8 Addition and subtraction of Multiplication and Fractions, ratios and

fractions division of decimals, proportions by
multiplication of fractions re-unitising
Advanced • Uses a broad range • Chooses appropriately • Chooses appropriately
propor- of mental partitioning from a broad range of from a broad range
tional strategies to estimate mental strategies to esti- of mental strategies
answers and solve prob- mate answers and solve to estimate answers
lems that involve adding problems that involve and solve problems
and subtracting fractions the multiplication and that involve fractions
including decimals division of decimals proportions and ratios.
• Combines ratios and pro- and the multiplica- E.g., 6:9 is equivalent to
portions. E.g., 20 coun- tion of fractions. E.g., 16:24 since 6 × 1 12 = 9
ters in ratio of 2:3 com- 4.2 ÷ 0.25 = (4.2 × 4) ÷ and 16 × 1 12 = 24
bined with 60 counters in (0.25 × 4) = 16.8 ÷ 1 or 9 × 2 23 = 24 and
ratio of 8:7 gives a com-
6 × 2 23 = 16
bined ratio of 1:1
the average attainment in number of different ethnic groups and of issues related
to the validity of the framework and reliability related to the test administration
by classroom teachers have supported the continuing development of the project
(see Thomas et al. 2006; Young-Loveridge 2006), and have led to the establishment
of standards for students’ learning in number for each of the first four levels of a
new national curriculum for years 1 through 8 (New Zealand Ministry of Education
2007b).
An example of activities in a New Zealand classroom will help clarify the type of
tasks that students undertake. In this example we describe what a 7-year old student,
who we called Mary, might have done as she engaged with a series of tasks drawn
from the Numeracy Project material for students advancing from Stage 4 (Advanced
Counting) to Stage 5 (Early Additive) strategy activity.
Mary’s initial task was to use the tens-frames (see Fig. 1) to help devise a sensible
non-counting strategy to work out 9 + 4.
Mary’s teacher would have begun by asking her to put counters on a tens-frame
to show 9 and then a further 4 counters on another tens-frame to show 4 (Fig. 1a).
Figure 1b shows the outcome of Mary’s actions that transform 9 + 4 into 10 + 3 so
leading to her recognition of 13 as the answer.
Mary might then have gone on to use the tens-frames to figure out the solution to
several similar tasks, 8 + 5, 7 + 6, and 9 + 7 before being challenged by her teacher
to see if she could work out 19 + 4, 27 + 6, and 38 + 7 without recourse to the
tens-frames. She could revert to using the tens-frames if she was unsure what to do.
She might be asked by her teacher to explain her thinking after each response. For
19 + 4, she might have said that she took one from the 4 and put it on the 19. So in
her mind she could see 20 and 3 making a total of 23. She might similarly transform
Fig. 1 Using tens-frames to develop early understanding of additive compensation
27 + 6 into 30 + 3 = 33 and 38 + 7 into 40 + 5 = 45 each time describing an additive

compensation strategy for which the underlying structure or generalization may be
represented algebraically as a + b = (a + c) + (b − c).
There are several aspects associated with Mary’s thinking that warrant further
analysis. Firstly, we argue that the role assigned to Mary’s use of the tens-frames
is one of ‘Image-Making’, an early level of understanding drawn from Pirie and
Kieren’s Model for the Growth of Mathematical Understanding (Pirie and Kieren
1989, 1994). In summarizing their model, Pirie and Kieren (1989, p. 8) claim that,
“Mathematical understanding can be characterized as leveled but non-linear. It is a
recursive phenomenon and recursion is seen to occur when thinking moves between
levels of sophistication. Indeed each level of understanding is contained within suc-
ceeding levels. Any particular level is dependent on the forms and processes within
and, further, is constrained by those without.” Mary’s Image-making here is largely
constrained by her existing number knowledge. She could count to 99, she knew
the pairs of whole numbers that add to ten, and she could interpret 2-digit whole
numbers according to the place-value system for tens and ones. Without this number
knowledge, which in Pirie and Kieren’s model is referred to as the ‘Primitive Know-
ing’ level of understanding, it is unlikely that Mary would make much progress with
building the images to establish a basis for fully understanding addition of whole
numbers.
In the Numeracy Project, the recursive theory of mathematical understanding,
proposed by Pirie and Kieren (1989) provides the basis of a teaching model, to
be used alongside Project teaching materials (New Zealand Ministry of Education
2007a). In this Numeracy Project Teaching Model, students begin with a ‘Using
Materials’ phase designed to build concrete images that reflect their thinking and
in so doing reflect the Image-Making level proposed by Pirie and Kieren. As Mary
manipulates the counters on the tens-frames, she notices the space/s to be filled on
one frame to make it complete and also where she could get the counter/s from to
do the filling of the space/s. She is likely to have begun to develop an awareness
of the consequence of this compensation action (Pirie and Kieren’s Image Having
Level) and so visualize the structure of the transformation that she uses subsequently
to solve the more challenging tasks, 19 + 4, 27 + 6, and 38 + 7. In the Teaching
Model, these tasks are included within a ‘Using Imaging’ phase and are designed
to challenge Mary’s thinking towards an awareness of the transformations involved

in the tasks but in the absence of the tens-frames. Further, she is also likely to have
begun to isolate the features common to each task. This level of Pirie and Kieren
is called ‘Property Noticing’ and corresponds to the ‘Using Number Properties’
phase in the Project’s Teaching Model. Here at this level Mary demonstrates her
understanding by explaining the roles of the various elements in the transformation.
Such explanation, where she uses both numbers and words to express generality of
the transformation, indicates she had engaged in algebraic thinking. It is helpful to
note that, as with the Pirie and Kieren’s model, the Numeracy Project’s Teaching
Model incorporates opportunities for folding-back. For example, if Mary had not
succeeded with the tasks at the Using Imaging phase she would fold back to work
with more tasks in the Using Materials phase. She would then advance once again
to the Using Imaging phase when she had demonstrated success with these tasks.
Students’ Algebraic Thinking in the Last Year of Intermediate

School (Age 11–12)
In 2002, we carried out an evaluation of one aspect of the Numeracy Project that
is reported more fully in Irwin and Britt (2005a). The goal of the evaluation was to
gauge if the students who had participated in the Numeracy Project used operational
strategies deemed algebraic in nature more successfully than students from the same
year cohort who had not participated in the project. We wanted to test our conjec-
ture that project students would show a greater awareness of the algebraic structure
of problems in arithmetic. We devised a 21-item test comprising six sections: com-
pensation in addition, x + y = (x + a) + (y − a); compensation in subtraction,
x − y = (x + a) − (y + a); compensation in the distributive law of multiplication
over addition, xy = x(y + a) − xa; equivalence with sums and differences in which
one of four structures is, ‘If x + a = b, then x = b − a’; compensation in multipli-
cation, xy = (ax)( ya ); and equivalence with fractional values, again in which one of
four structures is, ‘If ab = an
x , then x = b × n or bn’. The following shows the test
items for Section A: Compensation in addition.
Example 1. 47 + 25 = 50 + 22 = 72
Example 2. 67 + 19 = 66 + 20 = 86
Use a strategy like those shown above to work out each of the following.
(1) 97 + 56
(2) 268 + 96
(3) 4613 + 987
The study involved 899 Year 8 students from four schools of which 431 partici-
pated in the Numeracy Project in 2002 and 468 had not participated in this Project.
Year 8 students, the last year of intermediate school, were chosen for the study be-
cause they had not been taught formal algebra prior to the study. The test was sub-
jected to a Rasch analysis with reliability 0.88 as estimated by Kuder-Richardson’s
formula 20. A subsequent analysis of the students’ results showed that for every test
item, the proportion of students who participated in the project and were deemed
to have completed the item successfully was greater than the proportion that was
attained by those who had not participated in the project. The results suggested
strongly that students’ involvement in the Numeracy Project was likely to lead, not
only to improved outcomes for the arithmetic involved, but to improved algebraic
thinking skills for current and future algebraic activity. These results are in accord
with the views of a number of researchers such as MacGregor and Stacey (1999) and
Orton and Orton (1994) who have argued that the understanding of and skill in using
arithmetical relations are a necessary pre-requisite for learning algebra. While we
do not disagree with this we feel that we are able to explain why this is likely to be
the case. In our view, it is not merely that students involved in such arithmetic activ-
ity develop greater skill in this area but rather it is because they have already begun
their algebraic development as a consequence of their participation in the Numeracy
Project where a considerable amount of their time had been given to working with
the strategies as described in the Project Number Framework (see Tables 1 and 2).
That work, as we have previously argued, demands the development of awareness
of generality representing algebraic thinking.
The Growth of Algebraic Thinking from Numbers to Symbols:

A Longitudinal Study
Following the study above, in which we assessed the difference between students
engaged in the Numeracy Project and those not in the Numeracy Project, we car-
ried out a Longitudinal Study, Britt and Irwin (2008), in which we examined the
growth in algebraic thinking across the important move for students from interme-
diate school, which is part of the elementary school system in New Zealand, to
secondary school. Our hope was that the algebraic thinking skills gained in interme-
diate school would not be lost when students proceeded to secondary school, where
traditionally algebra has meant ‘doing arithmetic with letters’, which Mason et al.
(2005, p. 309) contend, “has proved fruitless for countless generations”.
We reasoned that students, who had developed an awareness of the generality in
a range of numerical operational strategies through their participation in the Numer-
acy Project, would be able to extend their algebraic thinking to include the standard
alphanumeric symbols of algebra. Students who could generalize in this way are at
the ‘Representational Level of Generalizing’ (Rivera 2006). We argued that it was
likely they would be able to capitalize on these generalizing skills as they progressed
through secondary school.
A new test of algebraic thinking, devised specifically for this study, was trialled
and revised before it was given in each of three consecutive years to students aged
12, 13 and 14 at years 8, 9 and 10 respectively. The intention was to gather data
in each of the three years on students as they progressed from year 8 through 10.
The test explored similar properties of operations to those explored by Schifter et
Section A
Jason uses a simple method to work out problems like 27 + 15 and 34 + 19 in his
head.
Problem Jason’s calculation
27 + 15 30 + 12 = 42
34 + 19 33 + 20 = 53
1) Show how to use Jason’s method to work out 298 + 57
2) Show how to use Jason’s method to work out 35.7 + 9.8
3) Use Jason’s method to work out what goes in the space: 58 + n=60 + _______
4) Use Jason’s method to work out what goes in the space: 9.9 + k =10 + _______
5) Use Jason’s method to work out what goes in the space: a +b=(a +c)+______
Fig. 2 Section A: Additive compensation in the algebraic thinking test
al. (2008) with somewhat younger students. It comprised four sections, one for each
arithmetical operation. First, a solution was presented showing how a hypothetical
student had solved two problems using a compensation strategy for the operation.
The first item for each operation involved generalizing the demonstrated method
with 3-digit whole numbers, chosen to discourage students’ use of computational
approaches while at the same time encouraging the use of the illustrated compen-
sation approach. The second item required them to demonstrate the operation with
decimal fractions. The inclusion of decimal fraction items arose partly out of a previ-
ous study (Irwin and Britt 2004) in which we noted that decimal fraction questions
increased the level of complexity of the task. We also wanted to provide further
opportunity to gauge the effect of extending numerical generalizing beyond whole
numbers as is required when letters are used as generalized numbers. The third item
asked students to show how the operation would work when one of the numbers in
the particular item was represented by a letter as a generalized unknown, the fourth
item asked them the same question when the item involved letters and a decimal
fraction, and the final item asked them to identify missing letters in an algebraic
identity representing the compensation operation. Figure 2 shows the additive com-
pensation tasks of the algebraic thinking test.
Four intermediate schools and four secondary schools agreed to participate in
this study. The intermediate schools were all participants in the Numeracy Project.
The secondary schools were then chosen by finding out what secondary school the
majority of students from the intermediate school moved to. Intermediate and sec-
ondary schools were paired in this way. Of the four pairs of schools two pairs of
schools were on the outskirts of each of two large cities in New Zealand. Students
who came, on average, from middle-income families attended three of the pairs of
schools and students from higher income families attended one pair of schools. We
asked schools to give the same test to all students in year 8 at Intermediate school
in 2004, year 9 at secondary school in 2005, and year 10 in 2006. In addition, each
secondary school was asked to give the test to all of their year 9 and 10 students
in each of the three years. Our main focus was on the students who took the test
for each of the three years, thus giving us three data points for each of the students.
Irwin and Britt (2005b, 2006) reported on progress during the first two years of the
study. There were 116 students who participated in all three years of the Longitudi-
nal Study.
Teachers in all schools gave the algebraic thinking test to their students during
the final term of the year in question. The tests were then returned to the authors
for marking. The criterion for success was that an item was deemed correct only
if the compensation method was used correctly for that operation, disregarding an-
swers that were correct if students used a vertical algorithm but with no evidence of
generalization of compensation.
Three analyses were carried out to evaluate the data gathered from the tests of
algebraic thinking over three years. The first analysis explored the correlation of
individual student’s scores on the algebraic thinking test with the results of individ-
ual interviews (see the earlier section on algebraic thinking and the New Zealand
Numeracy Project and its evaluation) on the three strategy scores of the Numeracy
Project Diagnostic Interview at the end of year 9 (first year of secondary school).
The second analysis looked at the algebraic thinking test scores for students who
took the test in years 8, 9 and 10 in 2004, 2005 and 2006 respectively. The third
analysis was both quantitative and qualitative, and took a closer look at what was
happening in the one pair of schools where the mean scores on the algebraic think-
ing test were significantly superior to the other schools. This third analysis is not
discussed here but is discussed in Britt and Irwin (2008).
In 2005, the scores that year 9 students gained on the algebraic thinking test
and on the Numeracy Project Diagnostic Interview were correlated for the three
secondary schools in the Secondary Numeracy Project. These students had been
given the Numeracy Project diagnostic interview described earlier in the section
on ‘Algebraic thinking and the New Zealand Numeracy Project’. While scores for
strategy stages in additive, multiplicative, and proportional thinking are reported
individually in the diagnostic interview, we added the scores for the stages together,
reasoning that students needed to have good ability in all skills to be a competent
mathematics student in this age group.
The correlations between students’ scores for these strategy stages on the di-
agnostic interview, and on the algebraic thinking test was 0.47 with significance
< 0.01. Thus we were assured that individual interviews demonstrated that students
who were flexible in their thinking about numerical problems as fostered by the
Numeracy Project were more likely to be the same ones who could transfer from
using numbers flexibly to using letters to express the generalizations than those who
had not benefited from the Numeracy Project. The flexible thinkers we argued were
capable of algebraic thinking.
A further analysis of the diagnostic strategy scales shows that students who were
above the median on the Numeracy Project diagnostic test were those who demon-
strated algebraic thinking on our test. This means that students who could demon-
strate multiplicative and proportional reasoning as well as additive reasoning were
likely to transfer this flexible numerical thinking to algebraic items that involve stu-
dents expressing their generalizations algebraically.
Fig. 3 Each panel represents a randomly selected student (ID at top of panel). Waves 1, 2, and 3
refer to Years 8, 9, and 10 respectively. The data points show a student’s score at each wave, and
the straight lines are least-squares fits
Table 3 Correlation
coefficients among the Year 8 Year 9 Year 10
students’ scores over three
years Year 8 Pearson Correlation 1 .640* .639*
Year 9 Pearson Correlation .640* 1 .714*
Year 10 Pearson Correlation .639* .714* 1
* significant at the 0.01 level N 116 116 116
(2-tailed)
We were particularly interested in the growth of algebraic thinking across the

three years. This analysis was of the scores of the 116 students who took the alge-
braic thinking test on three occasions.
The students varied both in their initial attainment at the end of Year 8 and in
their rate of development over the three-year period. An indication of this variabil-
ity is shown in Fig. 3, which presents the results of a random selection of eight
students. The figure shows that some students improved over the three years, some
did not change, and some declined in performance. In order to take into account this
variability among students over each year, a random coefficient analysis was under-
taken. In this analysis, the coefficients for the intercept and slope of the regression
are allowed to differ among the students. (Figure 3 illustrates examples of the differ-
ent intercepts and slopes of the fitted line for each student.) Accommodating for this
variation is especially desirable in any Longitudinal Study because of the interde-
pendence among responses when the same person is measured on several occasions
(Twisk 2003). Table 3 shows the correlation coefficients of the scores over the three
years.
Figure 4 shows the average performance of all the students who attended a pair
of schools over three years. With the exception of School Pair 4, the mean scores
Fig. 4 Mean algebraic thinking test score of students attending each of the four pairs of schools
over three years
of students in each of the schools improved over the three years. In examining this
figure it should be borne in mind that the number of students from each pair of
schools differed widely: 13 students in school-pair labelled “1”, 14 students in the
school-pair “2”, 61 students in school-pair “3”, and 28 students in school-pair “4”.
The low figure for students who were tested on three separate occasions can be
accounted for in several ways. Some students at the intermediate schools chose to
move from their intermediate school to a secondary school other than the nom-
inated secondary school, some moved away from the school district, and others,
particularly those from schools from lower socio-economic homes, had poor school
attendance records.
Scores from intermediate school 3 were significantly different from intermediate
schools 2 and 4 (p < 0.05) but not from intermediate school 1 (p = 0.081) because
of the large variance in the small number of student scores at that school (t-test).
The difference between schools when all three years were taken into consideration
approached significance F (3, 112) = 2.620, (p = 0.054).
The random coefficients analysis of the data was undertaken by means of SPSS’s
mixed linear model. Two models were fitted. The simpler model estimated, without
differentiating among the schools, the average score at Year 8 and the average rate
of improvement over each subsequent year. The model’s estimated score at Year
8 was 5.7 and the estimated improvement per year was 1.3 points. The heavy line
labelled “all” in Fig. 5 shows this model.
Fig. 5 Two best-fitting models: (a) the heavy line, labelled “all”, shows the estimated score at Year
8 and the estimated improvement over three years of the average student, without differentiating
among the schools; and (b) a model of the estimated score at Year 8 and the estimated improvement
of the average student in each of the pairs of schools
The other model took into consideration the effect of attending different pairs
of schools. This model provided a marginally better fit to the data than the simpler
one. If this is taken to show a worthwhile improvement in the model’s fit, then the
resulting estimated initial score and improvement rate for students from each school
are shown in Fig. 5 by the lines labelled “1”, “2”, “3”, and “4”. For both models, the
rate of change (1.3 points per year) represented a significant improvement.
Discussion
In the study of students at the end of intermediate school (Irwin and Britt 2005a), we
showed that students who had participated in the New Zealand Numeracy Project
were more able than students who had not participated in the Numeracy Project to
demonstrate an awareness of the underlying algebraic structure of the operational
strategies they used to solve problems in arithmetic. We noted that this awareness
of structure amounted to an awareness of generality and argued that such students
were therefore engaging in algebraic thinking. We also claimed that, rather than us-
ing the literal symbols of algebra, these students who had no prior experience with
such symbols, were thinking of the numbers themselves as variables. We referred to
these as quasi-variables (Fujii and Stephens 2001) and subsequently argued that the
arithmetic that students used to solve problems was rooted in algebra. That is, with-
out algebra there could be no arithmetic (Hewitt 1998; Mason et al. 2005). In the
Longitudinal Study (Britt and Irwin 2008), we extended this relationship between
algebra and arithmetic by showing that students, who had developed algebraic think-
ing in multiplicative and proportional situations at intermediate school were capable
of algebraic thinking that involved representing numerical generalities with the spe-
cial symbols of algebra. We also showed that these algebraic thinking skills for such
students continually improved as they progressed through the first year of secondary
school. These analyses taken together suggest strong support for a positive relation-
ship between success in the Numeracy Project and subsequent algebraic thinking in
which students use the literal symbols of algebra to express algebraic generaliza-
tions.
The support for a link between the effect of the Numeracy Project and the growth
of algebraic thinking skills as students progress from Year 8 through Year 10 is yet
further strengthened by the analysis of the data shown in Fig. 4. While the numbers
of students contributing from different schools to the sample for the Longitudinal
Study differed widely, the two linear models (Fig. 5) that were fitted to the data
provided complementary evidence for an overall improvement of performance over
the three years. As a result of this analysis and the previous analyses of student per-
formance, we contend that we have demonstrated the existence of a strong ongoing
positive link between the effects of the Numeracy Project and the development of
algebraic thinking that involves expressing generality with the standard algebraic
symbols.
A Pathway for Algebraic Thinking
In New Zealand, it has been customary for students to begin formal algebra at sec-
ondary school. A typical algebra curriculum drawn from locally produced textbooks
for Year 9 began by focusing on functional thinking activities such as patterns in
growing match-stick designs, skill development with the syntax of operational al-
gebra, and solving simple linear equations. The teaching was typically instructional
rather than explorative with little or no attention given to generality. Students mostly
found algebra taught in this manner confusing and therefore difficult.
The New Zealand Numeracy Project has led to a new perspective of algebra
among many New Zealand teachers, particularly primary and intermediate-school
teachers. The developing awareness that the seeking of generality in any mathemat-
ical task is essentially algebraic in nature has offered primary teachers, many of
whom have a jaundiced view of algebra, a new emphasis in their teaching of math-
ematics. Careful consideration of the three studies detailed in this chapter together
suggests an alternative pathway for teaching and learning beginning algebra or al-
gebraic thinking. We argue that this pathway should provide opportunities for all
students to work with several layers of awareness of generality in all areas of their
mathematics curriculum prior to any formal introduction to algebra.
Such a pathway might firstly involve very young children engaging in activi-
ties of a proto-quantitative nature where counting is not needed and in which they
describe in their own words generalizations such as for example, those involved in
compensation thinking, “If P1 + P2 = W , then (P1 − k) + (P2 + k) = W ”. This is the
same emphasis offered by Dougherty (2008). Attention might then turn to students
developing an array of counting strategies before using those strategies and associ-
ated counting knowledge to help devise and broaden a range of operational strategies
in arithmetic where generality may be expressed through the use of quasi-variables.
Finally, students who have worked successfully with quasi-variables might be fur-
ther challenged to use the symbols of algebra to represent and work with mathemat-
ical generalizations drawn from a range of numerical and measurement situations as
well as figural representations (see Rivera 2006; Rivera and Becker 2008).
In New Zealand, first steps towards establishing a pathway for algebraic thinking
have been taken during a recent redevelopment of national curricula. The changes
in the New Zealand mathematics curriculum for students aged 5 to 13, where the
formerly separate Number and Algebra strands have been combined to form a sin-
gle Number and Algebra strand in the newly designated Mathematics and Statis-
tics curriculum (New Zealand Ministry of Education 2007b) have arisen as a result
of the overall success of the Numeracy Project initiative (see New Zealand Min-
istry of Education 2009b, http://www.nzmaths.co.nz/annual-evaluation-reports-and-
compendium-papers) and of the research findings related to algebraic thinking that
have been discussed earlier in this chapter.
As a consequence of this reconsideration of the role of algebra in the New
Zealand mathematics curriculum, a number of classroom activities that encourage
algebraic thinking have been devised and trialled in classroom settings. Figure 6
shows an example of a sequence of such tasks in which students, assessed at Stage
6 (See Table 2—Advanced Additive), are asked to look for and express generality
in words, with quasi-variables where numbers act as if they are variables, and with
the literal symbols of algebra.
Students working on these tasks are asked to use three semiotic systems to ex-
press layers of generality: firstly with numbers as quasi-variables, then with words
and finally with the literal symbols of algebra. For example, the design in Fig. A2
might be expressed numerically as 2 lots of 47 black circles plus 2 lots of 45 white
circles for the coaster with 47 circles on each edge and in words as, ‘The total num-
ber of circles is two lots of the number of circles on an edge (black circles) side
plus two lots of two fewer than the number of circles on a side (white circles). The
algebraic expression is then 2 × x + 2 × (x − 2) which is more usually written as
2x + 2(x − 2). The algebraic expressions for the designs in Fig. A3 and Fig. A4 are
4 + 4(x − 2) and 4(x − 1) respectively.
As a follow-up to the tasks in Coaster Design, students themselves figure out
how to reduce the different algebraic expressions to the same form, likely here to be
4x − 4, interpreted as, ‘four times the number of circles on an edge less one circle
that has already been counted twice for each corner’. Students might also be en-
couraged to devise algebraic expressions to represent a range of challenging figural
representations. For example, they might look for generality in coaster designs sim-
ilar to the square-shaped designs in Fig. 6 but with circles along the edges of regular
Fig. 6 Expressing generality with and without the literal symbols of algebra
triangles, pentagons, hexagons, and even n-gons. Rivera (2006) notes that students
who identify generality in figural representation and are able to express that gener-
ality algebraically may be classified as predominantly Figural Generalizers working
at the Representational Level of Generalizing.
While figural representations such as those shown in Fig. 6 can offer a rich source
of challenging tasks requiring students to seek and express generality, the tasks in
Fig. 7 also provide similar opportunities. These tasks do not need students to be
figural generalizers as proposed by Rivera (2006), but they nevertheless provide op-
portunity to learn the structure and syntax of algebraic generality where letters stand
for generalized unknown numbers. The tasks in Fig. 7 comprise separate learning
progressions that deepen awareness of generality while at the same time assist stu-
dents to hone their mental skills in working with compensation strategies in addition
and subtraction.
Each item in the separate progression illustrates the structure of a group of similar
items. The groups of items so formed are sequenced in a manner to draw attention
to the generality inherent in the particular strategy. Students express such generality
with quasi-variables and with algebraic symbols. Task B1 focuses on additive com-
pensation operations. In the first item, 7 + = 9 + , any number can go in the first
box but a number two less than the first chosen number must go in the second box
Fig. 7 Generalizing additive and subtractive compensation strategies using quasi-variables and
algebraic symbols
to preserve the equivalence. Task B2 involves compensation in subtraction. Thus in

55 − 34 = 56 − , 55 has been increased by 1 so the number in the box must also
increased by one in order to preserve the difference represented by 55 − 34. Alter-
natively, in 20 − ( ) = 15 − m, 20 has been decreased by 5. So, in order to maintain
the difference between the pairs of values on either side of the equation the number
missing between the parentheses must be 5 more than m, that is m + 5.
However, in order for all students to achieve greatest benefit from working with
activities that encourage algebraic thinking at all levels teachers themselves will
need to also develop their own algebraic thinking skills. This could be achieved
through the adoption in teacher development programs of an “algebrafication” strat-
egy, similar to that proposed by Kaput and Blanton (2001). In such programs, teach-
ers might continually seek out generality in all aspects of mathematics and find
helpful ways, including using the symbols of algebra, to express generality. They
might also be encouraged to devise and apply wherever possible, a range of sensible
flexible mental operational strategies to carry out everyday tasks in arithmetic that
embody algebraic thinking.
New Zealand teachers and their students have had a growing involvement in the
Numeracy Project since 1999. The initial experimental nature of the project together
with its strong evaluative component and accompanying teacher professional devel-
opment, has led to improvements in skill level and confidence in teaching numeracy.
Increasingly, teachers are developing awareness that their work entails considerably
more than helping students get answers to problems involving numerical situations.
They are supported by, the Numeracy Project resource material and a curriculum
that encourages the concept of algebra within arithmetic. This national “experiment”
has led to professional development for most elementary school teachers and for an
increasing number of secondary school teachers. The challenge now is to enable all
teachers to see the benefit of an approach that includes seeing algebra within arith-
metic and the development of school teaching programs in mathematics that reflect
the notion of seeking and expressing generality that forms the basis of the Pathway
for algebraic thinking described in this chapter.
References
Britt, M. S., & Irwin, K. C. (2008). Algebraic thinking with and without algebraic representation: A
pathway for algebraic thinking. ZDM—The International Journal on Mathematics Education,
40, 39–53.
Brizuela, B. M., & Earnest, D. (2008). Multiple notation systems and algebraic understandings:
The case of the “best deal” problem. In J. J. Kaput, D. W. Carraher, & M. L. Blanton (Eds.), Al-
gebra in the Early Grades (pp. 273–301). New York: Lawrence Erlbaum Associates & National
Carraher, D. W., Schliemann, A. D., Brizuela, B. M., & Earnest, D. (2006). Arithmetic and algebra
in early mathematics education. Journal for Research in Mathematics Education, 37(2), 87–
115.
Carraher, D. W., Schliemann, A. D., & Schwartz, J. L. (2008). Early algebra is not the same as al-
gebra early. In J. J. Kaput, D. W. Carraher, & M. L. Blanton (Eds.), Algebra in the Early Grades
(pp. 235–272). New York: Lawrence Erlbaum Associates & National Council of Teachers of
Mathematics.
Dougherty, B. (2008). Measure up: A quantitative view of early algebra. In J. J. Kaput, D. W.
Carraher, & M. L. Blanton (Eds.), Algebra in the Early Grades (pp. 389–412). New York:
Lawrence Erlbaum Associates & National Council of Teachers of Mathematics.
Filloy, E., & Rojano, T. (1989). Solving equations: The transition from arithmetic to algebra. For
the Learning of Mathematics, 9(2), 19–25.
Fujii, T., & Stephens, M. (2001). Fostering an understanding of algebraic generalizations through
numerical expressions. In Proceedings of the 12th Conference of the International Commission
on Mathematical Instruction: The Future of the Teaching and Learning of Algebra (Vol. 1, pp.
258–264). Melbourne: The University of Melbourne.
Herscovics, N., & Linchevski, L. (1994). A cognitive gap between arithmetic and algebra. Educa-
tional Studies in Mathematics, 27, 59–78.
Hewitt, D. (1998). Approaching arithmetic algebraically. Mathematics Teaching, 163, 19–29.
Irwin, K. C. (1996). Children’s understanding of the principles of compensation and covariation in
part-whole relationships. Journal for Research in Mathematics Education, 27, 25–40.
Irwin, K. C., & Britt, M. S. (2004). Operating with decimals as a part-whole concept. In Proceed-
ings of the 27th Annual Conference of the Mathematics Education Group of Australasia (Vol. 2,
pp. 312–319). Townsville/Sydney: MERGA.
Irwin, K. C., & Britt, M. S. (2005a). The algebraic nature of students’ numerical manipulation in
the New Zealand Numeracy Project. Educational Studies in Mathematics, 58, 169–188.
Irwin, K. C., & Britt, M. S. (2005b). Algebraic thinking in the Numeracy Project: Year one of
a three-year study. In J. Higgins, K. C. Irwin, G. Thomas, T. Trinick, & J. Young-Loveridge
(Eds.), Findings from the New Zealand Numeracy Development Projects 2004 (pp. 47–55).
Wellington: Learning Media.
Irwin, K. C., & Britt, M. S. (2006). Algebraic thinking in the Numeracy Project: Year two of a
three-year study. In F. Ell, J. Higgins, K. C. Irwin, G. Thomas, T. Trinick, & J. Young-Loveridge
(Eds.), Findings from the New Zealand Numeracy Development Projects 2005 (pp. 46–54).
Wellington: Learning Media.
Kaput, J., & Blanton, M. (2001). Algebrafying the elementary mathematics experience. Part 1:
Transforming task structures. In H. Chick, K. Stacey, J. Vincent, & J. Vincent (Eds.), Pro-
ceedings of the 12th Conference of the International Commission on Mathematical Instruction:
The Future of the Teaching and Learning of Algebra (Vol. 1, pp. 344–351). Melbourne: The
University of Melbourne.
Kaput, J. J., Carraher, D. W., & Blanton, M. L. (Eds.) (2008). Algebra in the Early Grades. New
York: Lawrence Erlbaum Associates & National Council of Teachers of Mathematics.
MacGregor, M., & Stacey, K. (1999). A flying start to algebra. Teaching Children Mathematics,
6(2), 78–85.
McIntosh, A., Reys, B. J., & Reys, R. E. (1992). A proposed framework for examining basic
number sense. For the Learning of Mathematics, 12(3), 2–8.
Mason, J. (2008). Making use of children’s power to produce algebraic thinking. In J. J. Kaput,
D. Carraher, & M. L. Blanton (Eds.), Algebra in the Early Grades (pp. 57–94). New York:
Lawrence Erlbaum Associates & National Council of Teachers of Mathematics.
Mason, J., Graham, A., & Johnston-Wilder, S. (2005). Developing Thinking in Algebra. London:
The Open University.
New Zealand Ministry of Education (2007a). Numeracy Professional development projects.
http://www.nzmaths.co.nz/teaching-numeracy. Retrieved 15 February 2007.
New Zealand Ministry of Education (2007b). The New Zealand curriculum: Mathemat-
ics and statistics. http://nzcurriculum.tki.org.nz/the_new_zealand_curriculum/learning_areas/
mathematics_and_statistics.
New Zealand Ministry of Education (2008). Book 1. The number framework, revised edition
2008. http://www.nzmaths.co.nz/sites/default/files/Numeracy/2008numPDFs/NumBk1.pdf.
Retrieved 6 January 2009,.
New Zealand Ministry of Education (2009a). Global strategy forms (GloSS). http://www.nzmaths.
co.nz/gloss-forms. Retrieved 6 January 2009.
New Zealand Ministry of Education (2009b). New Zealand Numeracy Project annual eval-
uation and compendium reports 2004–2008. http://www.nzmaths.co.nz/annual-evaluation
-reports-and-compendium-papers. Retrieved 4 October 2009.
Orton, A., & Orton, J. (1994). Students’ perception and use of pattern and generalization. In Pro-
ceedings of the 18th Annual Conference for the Psychology of Mathematics Education (Vol. 4,
pp. 407–114). Lisbon, Portugal.
Pirie, S., & Kieren, T. (1989). A recursive theory of mathematical understanding. For the Learning
of Mathematics, 13(1), 7–11.
Pirie, S. E. B., & Kieren, T. E. (1994). Beyond metaphor: Formalising in mathematical understand-
ing within constructivist environments. For the Learning of Mathematics, 14(1), 39–43.
Radford, L. (2006). Algebraic thinking and the generalization of pattern: a semiotic perspective.
In Proceedings of the 28th Annual Meeting of the North American Chapter of the International
Group for the Psychology of Mathematics Education (Vol. 1, pp. 2–21). Mérida, Mexico: Uni-
versidad Pedagógica Nacional.
Resnick, L. B. (1992). From proto-quantitative to operators: Building mathematical competence
on a foundation of everyday knowledge. In L. G. Leinhardt, R. Putnam, & R. A. Hatrup (Eds.),
Analysis of Arithmetic for Mathematics Teaching (pp. 373–430). Hillsdale NJ: Erlbaum.
Rivera, F. (2006). Structures of elementary algebraic thinking: Evidence from sixth grade students
and implications for classroom practice. Intersection, 7(2), 2–13.
Rivera, F. D., & Becker, J. R. (2008). Middle schools children’s cognitive perceptions of construc-
tive and deconstructive generalizations involving linear figural patterns. ZDM—The Interna-
tional Journal on Mathematics Education, 40, 65–82.
Schifter, D., Monk, S., Russell, S. J., & Bastable, V. (2008). Early algebra: What does understand-
ing the laws of arithmetic mean in the Elementary grades. In J. J. Kaput, D. W. Carraher, & M.
L. Blanton (Eds.), Algebra in the Early Grades (pp. 413–447). New York: Lawrence Erlbaum
Associates & National Council of Teachers of Mathematics.
Slavit, D. (1999). The role of operation sense in transitions from arithmetic to algebraic thought.
Steffe, L. P. (2001). What is algebraic about children’s numerical operating. In Proceedings of the
12th Conference of the International Commission on Mathematical Instruction: The Future of
the Teaching and Learning of Algebra (Vol. 2, pp. 556–563). Melbourne: The University of
Melbourne.
Steffe, L. P., Cobb, P. & von Glasersfeld, E. (1988). Construction of Arithmetical Meanings and
Strategies. New York: Springer.
Thomas, G., Tagg, A., & Ward, J. (2006). Numeracy assessment: How reliable are teachers’ judg-
ments. In F. Ell, J. Higgins, K. C. Irwin, G. Thomas, T. Trinick, & J. Young-Loveridge (Eds.),
Findings from the New Zealand Numeracy Development Projects 2005 (pp. 91–102). Welling-
ton: Learning Media.
Twisk, J. W. R. (2003). Applied Longitudinal Data Analysis for Epidemiology: A Practical Guide.
Cambridge, UK: Cambridge University Press.
Young-Loveridge, J. (2006). Patterns of performance and progress on the Numeracy Development
Project: Looking back from 2005. In F. Ell, J. Higgins, K. C. Irwin, G. Thomas, T. Trinick, &
J. Young-Loveridge (Eds.), Findings from the New Zealand Numeracy Development Projects
2005 (pp. 6–21). Wellington: Learning Media.
Wright, R. (1994). Mathematics in the lower primary years: A research-based perspective on cur-
ricula and teaching practice. Mathematics Education Research Journal, 6, 23–36.
Examining Students’ Algebraic Thinking
in a Curricular Context: A Longitudinal Study
Jinfa Cai, John C. Moyer, Ning Wang, and Bikai Nie
Abstract This chapter highlights findings from the LieCal Project, a longitudinal
project in which we investigated the effects of a Standards-based middle school
mathematics curriculum (CMP) on students’ algebraic development and compared
them to the effects of other middle school mathematics curricula (non-CMP). We
found that the CMP curriculum takes a functional approach to the teaching of alge-
bra while non-CMP curricula take a structural approach. The teachers who used the
CMP curriculum emphasized conceptual understanding more than did those who
used the non-CMP curricula. On the other hand, the teachers who used non-CMP
curricula emphasized procedural knowledge more than did those who used the CMP
curriculum. When we examined the development of students’ algebraic thinking re-
lated to representing situations, equation solving, and making generalizations, we
found that CMP students had a significantly higher growth rate on representing-
situations tasks than did non-CMP students, but both CMP and non-CMP students
The research reported in this chapter is part of a large project, Longitudinal Investigation of the
Effect of Curriculum on Algebra Learning (LieCal Project). The LieCal Project is supported by a
grant from the National Science Foundation (ESI-0454739). Any opinions expressed herein are
those of the authors and do not necessarily represent the views of the National Science
Foundation.
J. Cai () · B. Nie
B. Nie
J.C. Moyer
Department of Mathematics, Statistics and Computer Science, Marquette University, Milwaukee,
USA
N. Wang
Center for Education, Widener University, Chester, USA

162 J. Cai et al.
had an almost identical growth in their ability to solve equations. We also found
that CMP students demonstrated greater generalization abilities than did non-CMP
students over the three middle school years.
Algebra readiness has been characterized as the most important “gatekeeper” in

school mathematics (Cai and Knuth 2005). Given its gatekeeper role as well as
growing concerns about students’ inadequate preparation in algebra, algebra curric-
ula and instruction have become focal points of mathematics education research
(e.g., Bednarz et al. 1996; Carpenter et al. 2003; Katz 2007). In particular, re-
searchers have tried to understand the nature of students’ algebraic thinking as well
as the role that different types of curricula play in its development (Katz 2007;
Kieran 2007; National Research Council 2004). The purpose of this chapter is to
compare how two different types of curricula affect the development of students’
algebraic thinking over the three middle school years (grades 6–8).
Reviews of research on the teaching and learning of algebra clearly show the
need to study the development of students’ algebraic thinking over time (Jones et
al. 2002; Kieran 2007). NCTM (2000) suggests that algebraic thinking should be
developed across all grade levels, but we know very little about how we should
characterize the growth of students’ algebraic thinking. For example, in their review,
Jones et al. (2002) explicitly called for future research to build a general model that
characterizes the growth of students’ algebraic thinking over time.
Although research in cognition has shown that students’ experiences out of
school have a substantial effect on their learning and problem solving (e.g., Lave
1988; Resnick 1987), classroom instruction is still considered a central component
in the development and the organization of students’ thinking and learning (Cai
2004; Detterman 1993; Rogoff and Chavajay 1995; Wozniak and Fischer 1993).
Bruner (1990/1998) proposed that it is culture and education, not biology, that shape
human life and the human mind. Similarly, Gardner (1991) argued that once a child
reaches age six or seven, the influence of culture and schooling “. . .has become
so pervasive that one has difficulty envisioning what development could be like in
the absence of such cultural supports and constraints” (p. 195). Because classroom
instruction plays such a central role in the development of students’ thinking, we
situated our investigation of the development of middle school students’ algebraic
thinking within a context of comparative instruction using two different types of
curricula—Standards-based and traditional. By doing so, we have been able to ex-
amine how the use of different types of curricula affect the development of students’
algebraic thinking over the middle school years.
Standards-Based and Traditional Curricula in the United States
In the late 1980s and early 1990s, the National Council of Teachers of Mathe-
matics (NCTM) published its first round of Standards documents, which provided
recommendations for reforming and improving K-12 school mathematics. In the
Examining Students’ Algebraic Thinking in a Curricular Context: A Longitudinal Study 163
Standards and related documents, the discussions of goals for mathematics educa-
tion emphasize the importance of thinking, understanding, reasoning, and problem
solving, with an emphasis on connections, applications, and communication (e.g.,
NCTM 1989, 1991; National Research Council 1989). This view stands in con-
trast to the more conventional view of mathematics education, which involves the
memorization and recitation of decontextualized facts, rules, and procedures, with
an emphasis on the application of well-rehearsed procedures to solve routine prob-
lems.
With extensive support from the U.S. National Science Foundation, a num-
ber of school mathematics curricula aimed to align with the recommendations in
the Standards were developed and field-tested in the United States (see Senk and
Thompson 2003 for details of these NSF-funded curricula). The Connected Math-
ematics Program (CMP) is one of the so-called Standards-based school mathemat-
ics curricula developed with the support of the National Science Foundation. The
CMP curriculum is a complete middle-school mathematics program that was iden-
tified as an exemplar by the U.S. Department of Education. The stated intent of
CMP is to build students’ understanding of major ideas in number, algebra, ge-
ometry, measurement, data analysis, and probability through explorations of real-
world situations and problems (Lappan et al. 2002). Because NSF-funded curricula
like CMP not only look very different from commercially developed mathemat-
ics curricula, but also claim to have different learning goals, they provide an in-
teresting context for examining the impact of curriculum on the development of
students’ algebraic thinking. By situating our examination of the development of
algebraic thinking in a curricular context, we can investigate the role that curricu-
lum plays in students’ learning of mathematics in general and in their acquisition
of algebraic concepts in particular (NCTM 1989; National Research Council 2004;
RAND Mathematics Study Panel 2003; Senk and Thompson 2003; Usiskin 1999).
LieCal Project
This chapter draws on findings from our NSF-funded LieCal Project (Longitudinal
Investigation of the Effect of Curriculum on Algebra Learning). The LieCal Project
was designed to longitudinally compare the effects of the Connected Mathematics
Program (CMP) to the effects of more traditional middle school curricula (hereafter
called non-CMP curricula) on students’ learning of algebra. The project was con-
ducted in 14 middle schools in an urban school district serving a diverse student
population. When the project began, 27 of the 51 middle schools in the district had
adopted the CMP curriculum, and the remaining 24 middle schools had adopted
more traditional curricula. Seven CMP schools were randomly selected from the 27
schools that had adopted the CMP curriculum. After the seven CMP schools were
selected, seven non-CMP schools were chosen based on comparable demograph-
ics. According to district ratings, the CMP and non-CMP groups of schools each
comprised two schools that were above average in overall educational performance,
three schools that were average, and two schools that were below average. About
164 J. Cai et al.
700 CMP students in 25 classes and 600 non-CMP students in 22 classes partici-
pated in the study when they were 6th graders. We followed these 1,300 students as
they progressed from grade 6 to 8. Approximately 85% of the participants were mi-
nority students: 64% African American, 16% Hispanic, 4% Asian, and 1% Native
American. Male and female students were about evenly distributed.
By comparing longitudinally the effects of the CMP curriculum to the effects
of other middle-school mathematics curricula on students’ learning of algebra, the
LieCal Project has been able to provide: (1) A profile of the intended treatment of
algebra in the CMP curriculum with a contrasting profile of the intended treatment
of algebra in the non-CMP curricula; (2) a profile of classroom experiences that
CMP students and teachers have, with a contrasting profile of experiences in non-
CMP classrooms; and (3) a profile of student performance resulting from the use
of the CMP curriculum, with a contrasting profile of student performance resulting
from the use of non-CMP curricula. Accordingly, the project was designed to an-
swer three sets of research questions: (1) What are the similarities and differences
between the intended treatment of algebra in the CMP curriculum and in the non-
CMP curricula? (2) What are key features of the CMP and non-CMP experience
for students and teachers, and how might these features explain performance dif-
ferences between CMP and non-CMP students? and (3) What are the similarities
and differences in performance between CMP students and a comparable group of
non-CMP students on tasks measuring a broad spectrum of mathematical thinking
and reasoning skills, with a focus on algebra?
In the next two sections, we first highlight the differences between the CMP and
non-CMP curricula and then the differences between classroom instruction using
CMP and non-CMP curricula. The goal of these two sections is to provide a cur-
ricular context in which to examine and understand the development of students’
algebraic thinking. In the following two sections, we describe the methods we used
to investigate the development of students’ algebraic thinking and then present our
findings related to the development of students’ algebraic thinking.

Curricula
We conducted detailed analyses of CMP and one of the non-CMP curricula,1 Glen-
coe Mathematics: Concepts and Applications (Bailey et al. 2006a, 2006b, 2006c),
and found significant differences between them (Cai et al. 2010c; Nie et al. 2009).
Overall, our research revealed that the CMP curriculum takes a functional approach
to the teaching of algebra, and the non-CMP curriculum takes a structural approach.
The functional approach emphasizes the important ideas of change and variation
1 In
this section, we only discuss the differences between CMP and one of the non-CMP curricula
(Glencoe Mathematics). Although there are differences among the non-CMP curricula, the differ-
ences between the CMP and each of the non-CMP curricula are similar.
in situations and contexts. It also emphasizes the representation of relationships

between variables. The structural approach, on the other hand, avoids contextual
problems in order to concentrate on developing the abilities to generalize, work ab-
stractly with symbols, and follow procedures in a systematic way (Cai et al. 2010c).
In this section, we highlight specific differences in the ways that the CMP curricu-
lum and the non-CMP curriculum (1) define variables, (2) define equation solving,
(3) introduce equation solving, and (4) use mathematical problems to develop al-
gebraic thinking. We focused on these four aspects in this chapter because they are
fundamental to algebra learning.
Defining Variables
Because of the importance of variables in algebra, and in order to appreciate the

differences between the CMP and non-CMP curricula, it is necessary to understand
how the CMP and non-CMP curricula introduce variable ideas. The learning goals
of the CMP curriculum characterize variables as quantities used to represent rela-
tionships. In contrast, learning goals in the non-CMP curriculum characterize vari-
ables as placeholders or unknowns. The CMP curriculum does not formally define
variable until 7th grade. However, the CMP’s definition of variable as a quantity
rather than a symbol makes it convenient to use variables informally to describe
relationships long before formally introducing the concept of variable in 7th grade.
Once CMP defines variables as quantities that change or vary, it uses them to repre-
sent relationships. The non-CMP curriculum formally defines a variable in 6th grade
as a symbol (or letter) used to represent a number. It treats variables predominantly
as placeholders by using them to represent unknowns in expressions and equations.
Defining Equations
With its emphasis on relationships, CMP clearly approaches the concept of variable
functionally. On the other hand, the non-CMP curriculum’s focus on variable as a
symbol points toward their structural approach. It is not surprising, therefore, that
the concept of equation is defined functionally in CMP, but structurally in the non-
CMP curriculum.
In CMP, the functional approach to equation is a natural extension of its de-
velopment of the concept of variable as a changeable quantity used to represent
relationships. At first, CMP expresses relationships between variables with graphs
and tables of real-world quantities rather than with algebraic equations. Later, when
CMP introduces equations, the emphasis is on using them to describe real-world
situations. Rather than seeing equations simply as objects to manipulate, students
learn that equations often describe relationships between varying quantities that
arise from meaningful, contextualized situations (Bednarz et al. 1996).
166 J. Cai et al.
In the non-CMP curriculum, the definition of variable as a symbol develops nat-

urally into to the use of “naked” equations and puts an emphasis on procedures for
solving equations. These are all hallmarks of a structural focus. For example, non-
CMP curriculum defines an equation as “. . .a sentence that contains an equals sign,
=” through examples like 2 + x = 9, 4 = k − 6, and 5 − m = 4. Students are told
that the way to solve an equation is to replace the variable with a value that results
in a true sentence.
Introducing Equation Solving
CMP and the non-CMP curriculum use functional and structural approaches, re-
spectively, to introduce equation solving. These are consistent with the approaches
they use to define equations. In the CMP curriculum, equation solving is introduced
within the context of discussing linear relationships. The initial treatment of equa-
tion solving does not involve symbolic manipulation as found in most conventional
curricula. Instead, the CMP curriculum introduces students to linear equation solv-
ing by making visual sense of what it means to find a solution using a graph. Its
premise is that a linear equation in one variable is, in essence, a specific instance
of a corresponding linear relationship (equation) in two variables. It relies heavily
on the context in which the equation itself is situated and on the use of a graphing
calculator.
After CMP introduces equation solving graphically, the symbolic method of solv-
ing linear equations is finally broached. It is introduced within a single contextual-
ized example, where each of the steps in the equation-solving process is accompa-
nied by a narrative that demonstrates the connection between what is happening in
the procedure and in the real-life situation. In this way, CMP justifies the equation-
solving manipulations through contextual sense-making of the symbolic method.
That is, CMP uses real-life contexts to help students understand the meaning of each
step of the symbolic method of equation solving, including why inverse operations
are used, as shown in Table 1.
In the non-CMP curriculum, contextual sense-making is not used to justify the
equation-solving steps as it is in the CMP curriculum. Rather, the non-CMP curricu-
lum first introduces equation solving as the process of finding a number to make an
equation a true statement. Specifically, solving an equation is described as replacing
a variable with a value (called the solution) that makes the sentence true. Equation
solving is introduced in the non-CMP curriculum symbolically by using the additive
property of equality (equality is maintained if the same quantity is added to or sub-
tracted from both sides of an equation) and the multiplicative property of equality
(equality is maintained if the same non-zero quantity is multiplied by or divided into
both sides of an equation).
In 6th grade, the non-CMP curriculum (Bailey et al. 2006a) formally introduces
equation solving with inverse operations by way of an activity that uses a cup to
stand for an unknown. The appropriate number of cups and counters used as ma-
nipulatives in the activity are initially positioned to exactly represent the equation’s
Table 1 An example of equation solving in CMP (Lappan et al. 2002b, p. 55)

The Unlimited Store allows any customer who buys merchandise costing over $30 to pay on the
installment plan. The customer pays $30 down and then pays $15 a month until the item is paid for.
Suppose you buy a $195 CD-ROM drive from the Unlimited Store on an installment plan, How
many months will it take you to pay for the drive? Describe how you found your answer.
Thinking Manipulating the Symbol

“I want to buy a CD-ROM drive that costs $195. To pay for the 195 = 30 + 15N
drive on the installment plan, I must pay $30 down and $15 a
month.”
“After I pay the $30 down payment, I can subtract this from the 195 − 30 = 30 − 30 + 15N
cost. To keep the sides of the equation equal, I must subtract 30
from both sides”
“I now owe $165, which I will pay in monthly installments of 165 = 15N
$15.”
“I need to separate $165 into payments of $15. This means I 165
15 = 15N
15
need to divide it by 15. To keep the sides of the equation equal,
I must divide both sides by 15.”
“There are 11 groups of $15 in $165, so it will take 11 months.” 11 = N
Fig. 1 Introduction of equation solving in the non-CMP curriculum
symbols. They are then used to illustrate each step of the symbolic manipulations
(see Fig. 1).
Using manipulatives as described above is referred to as “Method 1” and is typ-
ically shown adjacent to an example illustrating the corresponding solution using
the strictly symbolic “Method 2.” In this way, the non-CMP curriculum illustrates
168 J. Cai et al.
Fig. 2 Symbolic
representation of solving an
equation in the non-CMP
curriculum
Table 2 Percentages of various tasks in CMP and non-CMP curricula

Memorization Procedures Procedures Doing
without with Mathematics
Connections Connections
CMP (n = 920) 0.43 27.93 61.52 10.11

Non-CMP (n = 2391) 4.6 74.57 18.24 2.59
how each manipulative step is comparable to a symbolic step in a solution based

on the algebraic properties of equality, which is shown through vertical work. The
following is an example of Method 2 showing how to solve a one-step equation (see
Fig. 2).
Using Mathematical Problems
The extent of the differences between the CMP and non-CMP curricula can also
be highlighted through an analysis of mathematical problems. Using a scheme de-
veloped by Stein et al. (1996), we classified the mathematical tasks in the CMP
curriculum and non-CMP curriculum (Bailey et al. 2006a, 2006b, 2006c) into four
increasingly demanding categories of cognition: memorization, procedures with-
out connections, procedures with connections, and doing mathematics. As Table 2
shows, significantly more tasks in the CMP curriculum than in the non-CMP cur-
riculum are higher-level tasks (procedures with connections and doing mathematics)
(χ 2 (3, N = 3311) = 759.52, p < .0001).
We further analyzed the problems in the CMP and non-CMP curricula that in-
volve linear equations by classifying them into three categories:
(1) One equation with one variable (1equ1va)—e.g., 2x + 3 = 5;
(2) One equation with two variables (1equ2va)—e.g., y = 6x + 7;
(3) Two equations with two variables (2equ2va)—e.g., the system of equations y =
2x + 1 and y = 8x + 9.
Figure 3 shows the percentage distribution of the problems involving linear
equations in the two curricula. These two distributions are significantly different
Fig. 3 Percentage distribution of problems involving linear equations in CMP curriculum and the
non-CMP curriculum
(χ 2 (2, N = 2741) = 1262.0, p < .0001). The CMP curriculum includes a signifi-
cantly greater percentage of “one equation with two variables” problems than does
the non-CMP curriculum (z = 35.49, p < .0001). On the other hand, the non-CMP
curriculum includes a significantly greater percentage of “one equation with one
variable” problems than does the CMP curriculum (z = 34.145, p < .0001). These
results resonate with the findings that we reported above. Namely, that the CMP cur-
riculum emphasizes an understanding of the relationships between the variables of
equations, rather than an acquisition of the skills needed to solve them. In fact, of the
402 equation-related problems in the CMP curriculum, only 33 of them (about 8%
of the linear equation solving problems) involve decontexualized symbolic manipu-
lation of equation solving. However, the non-CMP curriculum includes 1,550 prob-
lems involving decontexualized symbolic manipulation of equations (nearly 70% of
the linear equation solving problems in the curriculum).

Classroom Instruction
We conducted 620 detailed lesson observations of CMP and non-CMP lessons over
a three-year period. Approximately half of the observations were of teachers using
the CMP curriculum. The other half were observations of teachers using non-CMP
curricula. Two retired mathematics teachers conducted and coded all the observa-
tions. The observers received extensive training that included frequent checks for
reliability and validity throughout the three years. Over the course of the 6th -grade
year, for example, we checked the reliability of the observers’ coding three times.
The reliability achieved during the three sessions averaged 79% perfect agreement,
using the criterion that the observers’ coded responses were considered equivalent
only if they were identical (i.e., perfect match). The reliability averaged 95% using
170 J. Cai et al.
the following criteria: (a) If an item or sub-item was “scored” using an ordinal scale,
then the observers’ coded responses were considered equivalent if they differed by
at most one unit; (b) If an item or sub-item (e.g. representation) was “scored” by
choosing from a list of alternatives all the words/phrases that characterize it, then
the observers’ coded responses were considered equivalent if they had at least one
choice in common (e.g. symbolic and pictorial vs. pictorial). We reached similar
high reliabilities for the 7th and 8th grade observations.
Each class of LieCal students was observed four times, during two consecu-
tive lessons in the fall and two in the spring. The observers recorded extensive
information about each lesson in a 28-page project-developed observation instru-
ment. During each observation, the observer made a minute-by-minute record of
the lesson on specially designed form. This record was used later to code the
lesson. The coding section of the LieCal observation instrument has three main
components: (1) the structure of the lesson and use of materials, (2) the na-
ture of the instruction, and (3) the analysis of the mathematical tasks used in
the lesson (see Moyer et al. 2010 for details about the lessons observed and ob-
servation instrument used). The analyses revealed striking differences between
classroom instruction using the CMP and non-CMP curricula (Cai et al. 2009;
Moyer et al. 2009). In this section, we discuss the differences related to two im-
portant instructional variables: (1) the level of conceptual and procedural emphases
in the lessons, and (2) the cognitive demand of the instructional tasks implemented.
Conceptual and Procedural Emphases
The second component of the coding section includes twenty-one 5-point Likert
scale questions that the observers used to rate the nature of instruction in a lesson. Of
the 21 questions, four of them are designed to assess the extent to which a teacher’s
lesson has a conceptual emphasis. Another four of the questions are designed to
determine the extent to which a teacher’s lesson has a procedural emphasis. Factor
analysis of the LieCal observation data confirmed that the four procedural-emphasis
questions loaded on a single factor, as did the four conceptual-emphasis questions.
There was a significant difference across grade levels among the levels of con-
ceptual emphasis in the CMP and non-CMP instruction (F = 53.43, p < 0.001).
The overall (grades 6–8) mean of the summated ratings of conceptual emphasis
in CMP classrooms was 13.41, while the overall mean of the summated ratings
of conceptual emphasis in non-CMP classrooms was 10.06. The summated ratings
of conceptual emphasis were obtained by adding the ratings on the four items of
the conceptual-emphasis factor in the classroom observation instrument. That im-
plies that the mean rating on the conceptual-emphasis items was 3.35 ( 13.41 4 ) for
CMP instruction and 2.52 ( 10.06
4 ) for non-CMP instruction. That is, CMP instruc-
tion was rated 0.40 points above the midpoint, while non-CMP instruction was rated
0.5 points below the midpoint. The bottom line is that CMP instruction was rated
an average of about 4/5 of a point higher (out of 5) on each conceptual emphasis
Table 3 Emphasis on conceptual understanding

Grade 6* Grade 7* Grade 8* Overall*
(CMP: n = 96; (CMP: n = 101; (CMP: n = 108; (CMP: n = 305;
non-CMP: n = 87) non-CMP: n = 95) non-CMP: n = 92) non-CMP: n = 274)
CMP 14.51 (3.70)a 12.52 (3.70) 13.27 (3.65) 13.41 (3.76)

Non-CMP 9.44 (2.50) 10.11 (2.31) 10.61 (2.73) 10.06 (2.55)
*p < .001
a numbers in parenthesis are standard deviations
Table 4 Emphasis on procedural knowledge

Grade 6* Grade 7* Grade 8* Overall*
(CMP: n = 96; (CMP: n = 101; (CMP: n = 108; (CMP: n = 305;
non-CMP: n = 87) non-CMP: n = 95) non-CMP: n = 92) non-CMP: n = 274)
CMP 11.67 (3.03)a 11.70 (3.05) 11.48 (3.44) 11.61 (3.18)

Non-CMP 13.77 (3.58) 14.24 (3.32) 15.41 (3.27) 14.49 (3.44)
*p < .001
a numbers in parenthesis are standard deviations
item than was non-CMP instruction, which is a significant difference (t = 11.44;

p < 0.001).
On the other hand, non-CMP lessons had significantly more emphasis on the
procedural aspects of learning than did the CMP lessons. The procedural-emphasis
ratings for the non-CMP lessons were significantly higher than were the procedural-
emphasis ratings for the CMP lessons (F = 37.77, p < 0.001). Also, the overall
(grades 6–8) mean of summated ratings of procedural emphasis in non-CMP class-
rooms (14.49) was significantly greater than was the overall mean of the summated
ratings of procedural emphasis in CMP classrooms, which was 1.61 (t = −9.43,
p < 0.001). The summated ratings for the procedural emphasis were obtained by
adding the ratings on the four items of the procedural-emphasis factor. That implies
that the mean rating on the procedural emphasis items was 3.62 ( 14.49 4 ) for non-
11.61
CMP instruction and 2.91 ( 4 ) for non-CMP instruction. On average, non-CMP
instruction was rated about 7/10 of a point higher (out of 5) on each procedural
emphasis item than was CMP instruction, which is a significant difference.
Instructional Tasks
Using a scheme developed by Stein et al. (1996), we also classified the instruc-
tional tasks implemented in CMP and non-CMP classrooms into four increasingly
demanding categories of cognition: memorization, procedures without connections,
172 J. Cai et al.
Table 5 Percentages of tasks with different levels of cognitive demand in CMP and non-CMP
classrooms
Memorization Procedure Without Procedure With Doing
Connections Connections Mathematics
CMP (n = 623) 2 53 34 11
Non-CMP (n = 695) 11 79 9 1
Table 6 Percentages of tasks with high cognitive demand in each grade level
6th Grade 7th Grade 8th Grade Overall
CMP 62 28 42 45
Non-CMP 12 6 13 10
Note: High cognitive demand tasks refer to tasks involving procedures with connection or doing
mathematics
procedures with connections, and doing mathematics. Table 5 illustrates the percent-
age distributions of the cognitive demand of the instructional tasks implemented (or
enacted) in CMP and non-CMP classrooms. A chi-square test shows that the per-
centage distributions in CMP and non-CMP classrooms are significantly different
(χ 2 (3, N = 1318) = 219.45, p < .0001). The difference confirms that a larger per-
centage of high cognitive demand tasks (procedures with connection or doing math-
ematics) was implemented in CMP classrooms than was implemented in non-CMP
classrooms (z = 14.12, p < .001). On the other hand, a larger percentage of low
cognitive demand tasks (procedures without connection or memorization) were im-
plemented in non-CMP classrooms than was implemented in CMP classrooms (Cai
et al. 2010a, 2010b).
Table 6 shows the percentage of high cognitive demand tasks implemented in
the observed lessons at each grade level. Not only did CMP teachers implement a
significantly higher percentage of cognitively demanding tasks than did non-CMP
teachers across the three grades, but also within each grade (z values range from
6.06–11.28 across the three grade levels, p < .001).
Over 50% of the CMP lessons implemented at least one high level task (involving
either procedures with connections or doing mathematics), but only 19% of the non-
CMP lessons did so (z = 8.91, p < .001). Nearly 80% of the non-CMP lessons
implemented low-level tasks involving procedures without connections, which is
significantly greater than that of the CMP lessons (45%) (z = 8.13, p < .001).
Students’ Development of Algebraic Thinking: Methodological

Considerations
Given the significant differences between the CMP and the non-CMP curricula
themselves, as well as the resulting instruction, how does the development of al-
gebraic thinking differ between students in CMP and non-CMP classrooms? Before
we present results that address this question, we first discuss the methodological
consideration for examining students’ development of algebraic thinking.
The Focus of Algebraic Thinking
Middle school algebra lays the foundation for the acquisition of tools for analyz-
ing quantitative relationships, for solving problems, and for stating and proving
generalizations (Bednarz et al. 1996; Carpenter et al. 2003; Kaput 1999; Mathe-
matical Sciences Education Board 1998; RAND Mathematics Study Panel 2003).
There are many important features of algebraic thinking (e.g., Cuoco et al. 1996;
Kieran and Chalouh 1993; Mason 1996; NCTM 1989; Zazkis and Liljedahl 2002).
In this chapter, we focus on the following three components of algebraic knowledge:
representing situations, solving equations, and making generalizations. Mastering
these three components requires the acquisition of the fundamental habits of minds
(Cuoco et al. 1996) involved in thinking algebraically (Sfard 1995). Not only are
these three components commonly accepted as important, they are also listed in the
National Assessment of Educational Progress’s Framework for 8th graders in the
United States (NAEP 2006).
Representing Situations One of the important aspects of algebraic thinking is to

interpret and represent quantitative situations (Mayer 1987; Kieran 1996). To solve
word problems algebraically, students first need to generate equations to represent
the quantitative relationships involved. Quantitative situations usually involve addi-
tive propositions, relational propositions, or both. For example, “Jake and Tom have
30 marbles altogether” is an additive proposition. “One pound of shrimp costs $3.50
more than one pound of fish” is a relational proposition. Researchers (e.g., Cocking
and Mestre 1988) have found that students have difficulty interpreting and represent-
ing quantitative relationships, especially those that involve relational propositions.
Equation Solving Being able to solve equations is a basic and important aspect
of algebraic thinking. In fact, when the topic of algebra in school mathematics is
brought up, the first thing that usually comes to people’s minds is equation solving.
Historically, equation solving has played a central role in the development of other
aspects of mathematics, and in solving real-life problems. Even though there has
been a major shift in the landscape of school mathematics in recent years (Chazan
2008), learning to solve equations is still an essential element in the study of algebra
(Mathematical Sciences Education Board 1998).
Making Generalizations Making generalizations from patterns is at the heart of

mathematical thinking in general and of algebraic thinking in particular (Lee 1996;
Steen 1988). School mathematics has increasingly emphasized the use of pattern to
develop students’ ability to make and express generalizations (NCTM 1989; Orton
and Orton 1999).
174 J. Cai et al.
Tasks and Data Analysis
In this section, we present findings about the development of students’ algebraic

thinking. These findings are based on data that we collected in the LieCal Project.
In the LieCal Project, we used multiple measures to examine the effect of curriculum
on students’ learning. Besides using the state test, we developed 32 multiple-choice
items and 13 open-ended tasks to assess students’ Learning (see Cai et al. 2010a,
2010b for details). Of the 32 multiple-choice items, six of them were representing-
situations tasks and another six were equation-solving tasks. The open-ended tasks
required the students to provide answers and explain how they got them. These items
and tasks were administered to 1,300 middle school students four times (fall 2005,
spring 2006, spring 2007, and spring 2008).
Below, we report the results of the six representing-situations items and the six
equation-solving items. From one testing administration to another, two of the six
representing-situations tasks were identical, and the other four were parallel. The
design was similar for the six equation-solving tasks. We also report the results
on one of the open-ended tasks. The open-ended task was a making-generalizations
task that was common across all forms and administrations. Appendix shows sample
items.
For representing-situations tasks and equation-solving tasks, we mainly used
scaled scores in our analysis. A scaled score is a generic term for a mathemati-
cally transformed student raw score on an assessment. Even though we used parallel
items, it is still possible that students responded to the parallel items differently than
they did to the original items. Parallel items were created through piloting and expert
judgment. Using scaled scores, assessment results can be placed on the same scale
even though students responded to different (but parallel) tasks at different times. In
particular, we used the two identical representing-situations items to scale students’
performance on the representing situations, and used the two identical equation-
solving items to scale students’ performance on equation solving. Since the scaled
scores estimated the students’ ability to represent situations or solve equations as
a whole, the scaled scores could not be used in an item analysis. When an item
analysis was needed, we used raw scores.
For the making-generalizations task, we conducted a qualitative analysis that cap-
tured the correctness of students’ answers and the kinds of strategies they employed.
When we analyzed the students’ solution strategies, we paid particular attention to
their use of concrete or abstract strategies.
Findings about the Development of Students’ Algebraic Thinking
We summarize the data on the development of students’ algebraic thinking in three

sub-sections, each focusing on one of the three types of tasks: representing situa-
tions, solving equations, and making generalizations.
Representing Situations
Repeated Measures ANOVA Table 7 shows the mean scaled scores on represent-
ing-situations tasks for both CMP and non-CMP students across the four testing
administrations. Repeated measures analysis of variance showed that both CMP
and non-CMP students showed significant growth from the fall of 2005 (6th grade)
to the spring of 2008 (8th grade) (F = 275.73, p < .001). CMP students started
a bit lower than non-CMP students in the fall of 2005, but by the spring of 2008,
the CMP students performed better than did the non-CMP students. The CMP and
non-CMP students not only had different patterns of growth, but also different rates
of growth, as shown in Fig. 4. A repeated measure ANOVA with mixed design
analysis indicates that CMP students had a significantly higher growth rate than did
non-CMP students on representing-situations items (F = 2.61, p < .05).
HLM Analysis Because the data is hierarchical in nature, we also used multilevel
statistical models to capture student achievement changes over time and to analyze
Table 7 Mean scaled scores and standard deviations for CMP and non-CMP students on tasks
involving representing situations
Fall 2005 Spring 06 Spring 07 Spring 08
CMP Students 449 501 536 563

(n = 312)a (92) (94) (96) (91)
Non-CMP Students 461 502 544 554
(n = 309) (90) (92) (94) (89)
a The number of students reported in this table only includes those students who took all four
assessments from fall 2005 to spring 2008
Fig. 4 Mean scaled scores for CMP and non-CMP students on representing situations
176 J. Cai et al.
the development of algebraic thinking in a curricular context (Raudenbush and Bryk

2002). In particular, we used growth curve modeling to examine the longitudinal ef-
fect of curriculum, taking into account both student level variables (e.g., gender
and ethnicity) and classroom variables (conceptual and procedural emphases) (see
Cai et al. 2009 for details). Similar to what was found using the repeated measures
ANOVA, the growth curve modeling showed that CMP students had a significantly
higher growth rate on the representing-situations tasks than did non-CMP students
(t = 2.24, p < .05). The level of conceptual emphasis in instruction had a posi-
tive impact on the growth rate of students’ performance on these tasks (t = 2.79,
p < .05). In fact, we found that a unit increase in the level of conceptual emphasis
resulted in an increase of 4.26 scaled-score points per year in the students’ growth
rate on the representing-situation tasks. The level of procedural emphasis in instruc-
tion, however, did not have a statistically significant impact on the growth rate of
students’ performance on the representing situations tasks (t = −0.64, p = .53).
Using growth curve modeling techniques, we were able to control for class-
room variables related to conceptual and procedural analysis. When we did so,
we found that the difference between CMP and non-CMP students’ growth rate
on representing-situations tasks was no longer significant (t = 1.38, p = .17). This
is to say that when the teacher’s conceptual and procedural emphases were at the
same level, there was no difference between CMP and non-CMP students with re-
spect to their growth on representing situations tasks. Similarly, the impact of the
level of conceptual understanding in instruction also became insignificant when the
students’ CMP status was controlled (t = 1.35, p = .18). This is to say that when
the students were either all CMP students or all non-CMP students, there was no
difference between conceptual and procedural emphases with respect to students’
growth on representing situation tasks.
Item Analysis To further examine how CMP and non-CMP students performed
across the four assessments, we conducted item analyses using raw scores. For both
CMP and non-CMP students, Fig. 5 shows the means of the percent scores on items
involving additive and relational propositions. As was expected, both CMP and non-
CMP students performed better on the items involving additive propositions than on
Fig. 5 Mean percentages of CMP and non-CMP students on items involving additive and rela-
tional propositions
the items involving relational propositions. By the spring of 2008, the vast majority
of the students were able to answer questions related to additive propositions cor-
rectly, but only a bit over a half of them were able to answer questions involving
relational propositions correctly.
From the fall of 2005 to the spring of 2008, both CMP and non-CMP students
had greater increases on items involving relational propositions than on additive
propositions, even though they had higher mean percentages on items involving ad-
ditive propositions than on relational propositions. CMP students showed greater
increases than non-CMP students on both types of items, although the CMP stu-
dents’ additive proposition increase was not significantly greater than the non-CMP
students’. In fact, on items involving additive propositions, the increase for CMP
students was about 20% from the fall of 2005 to the spring of 2008, but the increase
was about 15% for the non-CMP students (z = 1.64, p = .10). On items involving
relational proposition, the increase for CMP students was over 30% from the fall
of 2005 to the spring of 2008, but the increase was about 25% for the non-CMP
students (z = 2.88, p < .01).
Even though the increase over the three years was higher for CMP students than
non-CMP students, an item analysis revealed that both CMP and non-CMP students
showed very similar error patterns. For example, when solving items involving re-
lational propositions, the most common error for both CMP and non-CMP students
was to make an incorrect “literal translation” (Cocking and Mestre 1988). For in-
stance, for the relational proposition “One pound of shrimp costs $3.50 more than
one pound of fish,” the most common error for both groups was to respond that
“shrimp’s cost per pound +$3.50 = fish’s cost per pound,” rather than “shrimp’s
cost per pound = $3.50+ fish’s cost per pound.”
Solving Equations
Repeated Measures ANOVA Table 8 shows the mean scaled scores on tasks in-
volving equation solving for both CMP and non-CMP students across the four test-
ing administrations. Both CMP and non-CMP students showed significant growth
Table 8 Mean scaled scores and standard deviations for CMP and non-CMP students on tasks
involving equation solving
Fall 2005 Spring 06 Spring 07 Spring 08
CMP Students 459 466 505 505

(n = 312)a (80) (76) (93) (91)
Non-CMP Students 498 504 543 543
(n = 309) (90) (92) (94) (89)
a The number of students reported in this table only includes those students who took all four
assessments from fall 2005 to spring 2008

178 J. Cai et al.
from the fall of 2005 (6th grade) to the spring of 2008 (8th grade) (F = 177.72,
p < .001). CMP students started lower than non-CMP students in the fall of 2005,
and the advantage of the non-CMP students continued for all three years. Overall,
the growth rates for CMP and non-CMP students were almost identical, about 45
scaled-score points from the fall of 2005 to the spring of 2008. A repeated mea-
sures ANOVA with mixed design analysis indicated that there was no significant
difference between the growth rates of CMP and non-CMP students over the three
years (F = .45, p = .717). This finding suggests that students had similar gains in
equation solving ability, regardless of the curriculum used.
HLM Analysis Similar to what we found using the repeated measures ANOVA,
growth curve modeling showed that CMP and non-CMP students had comparable
growth rates over the three years on equation-solving tasks (t = −0.39, p = .70).
Instructional differences in the level of conceptual emphasis did not have a signif-
icant impact on the growth rate of equation solving skills (t = 1.12, p = .26), nor
did instructional differences in the level of procedural emphasis (t = 0.50, p = .62).
Making Generalizations
In this section, we report the results from one of the 13 open-ended tasks, namely
the Doorbell task (see Appendix). The Doorbell task includes four questions. The
answer to each succeeding question requires more generalization than the previous
one. To effectively convey how CMP and non-CMP students’ ability to generalize
grew over the three years, we will report the differences between the fall of 2005
and the spring of 2008 data.
Success Rates on Questions A, B, and C Table 9 shows the success rates for
CMP and non-CMP students on questions A, B, and C. To answer question A, stu-
dents need to find the number of guests entering on the 10th ring. The CMP students’
success rate on this question increased significantly from 55% in the fall of 2005 (6th
grade) to 76% in the spring of 2008. The success rate of the non-CMP students also
increased significantly (62% to 73%) from the fall of 2005 to the spring of 2008.
However, the success rate of CMP students increased significantly more than did
the success rate of the non-CMP students (z = 3.52, p < .001).
Table 9 Percentages of
students having correct Fall 2005 Spring 08
answers for questions A, B, QA QB QC QA QB QC
and C of the Doorbell
problem CMP Students 55 4 3 76 19 15
(n = 296)
Non-CMP Students 62 5 3 73 21 10
(n = 299)
The success rates were lower on the later questions for both CMP and non-CMP
students. For example, in spring 2008 only about 20% of the CMP and non-CMP
students correctly answered question B (How many guests will enter on the 100th
ring?). Furthermore, question C (299 guests entered on one of the rings. What ring
was it?) was correctly answered by only 3% of the CMP and 3% of the non-CMP
students in the fall of 2005. By the spring of 2008, the percentages had increased
to 15% for CMP students and 10% for non-CMP students. Even though still only
small proportions of the CMP and non-CMP students were able to answer question
C correctly in the spring of 2008, the increase for CMP students was significantly
greater than that for the non-CMP students (z = 1.99, p < .05). This finding sug-
gests that both CMP and non-CMP students increased their generalization abilities
over the middle school years. It also suggests that, on average, CMP students devel-
oped greater generalization abilities than non-CMP students over the middle school
years.
Solution Strategies An examination of the students’ solution strategies confirmed

this finding. We coded the solution strategies for each of these questions into two
categories: abstract and concrete. Students who chose an abstract strategy generally
followed one of two paths. Students who followed the first path noticed that the
number of guests who enter on a particular ring of the doorbell equals two times
that ring number minus one (i.e., y = 2n − 1), where y represents the number of
guests and n represents the ring number. Students on the second path recognized
that the number of guests who enter on a particular ring equal the ring number plus
the ring number minus one (i.e., y = n + (n − 1)). Using their generalized rule, these
students were able to determine the ring number at which 299 guests entered. Those
who used a concrete strategy also generally took one of two paths. Those on the first
path made a table or a list, while those on the second path noticed that each time the
doorbell rang two more guests enter than did on the previous ring and so added 2’s
sequentially to find an answer.
In the fall of 2005, one CMP student and none of the non-CMP students used an
abstract strategy to correctly answer question A, but in the spring of 2008, nearly
9% of the CMP students and 9% of the non-CMP students used abstract strategies
to correctly answer question A.
Table 10 shows the percentages of CMP and non-CMP students who used con-
crete or abstract strategies to correctly answer questions B and C. In the spring of
2008, nearly 20% of the CMP students and 19% of non-CMP students used an ab-
stract strategy to correctly answer question B. Only a small proportion of the CMP
and non-CMP students used abstract strategies to correctly answer question C in the
spring of 2008. However, the rate of increase for the CMP students who used ab-
stract strategies from the fall of 2005 to the spring of 2008 was significantly greater
than that for non-CMP students (z = 2.58, p < .01). Thus, these results confirmed
that both CMP and non-CMP students increased their generalization ability over
the middle school years. However, on average, the CMP students developed their
generalization ability more fully than did non-CMP students.
180 J. Cai et al.
Table 10 Percentages of students having correct answers for questions B and C of the Doorbell
problem
Fall 2005 Spring 08
Concrete Abstract No Concrete Abstract No
Strategy Strategy Strategy Strategy Strategy Strategy
QUESTION B
CMP Students 27 1 72 14 20 66
(n = 296)
(n = 299)
QUESTION C
CMP students 9 0 91 11 9 80
(n = 296)
(n = 299)
Success Rate for Question D The Question D asked students to write a rule or
describe in words how to find the number of guests that entered on each ring. The
percentages of getting correct rule for Question D of Doorbell for CMP students are
.3% in the fall of 2005, 6.4% in the spring of 2006, 3.7% in the spring of 2007, and
15.9% in the spring of 2008. The percentages of getting correct rule for Question D
of Doorbell for Non-CMP students are 1.3% in the fall of 2005, 5.4% in the spring
of 2006, 6.0% in the spring of 2007, and 15.7% in the spring of 2008. There is no
significant difference between CMP and Non-CMP students in each grade level in
terms of the percentages of the students who got the correct rule to find the number
of guests that entered on each ring. The rates of increase from the fall of 2005 to the
spring of 2008 between CMP and non-CMP are similar (z = .40, p = .35).
Conclusions and Instructional Implications

Middle school algebra lays the foundation for the acquisition of tools for represent-
ing and analyzing quantitative relationships, for solving problems, and for stating
and proving generalizations. Given recent efforts at curriculum reform, there is an
urgent need to understand the role that curriculum plays in students’ learning of
mathematics in general and in the acquisition of algebraic concepts in particular
(NCTM 1989; National Research Council 2004; RAND Mathematics Study Panel
2003; Senk and Thompson 2003; Usiskin 1999). A study like the one presented in
this chapter is significant not only because it investigates the development of stu-
dents’ algebraic thinking in middle grades, but also because it examines explicit
connections between the acquisition of algebraic concepts and the manner in which
algebra is taught and learned using two different types of curricula.
The findings presented here showed that across the middle school years, no mat-
ter which type of mathematics curriculum the students in the LieCal Project used,
they all showed significant growth on representing problem situations, solving equa-
tions, and making generalizations. On the one hand, students became much more
capable of representing problem situations, solving equations, and making gener-
alizations in the spring of 2008 than they were in the fall of 2005. For example, it
is encouraging that the vast majority of the students were able to represent addi-
tive propositions correctly by 8th grade, and also that some 8th graders were able to
make and represent generalizations in the Doorbell problem situation. On the other
hand, in an absolute sense, the performance level attained by 8th grade is not very
encouraging. For example, only slightly over a half of the 8th graders were able to
successfully represent the relational propositions, and only about 70% of the 8th
graders were able to solve simple equations. In particular, only a very small propor-
tion of the 8th graders were able to use abstract strategies and generalize the pattern
in the Doorbell problem.
What should we expect from middle school students in terms of algebra? The
National Assessment of Educational Progress (2006) has listed a set of expectations
for 8th graders. For example, by grade 8, students are expected to write algebraic ex-
pressions, equations, or inequalities to represent a situation, to solve linear equations
or inequalities (e.g., ax + b = c or ax + b = cx + d or ax + b > c) and to gener-
alize a pattern appearing in a numerical sequence or table or graph using words or
symbols. We believe that it is reasonable to expect greater proficiency representing
situations, solving equations and making generalizations than the students in this
study have shown.
How can we expect and foster greater proficiency representing situations, solv-
ing equations, and making generalizations than what the students in this study have
shown? In this chapter, we examined the impact of instruction using two different
types of curricula on the development of students’ algebraic thinking over three
years (grades 6–8). Students who used the CMP curriculum showed significantly
higher growth rates than did non-CMP students on both the representing-situations
tasks and the making-generalization task across the three middle school years. On
the equation-solving tasks, the growth rates for CMP and non-CMP students were
similar. These findings suggest that the use of the CMP curriculum has a pos-
itive impact on students’ development of algebraic thinking, as measured by the
representing-situations tasks and the making-generalization task.
In Standards-based curricula like CMP, one important focus is on developing stu-
dents’ conceptual understanding and higher-order thinking skills. Will Standards-
based curricula’s attention to the development of students’ higher-order thinking
skills come at the expense of the development of basic mathematical skills? The
findings presented in this paper showed that the development of students’ higher-
order thinking skills does not necessarily come at the expense of the development
of basic mathematical skills when using Standards-based curricula like CMP. In
fact, students using the CMP curriculum had growth rates in basic mathematical
skills that were similar to those of the non-CMP students.
Why would the students using the CMP curriculum show significantly greater
growth than the students using non-CMP curricula on the representing-situations
182 J. Cai et al.
tasks and the making-generalization task? The answer to this question might be
related to the nature of curriculum and instruction using CMP and non-CMP cur-
ricula. CMP explicitly uses a functional approach to define equations and intro-
duce equation solving. In particular, in the introduction of equation solving, CMP
emphasizes the connections between a situation and an equation used to represent
it. Moreover, not only did the CMP curriculum include more cognitively demand-
ing mathematical problems than the non-CMP curricula, but also the teachers in
classrooms using CMP curricula implemented more cognitively demanding instruc-
tional tasks. Research has shown that tasks with higher cognitive demand provide
better learning opportunities for students (Doyle 1988; Hiebert and Wearne 1993;
Stein et al. 1996). Therefore the CMP students’ more frequent engagement in cog-
nitively demanding tasks is likely to have contributed to the CMP students’ superior
performance on the representing-situations tasks and the making-generalizations
task.
Appendix: Sample Tasks

Representing Situations
Which number sentence is correct?
One pound of shrimp costs $3.50 more than one pound of fish.
a. shrimp’s cost per pound = fish’s cost per pound + $3.50
b. shrimp’s cost per pound + $3.50 = fish’s cost per pound
c. shrimp’s cost per pound + fish’s cost per pound = $3.50
d. shrimp’s cost per pound = fish’s cost per pound − $3.50
Solving Equations
Find the value of x so that x − 5 = 5
(a). 0 (b). 1 (c). 10 (d). 25
Making Generalizations
Sally is having a party.
The first time the doorbell rings, 1 guest enters.
The second time the doorbell rings, 3 guests enter.
The third time the doorbell rings, 5 guests enter.
The fourth time the doorbell rings, 7 guests enter.
Keep going in the same way. On the next ring a group enters that has 2 more
persons than the group that entered on the previous ring.
A. How many guests will enter on the 10th ring? Explain or show how you found
your answer.
B. How many guests will enter on the 100th ring? Explain or show how you found
your answer.
C. 299 guests entered on one of the rings. What ring was it? Explain or show how
you found your answer.
D. Write a rule or describe in words how to find the number of guests that entered
on each ring.
References
Bailey, R., Day, R., Frey, P., Howard, A. C., Hutchens, D. T., McClain, K., Moore-Harris, B.,
Ott, J. M., Pelfrey, R., Price, J., Vielhaber, K., & Willard, T. (2006a). Mathematics: Applica-
tions and Concepts (teacher wraparound edition), course 1. Columbus, OH: The McGraw-Hill
Company.
Ott, J. M., Pelfrey, R., Price, J., Vielhaber, K., & Willard, T. (2006b). Mathematics: Applica-
Company.
Ott, J. M., Pelfrey, R., Price, J., Vielhaber, K., & Willard, T. (2006c). Mathematics: Applica-
Company.
and Teaching. The Netherlands: Kluwer Academic Publishers.
Bruner, J. S. (1990/1998). Acts of Meaning. Cambridge, MA: Harvard University Press.
Cai, J. (2004). Why do U.S. and Chinese students think differently in mathematical problem solv-
ing? Exploring the impact of early algebra learning and teachers’ beliefs. Journal of Mathe-
matical Behavior, 23, 135–167.
Special Issue.
Cai, J., Moyer, J. C., Wang, N., & Nie, B. (2009). Learning from classroom instruction in a curric-
ular content: An analysis of instructional tasks. In S. L. Swars, D. W. Stinson, & S. Lemons-
Smith (Eds.), Proceedings of the 31st Annual Meeting of the North American Chapter of the
International Group for the Psychology of Mathematics Education (Vol. 5, pp. 692–699). At-
lanta, GA: Georgia State University.
Cai, J., Wang, N., Moyer, J. C., & Nie, B. (April/May, 2010a). Impact of curriculum reform:
Evidence of student learning outcomes in the United States. Paper presented at the annual
meeting of American Educational Research Association. April 30–May 4, Denver, Colorado.
Cai, J., Wang, N., Nie, B., Moyer, J. C., & Wang, C. (April/May, 2010b). Student learning and
classroom instruction using Standards-based and traditional curricula: An analysis of instruc-
tional tasks. Paper presented at the annual meeting of American Educational Research Associ-
ation. April 30–May 4, Denver, Colorado.
Cai, J., Nie, B., & Moyer, J. C. (2010c). The teaching of equation solving: Approaches in Stan-
dards-based and traditional curricula in the United States. Pedagogies: An International Jour-
nal, 5, 170–186.
and Algebra in Elementary School. Potsmouth, NH: Heinemann.
Chazan, D. (2008). The shifting landscape of school algebra in the United States. In C. E. Greenes
& R. Rubenstein (Eds.), Algebra and Algebraic Thinking in School Mathematics (pp. 19–33).
Reston, VA: NCTM.
Cocking, R. R., & Mestre, J. P. (Eds.) (1988). Linguistic and Cultural Influences on Learning
Mathematics. Hillsdale, NJ: Lawrence Erlbaum Associates.
Cuoco, A., Goldenburg, P., & Mark, J. (1996). Habits of mind: An organizing principle for mathe-
matics curricula. Journal of Mathematical Behavior, 15, 375–402.
Detterman, D. K. (1993). Individual Differences and Cognition. Norwood, NJ: Ablex.
Doyle, W. (1988). Work in mathematics classes: The context of students’ thinking during instruc-
tion. Educational Psychologist, 23, 167–180.
Gardner, H. (1991). The Unschooled Mind: How Children Think and How Schools Should Teach.
New York: Basic Books.
Hiebert, J., & Wearne, D. (1993). Instructional tasks, classroom discourse, and students’ learning
in second-grade arithmetic. American Educational Research Journal, 30(2), 393–425.
184 J. Cai et al.
Jones, G., Langrall, C., Thornton, C., & Nisbet, S. (2002). Elementary students’ access to powerful
mathematical ideas. In L. D. English (Ed.), Handbook of International Research in Mathemat-
ics Education (pp. 113–142). Mahwah, NJ: Erlbaum.
Kaput, J. (1999). Teaching and learning a new algebra. In E. Fennema & T. Romberg (Eds.),
Mathematics Classrooms that Promote Understanding (pp. 133–155). Mahwah, NJ: Erlbaum.
Katz, V. J. (Ed.) (2007). Algebra: Gateway to a Technological Future. Washington: The Mathe-
matical Association of America.
Kieran, C. (1996). The changing face of school algebra. In C. Alsina, J. M. Alvarez, B. Hodgson,
C. Laborde, & A. Pérez (Eds.), Eighth International Conference on Mathematics Education.
Selected Lectures (pp. 271–290). Seville, Spain: S.A.E.M. Thales.
Kieran, C. (2007). Learning and teaching algebra at the middle school through college levels:
Building meaning for symbols and their manipulation. In F. K. Lester (Ed.), Second Handbook
of Research on Mathematics Teaching and Learning (pp. 707–762). Charlotte, NC: Information
Age Publishing; Reston, VA. National Council of Teachers of Mathematics.
Kieran, C., & Chalouh, L. (1993). Prealgebra: The transition from arithmetic to algebra. In D. T.
Owens (Ed.), Research Ideas for the Classroom: Middle Grades Mathematics (pp. 179–198).
New York: Macmillan Publishing Co.
Lappan, G., Fey, J. T., Fitzgerald, W. M., Friel, S. N., & Phillips, E. D. (2002). Connected Mathe-
matics. Upper Saddle River, NJ: Prentice Hall.
Lave, J. (1988). Cognition in Practice: Mind, Mathematics, and Culture in Everyday Life. Cam-
bridge: Cambridge University Press.
Lee, L. (1996). An initiation into algebraic culture through generalization activities. In N. Bednarz,
C. Kieran, & L. Lee (Eds.), Approaches to Algebra: Perspectives for Research and Teaching.
The Netherlands: Kluwer Academic Publishers.
Mason, J. (1996). Expressing generality and roots of algebra. In N. Bednarz, C. Kieran, & L.
Lee (Eds.), Approaches to Algebra: Perspectives for Research and Teaching. The Netherlands:
Kluwer Academic Publishers.
Mathematical Sciences Education Board (1998). The Nature and Role of Algebra in the K-14 Cur-
riculum: Proceedings of a National Symposium. Washington, DC: National Research Council.
Mayer, R. E. (1987). Educational Psychology: A Cognitive Approach. Boston: Little & Brown.
Moyer, J. C., Cai, J., Laughlin, C., & Wang, N. (2009). The effect of curriculum type on middle
grades instruction. In S. L. Swars, D. W. Stinson, & S. Lemons-Smith (Eds.), Proceedings
of the 31st Annual Meeting of the North American Chapter of the International Group for
the Psychology of Mathematics Education (Vol. 5, pp. 201–209). Atlanta, GA: Georgia State
University.
Moyer, J. C., Cai, J., Wang, N., & Nie, B. (April/May, 2010). Impact of curriculum reform: Ev-
idence of change in classroom practice in the United States. Paper presented at the annual
meeting of American Educational Research Association, April 30–May 4, Denver, Colorado.
National Assessment of Educational Progress (2006). Mathematics Framework for the 2007 Na-
tional Assessment of Educational Progress. Washington, DC: National Assessment Governing
Board.
National Council of Teachers of Mathematics (1989). Curriculum and Evaluation Standards for
School Mathematics. Reston, VA: Author.
National Council of Teachers of Mathematics (1991). Professional Standards for Teaching Math-
ematics. Reston, VA: Author.
National Research Council (1989). Everybody Counts: A Report to the Nation on the Future of
Mathematics Education. Washington, DC: National Academy Press.
National Research Council (2004). On Evaluating Curricular Effectiveness: Judging the Quality
of K-12 Mathematics Evaluations. J. Confrey & V. Stohl (Eds.). Washington, DC: National
Academy Press.
Nie, B., Cai, J., & Moyer, J. C. (2009). How a Standards-based mathematics curriculum differs
from a traditional curriculum: With a focus on intended treatments of the ideas of variable.
Zentralblatt fuer Didaktik der Mathematik (International Journal on Mathematics Education),

41(6), 777–792.
Orton, A., & Orton, J. (1999). Pattern and the approach to algebra. In A. Orton (Ed.), Pattern in
the Teaching and Learning of Mathematics (pp. 237–268). London: Cassell.
RAND Mathematics Study Panel (2003). Mathematical proficiency for all students: Toward a
strategic research and development program in mathematics education (MR-1643-OERI).
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical Linear Models: Applications and Data
Analysis Methods (2nd ed.). Thousand Oaks, CA: Sage.
Resnick, L. B. (1987). Learning in school and out. Educational Researcher, 16, 13–19.
Rogoff, B., & Chavajay, P. (1995). What’s become of research on the cultural basis of cognitive
development? American Psychologist, 50, 859–877.
Senk, S. L., & Thompson, D. R. (Eds.) (2003). Standards-Based School Mathematics Curricula:
What Are They? What Do Students Learn? Mahwah, NJ: Lawrence Erlbaum Associates.
Sfard, A. (1995). The development of algebra: Confronting historical and psychological perspec-
tives. Journal of Mathematical Behavior, 14, 15–39.
Steen, L. A. (1988). The science of patterns. Science, 240, 611–616.
Stein, M. K., Grover, B. W., & Henningsen, M. A. (1996). Building student capacity for mathe-
matical thinking and reasoning: An analysis of mathematical tasks used in reform classrooms.
American Educational Research Journal, 33, 455–488.
Usiskin, Z. (1999, Winter). Which curriculum is best? UCSMP Newsletter, 24, 3–10.
Wozniak, R. H., & Fischer, K. W. (1993). Development in Context: Acting and Thinking in Specific
Environments. Hillsdale, NJ: Erlbaum.
Zazkis, R., & Liljedahl, P. (2002). Arithmetic sequence as a bridge between conceptual fields.
Canadian Journal of Mathematics, Science and Technology Education, 2(1), 93–120.
Years 2 to 6 Students’ Ability to Generalise:
Models, Representations and Theory
for Teaching and Learning
Tom J. Cooper and Elizabeth Warren
Abstract Over the last three years, in our Early Algebra Thinking Project, we have
been studying Years 3 to 5 students’ ability to generalise in a variety of situations,
namely, compensation principles in computation, the balance principle in equiva-
lence and equations, change and inverse change rules with function machines, and
pattern rules with growing patterns. In these studies, we have attempted to involve
a variety of models and representations and to build students’ abilities to switch be-
tween them (in line with the theories of Dreyfus 1991, and Duval 1999). The results
have shown the negative effect of closure on generalisation in symbolic representa-
tions, the predominance of single variance generalisation over covariant generalisa-
tion in tabular representations, and the reduced ability to readily identify common-
alities and relationships in enactive and iconic representations. This chapter uses
the results to explore the interrelation between generalisation, and verbal and visual
comprehension of context. The studies evidence the importance of understanding
and communicating aspects of representational forms which allowed commonali-
ties to be seen across or between representations. Finally the chapter explores the
implications of the results for a theory that describes a growth in integration of mod-
els and representations that leads to generalisation.
From 1999–2001, we were asked by the writers of the new Queensland state math-
ematics syllabus for our advice with regard to algebra in the elementary years. We
This chapter is a revised version of an article published in ZDM—International Reviews on

Mathematical Education, 40(1), 23–37. DOI 10.1007/s11858-007-0066-8.
The research for this paper was funded by Australian Research Council Linkage grant:
LP0348820.
T.J. Cooper ()
Queensland University of Technology, Brisbane, Australia
E. Warren
Australian Catholic University, Brisbane, Australia

188 T.J. Cooper and E. Warren
developed a framework for early algebra based on our knowledge and beliefs at that
time. These were driven by a structural view of mathematics and algebra (Kieran
1990; Sfard 1991), a cognitive perspective on learning (Hiebert and Carpenter 1992;
English and Halford 1995), and an appreciation of students’ difficulties with vari-
ables and the cognitive gap between arithmetic and algebra (Linchevski and Her-
scovics 1996; Usiskin 1988). This framework (Warren and Cooper 2001) was used
as a basis of the new Queensland syllabus which was finalised in 2004 (Queensland
Studies Authority 2003).
In 2002, we undertook a pilot study of Year 2 students’ learning of early alge-
bra. With the support of the Queensland state education department, we successfully
applied for a national grant for a longitudinal project, entitled Early Algebra Think-
ing Project (EATP), to study Years 2 to 6 students’ learning of early algebra. This
study was based on our syllabus framework (Warren and Cooper 2001), our ideas
of effective teaching through connecting multiple representations (Dreyfus 1991;
Duval 2002; Halford 1993; Hiebert and Carpenter 1992), and Krutetskii’s (1976)
generic pedagogies of generalising, flexibility and reversing (see Cooper et al.
2006). The early algebra framework and the representations we used gave rise to
many different examples of generalisation.
This chapter is a sweep across the five years of the pilot and main study for a
cohort of Years 2 to 6 students, a sweep that describes what we found and where
we now wish to go. It initially describes a theoretical framework with regard to
our development of early algebraic thinking and generalisation and representation.
It also describes the focus and design of the project, and discusses generalisation
results with respect to patterning, functional thinking, equations and equivalence,
and arithmetic principles. Finally, it draws conclusions and implications, and relates
findings to an EATP theory with regard to a teaching/learning trajectory designed to
support the development of the ability to generalise (Warren and Cooper 2008).
Special attention is given, in the results section, to functional thinking and arith-
metic principles as these have not been analysed to the same extent as patterning
and equations and equivalence in other EATP publications. The chapter particularly
focuses on generalisation abilities of young children, factors that affect these gen-
eralisation abilities, instructional strategies that are effective in developing general-
isations, and sequences of models and representations that facilitate generalisation.
Perspectives on the Mathematics of Early Algebra

In its most powerful form, we viewed algebra as an abstract system, a system with
interactions that reflected the structure of Arithmetic (Usiskin 1988). Like Dienes
(1961), Skemp (1978) and Wilson (1976), we saw the importance of algebra in
terms of how it represented the principles (e.g., commutative principle and balance
principle), structures of mathematics (e.g., field, group and equivalence class) and
not in terms of the “behaviours” of algebra (such as simplification and factorisation).
In line with Radford (2006), we did not see algebra as the manipulation of letters
but rather as a system characterised by indeterminacy of objects, an analytic nature
of thinking and symbolic ways of designating objects.
Years 2 to 6 Students’ Ability to Generalise: Models, Representations and Theory 189
Structure and Approach In terms of structures, we determined the importance

in laying the groundwork for both the equivalence class and the field structures,
particularly as the equals sign and the operational principles have been identified as
sources of difficulty in the development of algebra for many years (Behr et al. 1980;
Herscovics and Linchevski 1994; Kieran 1990). We also appreciated the need to
develop the two approaches to mathematics, relation and change, from the start.
Mathematics has been categorised by Scandura (1971) as having only three
foci: things; relations between things; and transformations (changes) between these
things, but with every relation capable of being seen as a transformation and every
transformation capable of being seen as a relationship. This is particularly true for
number and operations; for example, the addition of 2 and 3 can be seen in relational
or static terms as “balance”, 2 + 3 = 5, and in transformational or dynamic terms
as “change”, 2 → +3 → 5. From the relation perspective, equals is equivalence or
“same value as” whilst from the transformation perspective, equals is seen as a two-
way mapping in which the change is reversed (a move from a uni-directional mode
of reading an equation to a multi-directional processing of information—Linchevski
1995).
Abstract Schema and Reification We were impressed by Ohlsson’s (1993) dis-

cussion of abstract schema (which we determined in mathematics as being the same
as the mathematical idea of principle), that states the most powerful ideas in math-
ematics are schema where meaning is encoded in the structure or relationships be-
tween the components and not in the form of the components (e.g., a + b = b + a
has meaning independent of what the a and b represent—the meaning is in how the
a and b are related within the rule). Ohlsson (1993) argued that abstract schemas
(and, therefore, for our needs, principles) allow for cross-domain transfer, are highly
portable, and provide the sets of rules by which formal abstract thinking operates.
Ohlsson (1993) saw Piaget’s (1985) process of reflective abstraction as a process
whereby abstract schemas are created from reflection on regularities in cognitive
operations.
Along with Piaget’s (1985) idea of abstraction, Sfard’s (1991) idea of reification
from operational to structural knowledge, where a concept is recognised as “a fully
fledged mathematical object” (p. 14), appeared to provide a perspective on algebraic
development that illuminated the difficulties in learning algebra (see Warren and
Cooper 2001, for a discussion of these difficulties). We appreciated that number
could be reified from counting everyday items to the object of thought that allows,
for example, numbers to be added without reference to sets of items. We also saw
that, at a later time, variable could be reified, in turn, to the object of thought when
students were experienced with mathematical ideas that are true for any number.
Thus, we came to regard algebra as a second level reification, a reification of a
previous reification, indicating that algebra difficulties might be reduced if number
was fully reified, that is, no longer a set of items, and generalisation was an outcome
for as many number activities as possible.
Framework and EATP In our discussions with the Queensland mathematics syl-
labus writing committee, a framework (Warren and Cooper 2001) was developed
that placed early algebra activities within a special Patterns and Algebra strand and
as part of the Number strand of the Queensland Years 1–10 mathematics syllabus
(Queensland Studies Authority 2003). This framework encompassed: (i) pattern and
functions, the study of repeating and growing patterns and of early functional think-
ing (focusing on change); (ii) equivalence and equations, the study of equivalence,
equations and expressions; and (iii) arithmetic generalisation, the study of number
that involves generalisation to principles. The framework also saw early algebra as
the development of mental models based on relationships between real world in-
stances, symbols, language, drawings and graphs, particularly those that enabled
the modelling of real situations that contained unknowns and variables.
It was evident that the basis of these activities is the ability of students to gener-
alise: (i) from position to term in numerical and visual growing patterns (e.g., War-
ren 2005b); (ii) from tables of values to relationships between numbers down and
across columns, and to graphs in patterning and functional situations (e.g., Warren
2005a; Warren and Cooper 2007); (iii) from particular examples to general princi-
ples (e.g., Dougherty and Zilliox 2003; Warren 2005b); and (iv) from real-world
situations to abstract representations (Carraher et al. 2006; Schliemann et al. 2001;
Warren 2005b). Thus, improving one’s ability to generalise lies at the foundation of
efforts to enhance participation in and learning of algebra. As Dienes (1961) argued:
“If this generalisation [from an initial class of small familiar numbers to ‘any num-
ber’] does not take place, algebra cannot possibly be understood” (pp. 289–290).
Generalisation also lies at the heart of mathematics; as Lannin (2005), supported by
Kaput (1999) and Mason (1996), argued: “Statements of generality and discovering
generality are at the very core of mathematical activity.”
In EATP, we have been studying the act of generalisation, in particular, pattern
rules with growing patterns, change and inverse change rules with function ma-
chines and tables of values, balance principle in equivalence and equations, com-
pensation principles in computation, and abstract representations of change (e.g.,
tables, arrow diagrams, graphs) and relationship (equations). We have studied these
acts of generalisation across a variety of situations, in diverse contexts, and with
a range of representations, looking particularly at the relationships between repre-
sentations and growth of algebraic thinking. This has reinforced our position that
generalisation appears to be a major determiner of growth in algebraic thinking and
preparation for later learning of algebra.
Representation and Generalisation

Our research in EATP was based on sequences of teaching experiments using the
conjecture driven approach of Confrey and Lachance (2000) with the aim to pro-
duce both theoretical analyses and instructional innovations (Cobb et al. 2000).
We postulated hypothetical learning trajectories within each class and formulated
conjectures (in terms of both mathematical content and pedagogy) about envisaged
learning processes and specific means that might support these processes. Our ba-
sis for these conjectures was multi-models (e.g., balance and number line together)
and multi-representations (i.e., connecting language, diagrams and figures, symbol

systems and graphs).
Models and Representations
There has been general consensus for some time that mathematical ideas are repre-
sented externally (i.e., concrete materials, pictures/diagrams, spoken words, written
symbols) and internally (i.e., mental models and cognitive representations of the
mathematical ideas underlying the external representations) (Putnam et al. 1990).
Models and representation are related; models are ways of thinking about abstract
concepts (e.g., balance for equivalence) and representations are various forms of the
models (e.g., physical balances, balance diagrams, balance language, equations as
balance). Mathematical understanding is the number and strength of the connections
in a student’s internal network of mental models and representations (Hiebert and
Carpenter 1992). It has long been argued that generalising mathematics structures
involves determining what is preserved and what is lost between the specific struc-
tures which have some isomorphism (Gentner and Markman 1994; Halford 1993).
An example of such a structure is what is common between the subtraction algo-
rithm for whole numbers, decimal numbers, common fractions and measures. This
is referred to in the literature as the Mapping Instruction approach (English and
Halford 1995; Peled and Segalis 2005).
Dreyfus (1991) argued that learning proceeds through four states, namely, using
one representation, using more than one representation in parallel, making links be-
tween parallel representations, and integrating representations and flexibly moving
between them. Duval (2002) extends Dreyfus’ argument; he argues that mathemat-
ics comprehension results from coordination of at least two representation forms
or registers and that there are four registers; the multifunctional registers of nat-
ural language and figures/diagrams, and the mono-functional registers of notation
systems (symbols) and graphs. Again extending Dreyfus, Duval contends that learn-
ing involves moving from treatments where students stay within one register (e.g.,
carrying out calculations while remaining strictly in the one notation system) to con-
versions where students change register without changing the objects being denoted
(e.g., passing from natural language of a relationship to using letters to represent it)
and finally to coordination of registers. He argues that learning also requires build-
ing understanding of the mathematical processing performed in each register (Duval
1999). A further distinction between Dreyfus and Duval is that Duval also suggests
that representations play an epistemological and pedagogical role in teaching and
learning. It is this distinction that guided us in our classroom interactions.
Generalisation
Literature with respect to patterns, principles and abstract representations, and math-
ematics induction provides a framework for generalisation. For patterns and tables
of values, Lannin (2005) distinguishes between two types of generalisation: recur-

sive (single variant or sequential between terms, e.g., add 2 to get the next term) and
explicit (covariant or between term and position, e.g., the 25th term is 2 × 25 − 1).
For each of these types, Radford (2003) argues that generalisation develops through
three levels: (i) factual, where the generalisation focus remains at the level of the ma-
terial to be generalised, such as “the four counters” and is gesture and rhythm driven
to show the generalisation; (ii) contextual, where the focus is on more abstract and
descriptive terms such as “the next figure” and is language driven to explain the gen-
eralisation; and (iii) symbolic, where algebraic notation (including letters) is used to
describe the generalisation.
For principles and abstract representations, research in Measure Up (Dougherty
and Zilliox 2003) and EATP (Warren 2005b; Warren and Cooper 2007) has shown
that very young students can generalise the Equivalence Class principles from ac-
tivity in numberless situations (work pioneered by Davydov 1975). Other research
(e.g., Carraher et al. 2006) has shown that young students can generalise to abstract
representations, and that such activity results in better understandings of mathe-
matical structures in later years (Morris 1999). Goodson-Espy (1998) argues for
Sfard’s (1991) sequence of interiorisation, condensation and reification and Ci-
farelli’s (1988) reflective abstraction levels of: recognition, re-presentation, struc-
tural abstraction and structural awareness for generalisation to principles. Peled and
Segalis (2005) argue for the Mapping Instruction approach to teaching (English and
Halford 1995) which focuses on identifying similarities between isomorphic proce-
dures (e.g., what is the same in the processes for 34–16 and 3 weeks 4 days subtract
1 week 6 days).
Overall, Harel’s (2001) two different forms of mathematics-induction generali-
sation appear applicable to all types of generalisation. He argued that there were
two forms: (i) results generalisation, developing a generality from a few examples;
and (ii) process generalisation, developing a generality from a few examples and
then justifying it in terms that show its applicability to all examples or any number.
His distinction appears similar to Radford’s (2003, 2006) induction-generalisation
and Lannin’s (2005) empirical-generic distinctions. In particular, Radford (2006)
suggested that true generalisation involves noticing a local commonality and then
generalising this across all terms (perceiving the particular and then using this to
conceive the general). With regard to these forms of generalisation, Lannin (2005)
distinguished between generalisation from iconic/visual and numerical representa-
tions, arguing that iconic is a better representation to lead to process generalisation.
Finally, Radford (2003, 2006) purported that generalisation involves two com-
ponents: (i) grasping a generality (phenomenological) through noticing how a lo-
cal commonality holds across all terms; and (ii) expressing a generality (semiotic)
through gestures, language and algebraic symbols. The act of grasping a generality
through extending a local commonality (process generalisation) has led to a study
of strategies of which some (e.g., trial and error) are not strongly supported and
others (e.g., restructuring visual presentations) have some support. In particular, re-
search into using tables has had advocates for use and non use (e.g., Herbert and
Brown 1997; Orton and Orton 1999; Warren 1996, 2006). Expression of general-
isation has also attracted support with Redden (1996) and Stacey and MacGregor
(1995) suggesting that natural language is a prerequisite for algebraic notation. By

contrast, Bloedy-Vinner (1995), Kaput (1992), Ursini (1991) and Warren (2005b)
have shared misgivings with regard to this conjecture. Fujii and Stephens (2001)
identified the notion of a quasi-variable as a bridge between arithmetic and alge-
braic notation. From their perspective this involves the recognition that a number
sentence or group of number sentences can indicate an underlying mathematical re-
lationship which remains true whatever the numbers used are. Our research suggests
that this idea is extendable to generalisation, to give a notion of quasi-generalisation
as a step towards full generalisation. In this extension, quasi-generalisation is where
students are able to express the generalisation in terms of specific numbers. It ap-
pears often to be the case that students can apply a generalisation to many numbers,
and even to an example of “any number”, before they can provide a generalisation in
language or symbols (Warren 2005b; Warren and Cooper 2007). We have found that
a quasi-generalisation in the elementary context appears to be a necessary precursor
to expressing the generalisation in natural language and algebraic notation.
Focus of EATP
In EATP, we designed instruction to: (i) have grasping and expressing generali-
sation as its major outcome, distinguishing between process, results, recursive and
explicit types of generalisation in patterns and tables, and ensuring that time is spent
on generalising principles and abstract representations; (ii) use a variety of models
(e.g., balance, line, function machine) and representations (e.g., natural language,
figures/diagrams, symbols and, in later years, graphs) to achieve this outcome; and,
(iii) build students’ abilities to switch between representations and models (Drey-
fus 1991; Duval 1999). Given the paucity of literature concerning the development
of algebraic thinking at the elementary level, Bruner’s theory (1966) was utilised
to assist us in selecting representations for the interventions. Our selection broadly
followed the enactive to iconic to symbolic sequence unless another imperative in-
tervened. We considered mathematical development as cumulative rather than re-
placement and thus integrated various models and representations from different
levels (e.g., an iconic picture of balance with enacting the number line). Based on
our belief that no one model or system of representations provides all of the an-
swers, we used comparison of and transition between models and representations
to support the emergence of algebraic thinking (e.g., using the balance and num-
ber line models in unison). Overall, the instruction focused on connecting Duval’s
four registers to real world situations and the acting out of these situations by the
students themselves with physical materials. It also focused on connecting particu-
lar representations emerging from different perspectives (relationship and change—
Scandura 1971), particularly the arrow based symbol system for change with the
equation based system for relationship.
Our use of models and their representations, particularly in their physical or con-
crete representation form, were endowed with two fundamental components (Filloy
and Sutherland 1996), namely, the ability to translate and generalise to abstraction.
Translation encompasses moving from the state of things at a concrete level to the
state of things at a more abstract level with the model acting as an analogue for
the more abstract. Abstraction is believed to begin with exploration and use of pro-
cesses or operations performed on lower level mathematical constructs (English and
Sharry 1996; Sfard 1991).
In line with Filloy and Sutherland (1996), we accepted that models often hide
what is meant to be taught and present problems when abstraction from the model
is left to the pupil. Thus, we saw teacher intervention as a necessity for detachment
from the model to construction of the new abstract notion. As we implemented the
models and representations, we engaged in classroom conversations with the young
students and continually explored new signs that would assist students to extract the
essence of the mathematics embedded in the exploration (Radford’s, 2003, semiotic
nodes). We saw expression and language as essential to this journey as they gave
subtle shades of meaning that arise from the students’ thinking (Tall 2004). Thus,
EATP was based in the socio-constructivist theory of learning, inquiry based dis-
course and the simultaneous use of multi-representations to build new knowledge
(Warren 2006).
The goals of the EATP were to: (i) investigate Years 2 to 6 (6 to 12 year old)
children’s abilities to reason algebraically and, in particular, to generalise arithmetic
and algebraic situations; (ii) identify key transitions in the children’s development of
algebraic reasoning and generalisation; (iii) construct a model of this development;
(iv) develop instructional strategies effective in facilitating this development; and
(v) develop professional development processes that facilitate teacher learning of
these approaches.
Focus of Chapter
This chapter will report generally on the design of EATP and its overall findings
with regard to generalisation, and particularly on Year 5 (9–10 years old) functional-
thinking and Years 3–4 (7–9 years old) mathematics-principles lessons as cases of
how generalisation was studied and how multi-representations were used. The par-
ticular research questions to be answered by the chapter are: (i) at what age can
young children generalise patterns, tables of values, principles and abstract repre-
sentations; (ii) what factors enable and inhibit the development of these abilities;
and (iii) what instructional strategies are effective in this development. For the final
question, the chapter will look across the sequence of models, representations used
and their relative efficacy to draw conclusions with regard to a theory of model and
representation use for effective generalisation (and abstraction).
Design of EATP
The methodology adopted for EATP was longitudinal and mixed method using a
design research approach, namely, a series of teaching experiments that followed a
cohort of students across elementary Year 2 to Year 6 based on the conjecture driven
approach of Confrey and Lachance (2000). It was predominantly qualitative and
interpretive (Burns 2000) but with some quantitative analysis of pre-post tests. In
each year, the teaching experiments investigated the students’ learning in lessons on
patterning and functional thinking (using the change perspective), equivalence and
equations (using the relationship perspective), and principles of arithmetic (related,
in particular, to the Equivalence Class and Field structures).
EATP was based on a re-conceptualisation of content and pedagogy for algebra
in the elementary school and as such the teaching experiments were exploratory
in nature; seeking to identify the fundamental cognitive building blocks on which
student understanding of function could be constructed. The representations cho-
sen were intended to be inclusive of all children; however, the necessity to respond
to individual student needs was a position we acknowledged from the outset. Mul-
tiple sources of data were collected, and only those findings for which there was
triangulation were considered in the analysis. Adequate time was spent in the field
observing the lessons to substantiate the reliability of the collected data (Davis and
Maher 1997).
Participants and Procedure The participants across the years were a cohort of
students, and their teachers, from 5 inner city Queensland state schools with socio-
economic status varying from upper to working class. These students moved through
Year 2 to Year 6 during the time of the study (although in the Pilot Year, Year 2 (6–7
years old), there was only one middle class state school involved). Thus the sample
comprised 3 classes (70–85 students) and 3 teachers in Year 2 and 10 classes (220–
270 students) and 10 teachers in Years 3 to 6. All schools were following the new
Queensland Years 1–10 Mathematics Syllabus (Queensland Studies Authority 2003)
which has a new “Patterns and Algebra” strand from Year 1.
For each teaching experiment, the procedure was that we developed and taught
four one-hour lessons to two classes and prepared detailed lesson plans for the other
teachers to follow. In this way, all classes could receive instruction from all the
teaching experiments, and the teachers could have professional learning with re-
spect to a new area of content that they had to teach in the future. Detailed lesson
plans and professional learning sessions to introduce them were required because,
although all teachers were well credentialed (all had four years training, in line with
Queensland policy), the mathematics component of their training was small and,
like most elementary teachers in Queensland, they were not confident in teaching
mathematics (Nisbet and Warren 2000). As algebra was a new mathematics content
area for them, requiring thinking that had not previously been explored, the teachers
had no teaching experience and were unsure of how to conduct these lessons.
In common with normal practice in Queensland elementary schools, the mathe-
matics taught in the classes had a preponderance of activities related to the develop-
ment of pen and paper algorithms and the acquisition of computational skills. The
teachers used some materials in their teaching, but the use of these materials was
highly prescribed, accompanied by stipulated language and recording processes,
and imitative and teacher directed (Baturo et al. 2003).
The teaching in this study was based on models and representations (see Figs. 1
to 6 for models and representations) that have been used before, particularly in the
sixties and seventies. However, the use of these materials was linked to: (i) a social
semiotic approach to teaching; (ii) an inquiry approach to classroom management
in which students shared their thinking; and (iii) lesson plans and worksheets that
connected representations in terms of drawings, real life stories and symbols.
Instruments and Analysis The instruments used in the teaching experiments

were: (i) observations, all lessons in the teaching experiments taught by the re-
searchers and some lessons taught by teachers were recorded with field notes and
videotaped, with one camera fixed on the teacher and class as a whole and another
camera moving around the class focusing on student activity of interest; (ii) inter-
views, planned yearly interviews with teachers and a sample of students and ad hoc
unplanned interviews with students who showed interesting activity during lessons
were conducted and audiotaped; (iii) reflections, feedback from teachers and other
researchers/research assistants on their perceptions of the lessons and student learn-
ing were recorded in written and/or oral form (audiotaped); (iv) tests, yearly written
tests were conducted to ascertain students’ development of algebraic thinking across
the years, and pre-post written tests covering the content in teaching experiments
were administered before and after the experiments with all pen and paper tests col-
lected and marked; and (v) artefacts, all worksheet materials completed by students
(students directed to show working and not to rub out any errors) were collected
during all lessons, marked and collated.
For analysis, the videotapes of the classroom observations and the audiotapes
of the interviews were transcribed, and the pre-post tests and students’ worksheets
were graded and results placed in Excel spreadsheets. An interpretive, descriptive
protocol analysis was used in exploring the data. This entailed a situated discourse
analysis which provides salient segments omitting where necessary students’ repe-
titions.
These data were combined with field notes and written reflections to provide rich
descriptions of each lesson and each teaching experiment, descriptions that contain
information on relationships between teachers’ teaching actions and students’ learn-
ing responses in relation to records of performance and performance change. These
descriptions were then analysed for evidence of student learning and generalisation
processes that followed from that learning. Finally, data across lessons and teach-
ing experiments were compared for similarities and differences in order to construct
theory with regard to integration of models and representations that leads to Years
2–6 students’ development of algebraic reasoning.
Findings and Discussion

EATP teaching experiments have focused on generalisation in patterns and functions
(the change perspective) and principles and abstract representations (the relationship
perspective). We will discuss each of these in turn.
Patterns
Results in patterns have shown that young students can determine pattern rules (up
to symbols with letters) for growing patterns from tables of numbers and visual
structures of counters (Warren 2005b; Warren and Cooper 2007). Results also in-
dicated that generalising patterns with tables is easier to teach while generalising
with visuals provides more equivalent solutions (Warren 2005a). Further, results
show that young students develop these abilities from recursive to explicit (using
Lannin’s 2005 terms) and via quasi-generalisation (building on Fujii and Stephens’
2001 term) (Warren 2005b; Warren and Cooper 2007, 2008). With further analy-
sis, it also appears likely that results will show pattern-generalisation development
moves from factual to contextual to algebraic (as in Radford 2003) and that pattern
visuals result in better process generalisation (similar to Lannin 2005).
Results have also shown that students can generalise relationships between dif-
ferent materials within repeating patterns across many repeats. In these activities,
students break the pattern into repeats, record in a table the number of each mate-
rial against the number of repeats, and generalise across the columns. The students
have also shown an ability to use these tables to generalise equivalent fractions and
equivalent ratios (Warren 2005a).
A lesson that illustrates some of these findings is a Year 5 (9–10 years old) lesson
that focused on teaching growing pattern generalisation without tables. Students
were required to find the general pattern rule for growing patterns built from blocks
and sticks (as in Fig. 1) directly from the visual geometry of the objects.
This proved very difficult and most students found it hard to notice a commonal-
ity in the examples given, which they could extend to a general rule. Students were
asked to find a step sufficiently large enough (e.g., the 22nd step) that directed atten-
tion towards commonality but not so large that it prevented any commonalities being
noticed (Radford 2006). They were then asked to state the rule for large numbers
(e.g., the 237th step), in language, and for a variable n. Only a few students could
find the rule for a large number; even quasi-generalisation appeared to be difficult
for this cohort of students.
The method that was successful was asking the few students who did find a cor-
rect general rule to describe their thoughts about the pattern, thoughts that enabled
them to generalise a commonality. This was identified by the teacher as a form of
visual analysis and labelled with the student’s name (see Fig. 2). In the next pattern,
Fig. 1 Typical growing patterns

Fig. 2 Three forms of visual analysis for the 3, 5, 7, . . . growing pattern
students were directed to use one of these successful students’ methods to analyse
the pattern. Pre-post tests indicated that students became successful in identifying
pattern rules for growing patterns (nearly to the same level as when tables were al-
lowed) but had an enhanced ability to determine more than one rule (Warren 2005b).
From Fig. 2 it can be seen that the three ways of viewing the commonality evident
in Step 3 can be generalised to Step n. All represent different (although equivalent)
rules: n + n + 1, 2 × n + 1, and 2 × (n + 1) − 1. This indicates that visual analysis
is more effective in encouraging the process of generalisation (Harel 2001). But
in each instance students seemed to need to physically separate the pattern into its
components (e.g., separate the two columns from the one tile for 2n + 1) in order to
validate these generalisations.
In a follow-up teaching experiment, visual analysis was taught as a skill to stu-
dents before patterning started. This intervention markedly improved the perfor-
mance of the students. It was evident that students were unfamiliar with the use of
visuals to analyse and trial a variety of solutions.
Change and Functions
The results in the functions segment of EAPT indicate that students can, for more
than one operation, identify change rules using the sequence quasi-variable to gen-
eralised language to algebraic symbols (with letters) and these generalisations move
from recursive to explicit as they become more familiar with functional thinking.
Results also show that students can represent real world situations in terms of change
and inverse change using all of Duval’s representation forms, that is, natural lan-
guage, figures/diagrams (activity with function machines and drawings of function
machines), notation systems (tables, arrow diagrams and equations), and graphs
(Warren and Cooper 2007 has some early results). However, enabling students to
construct real-world situations for these representations has proved very difficult.
This can be illustrated by the functions lessons undertaken with Year 5 students
(aged 9–10 years). These lessons were not the traditional generalisation lessons
commonly used to introduce change (lessons where a given set of values is gen-
eralised to a change rule), but lessons where abstract representations were being
developed for a real-world change situation with a variable and two operations. The
Fig. 3 Representations for change (functions)
lessons commenced with a problem which was modelled with two function ma-
chines as a change and this change, and its inverse, were represented on diagrams of
function machines, Input-Middle-Output tables and with arrow diagrams and equa-
tions (see Fig. 3). The lessons moved to recording function machines on worksheets
that focused on developing inverse change (backtracking), representing this change
with letters on the Input-Middle-Output table and with arrow diagrams and equa-
tions. Then, the activity was reversed, commencing with an arrow diagram and de-
veloping tables of values, diagrams of possible function machines and, finally, with
the posing of a problem that the arrow diagram could represent.
The lessons were conducted in a Year 5 (9–10 years old) classroom with 29 stu-
dents in a middle class school. The Year 5 lesson was the third in a sequence across
Years 3, 4 and 5 of four teaching experiments that connected real-world problems,
action with function machines, drawings of function machines, tables of values,
arrow diagrams, equations and graphs, and included opportunities to discuss and
record change with letters as variables. The function machines were robot-like cre-
ations made from large cardboard cartons with input and output holes represented by
‘ears’ and operations (e.g., ×3 and +4) on their fronts. Each machine was capable
of holding two students.
The lessons initially focused on generalising from change problems (e.g., “I
bought pies for the visitors for $3 each and a chocolate worth $4”) to abstract
representations (e.g., arrow diagrams and equations) and algebraic notation (e.g.,
n × 3 + 4). Students in turn came to the front of the class with input numbers
(e.g., 7) which they placed in the first robot’s right ear. The student in/behind the
first carton/robot exchanged this number for the appropriate change card (e.g., 21)
and passed it through the robot’s left ear to the student who carried it to the next
function machine. The remaining students predicted the changes and tracked them
on calculators, calling out, when allowed, the changes in unison. Students were also
brought to the front of the classroom with output numbers (e.g., 16) and were di-
rected to “back up” in the opposite direction past the robots (the students actually
walked backwards) and passed their cards into the robots in the inverse direction (to
introduce backtracking). Once again, the remaining students were asked to predict
the changes and to check them with calculators.
After modelling of the function machines had occurred, the lesson continued with
students completing a worksheet that required them to: (i) draw a diagram to repre-
sent the function machines; (ii) complete an Input-Middle-Output table with num-
bers and letters in the input, output and middle columns; (iii) represent the change
and the inverse change with arrow diagrams and equations; and (iv) write down
in general terms “what is happening” with respect to the change and the inverse
change. Finally, the tasks were revised in the form of a worksheet where there were
columns for all the different representations and only some columns were filled in,
often leaving the problem column empty so that students had to invent problems
appropriate to the other representations.
The interaction between students and teacher showed that students understood
the workings of the function machines in terms of numbers (quasi-generalisation):
Teacher: [standing in between the two function machines with 21 on a card] I’ve got 21.
If it’s times 3 here [pointing to first robot] and plus 4 here [pointing to second
robot], who can tell me what the number would go to [pointing to the end of
the line of function machines]
Student: 25 [selected from many raised hands]
Teacher: If it’s 21 here, what would it be that side [pointing to the start of the machines]?
Student: 7 [selected from many raised hands]
Students were presented with a worksheet consisting of three columns, namely

input, middle and output. They were also informed that the first machine multiplied
the number by 3 (e.g., changed 4 in the input column to 12 in the middle column) and
the second machine added 4 to the result (e.g., changed 12 in the middle column to
16 in the output column). The worksheet had numbers or letters in different positions
in each row. Students were asked to complete the table and then write the rule.
Table 1 summarises the results for this worksheet. A correct response involved the
student correctly completing the row.
Table 1 Students’ responses on the Input-Middle-Output table for ×3 + 4 change (n = 29)

The position of the number on the sheet Number Number incorrect/not
correct answered
Number 32 placed in Input column 27 2

Number 178 placed in Input column 24 5
Number 34 placed in Output column 28 1
Number 24 in Middle column 22 7
Letter n placed in Input column 12 17
Letter p placed in Output column 10 19
3 × q placed in Middle column 8 21
What is happening here? 21 8
What happens when things are reversed? 11 18
In spite of the success on the worksheet, only some students could repre-
sent the change algebraically. Students experienced little difficulty with represent-
ing the problem with arrow diagrams, inverse arrow diagrams or equations (e.g.,
3 × 7 + 4 = 25), but reversing the equation ((25 − 4) ÷ 3 = 7) caused difficulty.
This was due to students not understanding the BOMDAS arithmetic convention
(Brackets, Of, Multiplication, Division, Arithmetic, Subtraction) which gives the
order of operations. They also did not understand the conventions for reading equa-
tions, that is, the need to write the inverse equation left to right (and not right to
left as is the case in the reverse arrow diagram). Some of the explanations provided
by the students were interesting, for example, a female student stated “The input
is multiplied by 3 which gets the middle then add 4 to the middle” while a male
student stated “You get the same sum you had before except you get your answer
minus 4 then divide by 3 then you should have the same number you started with in
the first one.”
Finally, the worksheets also contained a section where students were asked to
create a real world problem for the linear function represented by the function ma-
chines. They found this task to be particularly difficult. We conjecture that this was
due to the fact that many classroom teachers unpack word problems as equations
but rarely investigate the reverse process.
Equations and Equivalence
Results in the equivalence and equations segment of EATP indicate that very young
students can: (i) represent equivalence in equation form in un-numbered and num-
bered situations, (ii) generalise the Equivalent Class principles for equivalence in
un-numbered situations, and (iii) generalise the balance principle for simple equa-
tions. Results also show that older students can: (i) represent equivalence with un-
knowns in equation form; (ii) generalise the balance principle for all operations; and
(iii) use the balance principle to solve for unknowns in linear equations.
These findings are illustrated in lessons developed to teach the inverse principle
for addition and subtraction as part of the process that leads to solving simple addi-
tion and subtraction problems for unknowns. We define the addition and subtraction
inverse principles as; if any number is changed through addition/subtraction, then
the opposite change (subtraction/addition respectively) by the same amount results
in returning to the original number (i.e., x = x + p − p or x − q + q in algebraic
symbols). The lessons were conducted in a Year 3 (7–8 years old) classroom in
a middle class school and a Year 4 (8–9 years old) classroom in a working class
school. The sample was 22 Year 3 and 28 Year 4 students and 2 teachers. The Year
3 lesson was at the end of a sequence of lessons introducing the balance rule for
addition and subtraction and focused on the students recognising that, for example,
2 has to be subtracted from ? + 2 = 5 to have the unknown (represented by ?) on its
own and find its value. The lesson was designed to be taught with objects, bags (to
represent the unknown) and a balance beam, and diagrams of this beam (to allow
Fig. 4 Representations for equivalence and equations
Table 2 Number of correct

responses in terms of inverse Item: What do you do to both sides? Correct action
balancing action (n = 22)
? + 11 = 36 22
?−7=6 19
8+? = 3 19
? − 30 = 54 15
2×? = 12 4
?÷3=6 5
3×? + 4 = 19 1
for all operations), in order for the students to generalise the process to equations
with algebraic symbols (see Fig. 4).
Previous lessons had: (i) connected the beam balance representation with objects
to number equations (see Fig. 5); (ii) introduced the balance rule (i.e., adding or
removing objects from one side of the equation requires the same action with the
same number of objects to the other side); and (iii) introduced the notion of unknown
with the cloth bag. This lesson discussed how the value of the unknown could be
found by using the balance rule, that is, for ? + 2 = 5, determining the inverse of
the operation and subtracting 2 from both sides. This was reinforced by worksheets
requesting the balancing action and the value of the unknown in picture and equation
form.
Viewing of the videotape showed that most students could determine, for the
example ? + 2 = 5 represented on the balance, that subtracting 2 from both sides
gives the answer 3 for the unknown. This was repeated for the picture worksheet. In
this worksheet students were asked to present the action that would result in finding
the unknown (e.g., for ? + 11 = 36 the correct action is subtract 11 from both sides).
Table 2 presents the number of students for each task who could provide the correct
action.
The correct responses reduced markedly for the questions involving multiplica-
tion and division but it should be noted that no reference had been made during the
lesson to situations involving multiplication, division or cases with two operation
actions and little reference was given to subtraction situations. The multiplication
Fig. 5 Beam balance and number line representations for expressions
and division problems were given to see if any students could generalise the balance
principle from their experiences with addition to the other operations. Multiplication
and division were covered in later lessons and later Years.
The results indicate that some of these young students could successfully trans-
fer the balance principle to other operations. In the Year 4 (8–9 years old) lessons,
a new approach was tried, an approach involving subtraction in conjunction with
addition. The students first discussed what was required to find solutions to addition
and subtraction equations with unknowns. This revolved around determining an ac-
tion that would leave the unknown on its own. To do this, the lesson focused on
the expression that contained the unknown and the operation. This expression was
represented in two ways: first by extending the balance representation in Fig. 5 to
expressions by removing the balance and the objects for the total (i.e., considering
one side of the balance scale), and second by using a number line (see Fig. 5).
The beam balance activity was similar to the Year 3 lesson except that the focus
of the discussion and the worksheets was only on the balancing action not finding
the unknown’s value. The number line activity was new and required the students
to place the unknown anywhere and move right for addition and left for subtraction
(the students had not used a blank number line in this way before). After this skill
was achieved through discussion and worksheets, the students were challenged to
determine the change that would return their action to the unknown. Discussion fo-
cused on generalising the principle that the unknown could be reached by the inverse
operation (−4 for ? + 4 and +3 for ? − 3) as this was equivalent to adding zero. At
this point, the work already completed in functions (see Warren and Cooper 2007)
in identifying inverses reinforced the generalisation as did the Mapping Instruction
approach of comparing addition and subtraction changes.
A final worksheet was used to ascertain students’ understanding of the inverse
principle. It contained items that asked students to draw, for example, +6 on the
number line and to identify the operation that would return one to the unknown. The
results were overwhelming; all 28 students were successful for all items. However,
students were not asked to write a generalisation and there were no items that re-
ferred to, for example, ? + n, the students were only able to show quasi (Fujii and
Stephens 2001) or contextual (Radford 2003) generalisation at best. A viewing of
the videotape showed that some children were able to justify their answers in a way
that indicates that process generalisation was utilised (Harel 2001).
Interestingly, the inverse and balance principles have the opposing actions (the
“opposite” operation for inverse and the same operation for balance). After the suc-
cessful generalising lesson described above which explicitly identified the inverse
principle for expressions with unknowns, some students became confused when this
principle was used in conjunction with the balance principle to solve for unknowns
in later lessons (we call this a compound difficulty). This was particularly so in
the initial teaching of finding solutions to unknowns (e.g., solving ? + 4 = 7) be-
cause getting ? alone requires the inverse operation while the balance rule requires
the same operation. Some students interchanged the processes; one student did this
twice, he solved the problem by saying that finding the value of ? in ? + 4 = 7
meant that 4 had to be added to the left hand side which in turn meant that 4 had to
be subtracted from the right hand side.
Generalising Principles and Abstract Representations
EATP teaching experiments have also focused on supporting students to generalise

principles and abstract representations. Results have shown that students can gener-
alise the Equivalence Class principles for un-numbered situations using unmeasured
lengths and masses (Warren 2005a) and principles associated with identity and in-
verse (e.g., compensation—Warren 2003). They have also shown that students can
generalise to the balance principle for more than one operation. More detailed analy-
sis will be necessary before determining whether this generalisation follows Sfard’s
(1991) or Cifarelli’s (1988) sequence. Students have exhibited the ability to use for-
mal equations with unknowns, represented as a box with a question mark inside, for
both relationship and change situations.
Lessons to illustrate these findings were the first lessons involving the teaching of
the compensation principle for addition and subtraction (Warren 2003; Warren and
Cooper 2003). The addition compensation principle is when adding two numbers
“do the opposite”; if the first number is increased/decreased by an amount, then the
second number is oppositely decreased/increased by the same amount respectively
to keep the sum of two numbers the same (i.e., a + b = (a + k) + (b − k) and (a −
m) + (b + m) in algebraic symbols). The subtraction compensation principle when
subtracting two numbers is “do the same”; if the first number is increased/decreased,
then the second number is increased/decreased the same amount respectively to keep
the difference between two numbers the same (i.e., a − b = (a + k) − (b + k) and
(a − m) − (b − m) in algebraic symbols).
The lessons were conducted in two Year 3 (7–8 years old) classrooms, one from
each of two middle class schools. The sample comprised 45 students and 2 teach-
ers. The lessons were designed to be taught with strips of papers (an un-numbered
situation), then sets of counters, and finally, for addition situations, to having the stu-
dents act out relay races and use number lines and measuring cylinders (see Fig. 6
for representations for addition) to model these situations. Worksheets were spe-
cially developed to reinforce the principles; students were asked to predict, justify
and generalise their findings.
The students were asked to generalise at the end of the strips and number activ-
ity. As Warren (2003) evidences, even though length generalisation had been done
Fig. 6 Models and representations for addition compensation
before number generalisation was discussed, they were more successful at general-
ising for the strips of paper than they were for the number situations, showing the
strength of un-numbered situations in generalisation (in line with the findings of
Davydov 1975; Dougherty and Zilliox 2003). The number line model was very ef-
fective after the students had experienced the strip model as it clearly showed that a
change to one length automatically indicated the compensation that had to be made
to the other length. However, the measuring cylinder model did not work because
although the idea of pouring from cylinder A to B had inbuilt compensation, there
was no connection between A and B (as there was in the number line) and no rela-
tion to a total. Mapping instruction activities were also attempted at the end of both
lessons; the students were asked to compare the length and number activity at the
completion of the first activity and to compare increasing and decreasing numbers
at the end of the second lesson.
The researchers followed the addition-compensation lessons with subtraction
compensation; however, students had great difficulty with the principle. Once again
the same sequence of steps occurred, the papers strips were followed by sets of coun-
ters and then equations with small and larger numbers. However, to show the three
numbers, subtrahend, minuend and difference, required the materials to be used for
subtraction as comparison not takeaway, a more difficult meaning for subtraction.
Secondly, the generalisation was opposite to addition (subtraction compensation re-
quires the same thing to be done to both numbers) and this confused the students.
Little headway was made until the researcher/teacher stopped the sequence and
organised the students to act out the situation using the model of a relay race. Ad-
dition compensation was conducted first with pairs of relay walkers completing a
distance. Through actions, the students came to realise that when the first walkers
increased their distance, the second had less to walk. Then the activity was changed
to one walker walking too far and the second returning to the ‘finish position.’ Again,
through actions, the students identified that when the first walker increased the dis-
tance they walked past the ‘finish line’, the second had more to walk back to the
‘finish line’. This worked successfully at least in ensuring that students understood
that there was a difference between the addition and subtraction compensation pro-
cesses.
Thus, the lesson showed three things. First, it again reinforced the efficacy of the
Mapping Instruction approach (English and Halford 1995) in teaching generalisa-
tion. Second, it showed the power of kinaesthetic activity as a beginning general-
isation activity (supporting gesturing as a beginning generalisation step—Radford
2003). Third, similar to the equations and equivalence lessons, it also showed the
“compound” difficulty that occurs when two principles have opposite effects (show-
ing that structural understanding is difficult to build in small steps).
Conclusions and Implications

Analysis across all years, topics and representations for the EATP data set is on-
going and thus conclusions are tentative. However, evidence for implications for
young children’s learning of algebraic thinking exist in terms of (i) models and
representations, (ii) generalisation overall, and (iii) a theoretical framework for in-
struction integrating all three segments that has application beyond Years 2–6 early
algebra. Some of these are specific to the models and representations, and activities,
presented to students in EATP. Other implications appear to apply more generally
across mathematics teaching and learning, particularly those associated with the
theoretical framework.
Models and Representations
In summary, our use of models and representations in EATP appears to have been
successful with regard to learning and generalisation. The use of blocks and sticks
and step cards (Fig. 1) allowed visual analysis of growing patterns (Fig. 2), lead-
ing to noticing of local commonalities and generalisation to pattern rules. Using
objects and repeat cards has also been successful with repeating patterns in lead-
ing to generalisation of object relationships and equivalent fractions and equivalent
ratios (Warren 2005b). The sequence of function model representations (function
machines, diagrams of machines, tables, arrow diagrams, equations and graphs—
Fig. 3) has been successful in building understanding except in inverse equations
and posing of problems. The function machine, particularly in the form of a large
robot made out of cartons, resulted in highly motivational function lessons. Back-
tracking was illuminated by physically backing students in a reverse direction past
the function machines.
The balance models and the beam representation with different coloured small
cans of food and cloth bags has motivated the study of equations and equals and
this model effectively could be mapped to diagrams and equations (see Fig. 4).
The number line models have also proved effective (see Fig. 5). There has also
been a positive effect on understanding compensation principles in using all three
models, balance (mass), number line (length) and set (objects), at the same time.
The balance/mass model has been particularly effective with regard to the balance
rule.
With regard to the effect of model and representation on generalisation, the im-
portance of making connections between representations (Hiebert and Carpenter
1992) and conversions between registers and domains (Duval 1999) has been high-
lighted. Central to this are the socio-constructivist theory of learning, inquiry based
discourse and the simultaneous use of multi-representations to build new knowl-
edge. The major representations used in EATP were effective, particularly in the
way sequences of representations were used from acting out with materials through
diagrams to language and symbols. In particular, the following model-representation
sets were very effective in motivating students, solving problems and building prin-
ciples and structure: (i) function machines, Input-Output tables and arrow diagrams;
(ii) beam balances, cloth bags and objects and their pictures; and (iii) walking relays,
paper strips and number lines.
Enactive, iconic and symbolic sequences of representations seem to be important
in building towards generalisation (the function machine, diagram, table and arrow
diagram sequence in functions). However, direct teaching of analytic skills in ar-
eas where experiences were weak was found to be necessary. For example, students
found difficulty generalising visual representations in patterns because they had lit-
tle ability in perceiving the visuals in different ways. More success occurred when
a variety of perceptions for visuals was directly taught. This enabled the students
to quickly trial different perceptions to see the highlighted commonalities in the vi-
suals. Before this teaching, generalisation from visual or iconic contexts was less
robust than for symbolic (Radford 2006).
The use of quasi-generalisations appears to be a powerful symbolic strategy to-
wards generalisation in symbolic contexts that, as yet, has no iconic or enactive
counterpart. However, the negative effects of closure in numbered situations ap-
peared to prevent generalisations that were facile in un-numbered situations (e.g.,
the Equivalence Class principles). Finally, understanding and communicating as-
pects of representational forms in a variety of contexts appears to allow commonali-
ties to be seen across or between representations (e.g., inverses in function machines
and number lines).
Generalisation
It is evident, even in the few lessons analysed in this paper, that some findings appear
to have constant application in generalisation.
First, students can learn to understand powerful mathematical structures, usually
reserved for secondary school, in the early and middle years of elementary school if
instruction is appropriate (at least in language and quasi-generalisation form). This
is particularly so for the principles associated with the Equivalence Class and Field
structures. In EATP, because of the separate focus on relationships through equa-
tions and change through function machines, there was cross over for the identity
and inverse principles that reinforced these structures in both perspectives. A teach-
ing focus on structure is a highly effective method for achieving immediate and long
term mathematical goals.
Second, Radford’s (2003, 2006) distinction between grasping and expressing
generalities is important; these are two different things and can be confused by the
teacher. Many times, students’ problems are in expression not identification. Stu-
dents often lacked the language with which to discuss generalisation and lessons
often became language development (e.g., down patterns and across patterns to dis-
tinguish between recursive and explicit pattern rules).
Third, although they were developed for older students, some theories regarding
development of generalisation appear to have strong application in early general-
isation. This is particularly so for: (i) Harel’s (2001) theory regarding results and
process generalisation; (ii) Radford’s (2003, 2006) theory regarding factual, con-
textual and symbolic levels of generalisation; and (iii) Fujii and Stephens (2001)
notion of quasi variable (which we have extended to quasi-generalisation). Harel
directs us towards justifying as well as identifying generalisation, Radford towards
role of gestures (action, movement) and language in early generalisation and Fuji
and Stephens towards the acceptability of number-based descriptions of generalisa-
tions. As well, Lannin’s distinction between recursive (single variance) and explicit
(covariance) has strong application in tables of values.
English and Halford’s (1995) Mapping Instruction teaching approach to princi-
ple generalisation has also proved its efficacy in many lessons. It directs us towards
comparing activity from different domains (e.g., addition and subtraction) and ac-
tivity from different representations (e.g., balance and length). Some attempts have
been made with the Mapping Instruction teaching approach to develop generalisa-
tions by having students focus on similarities between inverse operations in func-
tions and inverse operations in equations. This cross over between the relationship
and change perspectives appears to be one reason why there has been so much suc-
cess with the symbolic work in backtracking and the use of the balance rule for
finding solutions to unknowns in linear equations.
Fourth, EATP has shown that learning can be enhanced by creative represent-
ation-worksheet partnerships that can reinforce connections between representa-
tions. Often teachers restrict worksheets to the symbolic register. EATP has shown
that creative use of pictures and directions can allow a worksheet to reinforce un-
derstandings as well as procedures and to highlight principles.
Fifth, some activities necessary for building structure affect cognitive load. This
is particularly so when large numbers are used to prevent guessing and checking as
a strategy for determining answers and direct students towards the principle. Fur-
thermore, the two examples in this paper have shown the compounding effect of
building structure through small steps. It is necessary to build a superstructure into
which to place conflicting principles such as compensation for addition and subtrac-
tion and inverse and balance for solutions of linear equations.
Finally, although EATP involved much creative lesson development and many
new activities and outcomes, the problems did not really exist with the new work but
with the basic arithmetic prerequisites. Once numbers appeared students attempted
to close on operations and did not attend to pattern and structure to the same extent
as in un-numbered situations (as found in previous research by Davydov 1975).
Furthermore, students’ abilities to interpret and create real world situations in terms
of the actions with materials, diagrams/figures and symbols of early algebra, lagged
far behind their abilities to process the representations and was a constant difficulty
in EATP, a difficulty that increased as the cohort of students moved into their middle
school years. The reason for this appeared to be the lack of teaching of this creation
within the schools. The increased difficulty was a result of examples in higher years
having more complexity (i.e., involving multiplication and division, and more than
one operation).
Theoretical Framework
In Warren and Cooper (2009), we analysed the instructional sequence of equiva-

lence and equations activities described in this paper in greater depth in order to
find hypotheses that would explain and predict teaching and learning behaviour. We
identified nine conjectures that related to the development of equations-equivalence
knowledge and growth in generalisation ability. These conjectures represented a
growth in integration of models and representations that moved towards less reality
(physicality) of representations as the complexity of the tasks and generalisations in-
creased. From a reappraisal of their relationships, the following 6 theory hypotheses
emerged as a basis for an EATP theory for a teaching/learning trajectory designed to
generalise to abstract representations. As this brief analysis here shows, these theory
hypotheses are also supported by the whole of the EATP project; instructional ac-
tivities that were effective for pattern, function machine and generalisation as well
as those for equivalence and equations.
Theory Hypothesis 1 Generalisation to abstraction occurs not within a model or

representation but across models and representations that follow a structured se-
quence. This is a hypothesis that is a consequence of all activity across the topics
and Year levels. There appears to be no ‘magic bullet’; abstraction is generalised
from model to model and representation to representation. This was particularly
evident in the building of inverse and backtracking (using function machine repre-
sentations), the equivalence and balance principle (using balance and number line
models) and the compensation principle (using length and set models) across Years
2 to 6, through progressively more abstract representations and by integrating bal-
ance, number line, set and function machine models and representations (see Figs. 3,
4, 5 and 6).
Theory Hypothesis 2 Effective models and representations show underlying

structure of the mathematics ideas and easily extend to new components and ex-
pand to new applications. This particularly applies to the models and representations
that begin the teaching/learning trajectory; here the criteria for determining effec-
tive models is: (i) strong isomorphism between the desired internal mental model
outcome and initial external model that covers the important aspects of the mental
model; (ii) lack of distracters to direct attention away from isomorphisms; and (iii)
many options in terms of representations that enable the model to extend to new
components (such as variables) and expand to new applications (such as finding so-
lutions to problems). Both the balance and number line models have these attributes;
the number line model was stronger in inverse but the balance made up for this in
its powerful portrayal of equivalence as balance. The function machine model was
particularly strong in its representation options from drawings, tables and arrow di-
agrams to graphs. In a slightly different manner, the grouping of objects into repeats
in repeating patterns enabled the underlying structure in growing patterns to seam-
lessly emerge.
Theory Hypothesis 3 An effective way to structurally sequence models and rep-

resentations to generalise to abstraction is to have models with representations that
develop in four ways: (i) increased flexibility, following the general sequence con-
crete to dynamic diagram to static diagram to symbols (e.g., physical balance to
drawing of balance, blocks for number to symbols); (ii) decreased overt structure,
following the sequence ‘structure in action’ to ‘structure alluded to in picture’ to
‘structure imagined in the mind’; (iii) increased coverage, later representations com-
pensate for limitations in earlier representations (e.g., the balance drawing handles
greater flexibility in operations than the physical balance); and (iv) connectedness
to reality, always relating the form of the representation to real world instances. The
balance model is particularly powerful in terms of its sequence of increased flexibil-
ity and coverage as it moves from physical to diagram representations (see Fig. 4);
if begun with strips of paper, the number line model was also powerful in showing,
in particular, increased flexibility and decreased structure (see Fig. 6). Similarly, the
function machine model showed all three modes of development as it moved from
physical to visual to mental representation form (see Fig. 3).
Theory Hypothesis 4 Sequencing should ensure consecutive steps are nested.

This is a particularly important hypothesis. The nesting of models and represen-
tations to ensure that later versions are subsets of earlier versions causes difficulties
and conflicts if not acceded to. This was most clearly evidenced by the limiting
of generalising ability created by teachers giving prominence to arithmetic com-
putation before equivalence was taught. Covering forms like 4 + 5 = 11 − 2 and
3 × ? − 7 = 8 − 2 × ? means equation is a much more general equivalence form
than computation (e.g., 4 + 5 = ?); computation is a subset of equation. This hy-
pothesis is also important because it implies that the engagement with unnumbered
situations before numbered enables students to effectively attend to mathematical
structure, thus reinforcing the work of Davydov (1975). This was seen in the reduc-
tion of generalising thinking seen when the balance and function machine models
moved from objects (sugar = juice + soap; changing colour) to numbers (6 = 2 + 4;
adding 3). In patterning, this hypothesis supported generating patterns from visual
analysis before numeric tables; the visual methods were stronger in terms of flexi-
bility of results and depth of generalisation (see Fig. 2).
Theory Hypothesis 5 Complex procedures can be facilitated by integrating more

than one model. This is best evidenced by the way balance and number line mod-
els were used together at the point of solving linear equations with an unknown.
However, such integrations can give rise to compound difficulties which require the
development of superstructures (see discussion under generalisation heading). As
shown in Figs. 4 and 5, compensation in balance is direct while compensation on
the number line is inverse causing difficulties for students. Similarly, the compen-
sation rules for addition and subtraction are likewise contradictory and require an
understanding that subtraction is the inverse of addition. However, the notion of
such superstructures is not well developed in the literature, especially with regard to
integrating models to develop ‘deep’ understanding of concepts.
Theory Hypothesis 6 Abstraction is facilitated if comparisons of different models

and representations of the same mental model show commonalities that encompass
the kernel of the mental model. This reflects the success of using the number line and
balance models for the same purpose (solving the equation), particularly in terms
of the extension to variables on both sides of the equation (the didactic cut) and
simultaneous equations. It also implies that effective structured sequences of models
and representations are dual, built around at least two models that act as a spine for
the development of the mathematical idea. This is seen in the integration of visual
and numeric methods for finding patterns as position rules, the different models and
representations (balance, number line; visual and graphical) for the balance model
(see Fig. 4) and the function machine model (see Fig. 3), and the way calculators
are integrated with the function-machine processes.
Final Point The hypotheses above offer promise as the beginning of a theory of
model and representation use in learning-teaching trajectories for generalisation to
abstract representations. This is seen across topics as well as within topics. For ex-
ample, we found it more effective in the later Year levels to cover functional thinking
before equivalence and equations within each year. This was because function ac-
tivity built a strong superstructure around the inverse and identity principles which:
(a) assisted in finding solutions of linear equations with an unknown; and (b) pre-
vented conflict between inverse and balance, and the development of compound
difficulties, in the solution process (Cooper and Warren 2008).
References
Baturo, A. R., Warren, E. A., & Cooper, T. J. (2003). Queensland Numeracy Research and De-
velopment Project Final Report: Teachers Enhancing Numeracy. Report to the Department of
Science and Training, Canberra, ACT.
Behr, M., Erlwanger, S., & Nichols, E. (1980). How children view the equals sign. Mathematics
Bloedy-Vinner, H. (1995). An algebraic interpretations of algebraic expressions: Functions or pred-
icates. In Proceedings of the 19th International Conference for the Psychology of Mathematics
Education (pp. 301–321).
Bruner, J. S. (1966). Towards a Theory of Instruction. New York: Norton.

Burns, R. B. (2000). Introduction to Research Methods (4th ed.). French’s Forest, NSW: Pearson.
Carraher, D., Schliemann, A. D., Brizuela, B. M., & Earnest, D. (2006). Arithmetic and algebra in
early mathematics education. Journal of Research in Mathematics Education, 37(2), 87–115.
Cifarelli, V. (1988). The role of abstraction as a learning process in mathematical problem solving.
Unpublished doctroral dissertation, Purdue University, Indiana.
Cobb, P., Yackel, E., & Mclain, K. (Eds.) (2000). Symbolising and Communicating in Mathe-
matics Classrooms. Perspectives on Discourse, Tools and Instructional Design. Mahwah, NJ:
Lawrence Erlbaum.
Confrey, J., & Lachance, A. (2000). Transformative teaching experiments through conjecture-
driven research design. In A. E. Kelly & R. A. Lesh (Eds.), Handbook of Research Design
in Mathematics and Science Education (pp. 231–265). Mahwah, NJ: Lawrence Erlbaum.
Cooper, T. J., & Warren, E. (2008). The effect of different Representations on Year 3 to 5 students’
ability to generalise. The International Journal of Mathematics Education, 40(1), 23–38.
Cooper, T. J., Baturo, A. R., & Grant, E. J. (2006). Collaboration with teachers to improve numer-
acy learning: Pedagogy at three levels. In J. Novotna, H. Moraova, M. Kratka, & N. Stehlikova
(Eds.), Proceedings of the 30th Annual Conference of the International Group for the Psychol-
ogy of Mathematics Education (pp. 361–368). Prague, Czech Republic: Charles University.
Davis, R. B., & Maher, C. A. (1997). How students think: The role of representations. In L. English
(Ed.), Mathematical Reasoning: Analogies, Metaphors, and Images (pp. 93–115). Hillsdale,
NJ: Lawrence Erlbaum.
Davydov, V. V. (1975). The psychological characteristics of the “prenumeral” period of mathemat-
ics instruction. In L. P. Steffe (Ed.), Children’s Capacity for Learning Mathematics (Vol. VII,
pp. 109–205). Chicago: University of Chicago.
Dienes, Z. P. (1961). On Abstraction and Generalization. Boston: Harvard University Press.
Dougherty, B., & Zilliox, J. (2003). Voyaging from theory and practice in teaching and learning:
A view from Hawaii. In N. Pateman, B. Dougherty, J. Zilliox (Eds.), Proceedings of the 27th
Conference of the International Group for the Psychology of Mathematics Education (Vol. 1,
pp. 31–46). College of Education, University of Hawaii.
Dreyfus, T. (1991). Advanced mathematical thinking processes. In D. Tall (Ed.), Advanced Math-
ematical Thinking (pp. 25–41). Dordrecht: Kluwer Academic Publishers.
Duval, R. (1999). Representations, vision and visualization: Cognitive functions in mathematical
thinking. Basic issues for Learning. In F. Hitt & M. Santos (Eds.), Proceedings of the 21st
Conference of the North American Chapter of the International Group for the Psychology of
Mathematics Education (Vol. 1, pp. 3–26).
Duval, R. (2002). The cognitive analysis of problems of comprehension in the learning of math-
ematics. Paper presented at the Semiotics Discussion Group: 25th Annual Conference of the
International Group for the Psychology of Mathematics Education. Norwich, UK: PME.
English, L., & Halford, G. (1995). Mathematics Education: Models and Processes. Mahwah, NJ:
Lawrence Erlbaum.
English, L. D., & Sharry, P. (1996). Analogical reasoning and the development of algebraic ab-
straction. Educational Studies in Mathematics, 30, 135–157.
Filloy, E., & Sutherland, R. (1996). Designing curricula for teaching and learning algebra. In A.
Bishop (Ed.), International Handbook of Mathematics Education (Vol. 1, pp. 139–160). Dor-
drecht: Kluwer Academic Publishers.
Fujii, T., & Stephens, M. (2001). Fostering understanding of algebraic generalisation through nu-
merical expressions: The role of the quasi-variables. In H. Chick, K. Stacey, J. Vincent, & J.
Vincent (Eds.), The Future of the Teaching and Learning of Algebra. Proceedings of the 12th
ICMI Study Conference (Vol. 1, pp. 258–264). Melbourne, Australia.
Gentner, D., & Markman, A. B. (1994). Structural alignment in comparison: No difference without
similarity. Psychological Science, 5(3), 152–158.
Goodson-Espy, T. (1998). The roles of reification and reflective abstraction in the development
of abstract thought: Transitions form arithmetic to algebra. Education Studies in Mathematics,
36(3), 219–245.
Halford, G. S. (1993). Children’s Understanding: The Development of Mental Models. Hillsdale,

NJ: Erlbaum.
Harel, G. (2001). The development of mathematical induction as a proof scheme: A model for
DNR-based induction. Cited Feb 2007. math.ucsd.edu/harel.
Herbert, K., & Brown, R. (1997). Patterns as tools for algebraic reasoning. Teaching Children
Mathematics, 340–344.
tional Studies, 27, 59–78.
Hiebert, J., & Carpenter, T. P. (1992). Learning and teaching with understanding. In D. Grouws
(Ed.), Handbook for Research on Mathematics Teaching and Learning (pp. 65–97). New York:
Macmillan.
Kaput, J. (1999). Teaching and learning a new algebra. In E. Fennema & T. Romnberg (Eds.),
Mathematics Classrooms that Promote Understanding (pp. 133–155). Mahwah, NJ: Lawrence
Erlbaum.
Kaput, J. (1992). Patterns in students’ formalization of quantitative patterns. In G. Harel & E. Dub-
sinsky (Eds.), The Concept of Function. Aspects of Epistemology (pp. 291–331). Washington,
DC: Mathematics Association of America.
Kieran, C. (1990). Cognitive processes involved in learning school algebra. In P. Nesher & J. Kil-
patrick (Eds.), Mathematics and Cognition: A research synthesis by the International Group for
the Psychology of Mathematics Education (pp. 97–136). Cambridge: Cambridge University.
Krutetskii, V. A. (1976). The Psychology of Mathematical Abilities in Schoolchildren. Chicago, IL:
University of Chicago Press (J. Kilpatrick & I. Wirszup, Trans.).
Lannin, J. (2005). Generalization and justification: The challenge of introducing algebraic reason-
ing through patterning activities. Mathematical Thinking and Learning, 7(3), 231–258.
Linchevski, L. (1995). Algebra with number and arithmetic with letters: A definition of pre-algebra.
Journal of Mathematical Behavior, 14, 113–120.
Linchevski, L., & Herscovics, N. (1996). Crossing the cognitive gap between arithmetic and alge-
bra: Operating on equations in the context of equations. Educational Studies, 30, 36–65.
Mason, J. (Ed.) (1996). Expressing Generality and Roots of Algebra. Dordrecht: Kluwer Academic
Publishers.
Morris, A. K. (1999). Developing concepts of mathematical structure: Pre-arithmetic reasoning
versus extended arithmetic reasoning. Focus on Learning Problems in Mathematics, 21(1), 44–
72.
Nisbet, S., & Warren, E. (2000). Primary school teachers’ beliefs relating to teaching and assess-
ing mathematics and factors that influence these beliefs. Mathematics Teacher Education and
Development, 2, 34–47.
Ohlsson, S. (1993). Abstract schemas. Educational Psychologist, 28(1), 51–66.
Orton, J., & Orton, J. (1999). Pattern and the approach to algebra. In A. Orton (Ed.), Pattern in the
Teaching and Learning of Mathematics (pp. 104–120). London: Cassell.
Peled, I., & Segalis, B. (2005). It’s not too late to conceputalise: Constructing a generalized subtrac-
tion scheme by abstracting and connecting procedures. Mathematical Thinking and Learning,
7(3), 207–230.
Piaget, J. (1985). The Equilibration of Cognitive Structures: The Central Problem of Intellectual
Development. Chicago, IL: University of Chicago Press.
Putnam, R. T., Lampert, M., & Peterson, P. L. (1990). Alternative perspectives on knowing math-
ematics in Elementary schools. Review of Research in Education, 16, 57–149.
Queensland Studies Authority (November, 2003). Mathematics Years 1–10 Syllabus. Prepublica-
tion. www.qsa.qld.edu.au.
Radford, L. (2003). Gestures, speech, and the sprouting of signs: A semiotic-cultural approach to
students’ types of generalization. Mathematical Thinking and Learning, 5(1), 37–70.
Radford, L. (2006). Algebraic thinking and the generalization of patterns: A semiotic perspective.
Paper presented at the 28th annual meeting of the International Group for the Psychology of
Mathematics education (NA), Merida, Mexico.
Redden, T. (1996). Patterns language and algebra: A longitudinal study. In P. Clarkson (Ed.), Tech-
nology in Mathematics Education. Proceedings of the 19th Annual Conference of the Mathe-
matics Education Research Group (pp. 469–476). Rotorua: MERGA.
Scandura, J. M. (1971). Mathematics: Concrete Behavioural Foundations. New York: Harper &
Row.
Schliemann, A. D., Carraher, D., & Brizuela, B. M. (2001). When tables become function tables.
In M. van den Heuvel-Penhuizen (Ed.), Proceedings of the Twenty-fifth Conference of the Inter-
national Group for the Psychology of Mathematics Education (Vol. 4, pp. 145–152). Utrecht:
Drukkerij Wilco.
Sfard, A. (1991). On the dual nature of mathematical concepts: Reflections on processes and ob-
jects as different sides of the same coin. Educational Studies in Mathematics, 22(1), 191–228.
Skemp, R. R. (1978). Relational understanding and instrumental understanding. Arithmetic
Teacher, 26(3), 9–15.
Stacey, K., & MacGregor, M. (1995). The effects of different approaches to algebra in students’
perception of functional relationships. Mathematics Education Research Journal, 20, 147–164.
Tall, D. O. (2004). Thinking through three worlds of mathematics. In Proceedings of the 28th
Conference of PME (pp. 158–161), Bergen, Norway.
Ursini, S. (1991). First steps in generalisation processes in algebra. In Proceedings of the 15th
International Conference for Psychology of Mathematics Education (Vol. 3, pp. 316–323).
Usiskin, Z. (1988). Conceptions of School algebra and uses of variables. In A. Coxford & A. P.
Schulte (Eds.), Ideas of Algebra, K-12. Reston, VA: NCTM (1988 Yearbook: 8–19).
Warren, E. (1996). Interactions between instructional approaches, students’ reasoning processes,
and their understanding of elementary algebra. Unpublished PhD thesis, QUT.
Warren, E. (2003). Young children’s understanding of equals: A Longitudinal Study. In N. Pate-
man, G. Dougherty, & J. Zilliox (Eds.), Proceedings of the 27th Conference of the International
Group for the Psychology of Mathematics Education, Hawaii (pp. 379–387).
Warren, E. (2005a). Patterns supporting the development of early algebraic thinking. In P. Clarkson
et al. (Ed.), Proceedings of the 28th Conference of Mathematics Education Research Group of
Australasia (Vol. 2, pp. 759–766). Sydney: Merga.
Warren, E. (2005b). Young children’s ability to generalize the pattern rule for growing patterns. In
H. Chick & J. Vincent (Eds.), Proceedings of the Twenty-ninth Conference of the International
Group for the Psychology of Mathematics Education (Vol. 4, pp. 305–312). Melbourne, Vic:
University of Melbourne.
Warren, E. (2006). Teacher actions that assist young students write generalizations in words and
in symbols. In J. Novotna et al. (Ed.), Proceedings of the 30th Conference of the International
Group for the Psychology of Mathematics Education (Vol. 5, pp. 377–384). Prague: PME.
Warren, E., & Cooper, T. (2001). Theory and practice: Developing an algebra syllabus for P-7. In
H. Chick, K. Stacey, J. Vincent, & J. Vincent (Eds.), The Future of the Teaching and Learning
of Algebra. Proceedings of the 12th ICMI Study Conference (Vol. 2, pp. 641–648), Melbourne,
Australia.
Warren, E., & Cooper, T. (2003). Arithmetic pathways towards algebraic thinking: Exploring arith-
metic compensation in year 3. Australian Primary Mathematic Classroom, 8(4), 10–16.
Warren, E., & Cooper, T. J. (2007). Repeating patterns and multiplicative thinking: Analysis of
classroom interactions with 9 year old students that support the transition from known to the
novel. Journal of Classroom Instruction, 41(2), 7–11.
Warren, E., & Cooper, T. J. (2008). Generalising the pattern rule for visual growth patterns: Actions
that support 8 year olds thinking. Education Studies in Mathematics, 67(2), 171–185.
Warren, E., & Cooper, T. (2009). Developing mathematics understanding and abstraction: The case
of equivalence in elementary years. Mathematics Education Research Journal, 21(2), 75–95.
Wilson, R. L. (1976). A structural characterization of the simple lie algebra of generalized Cartan
type over fields of prime characteristics. Journal of Algebra, 40, 418–465.
Algebra in the Middle School: Developing
Functional Relationships Through Quantitative
Reasoning
Amy B. Ellis
Abstract Understanding function is a critical aspect of algebraic reasoning, and

building functional relationships is an activity encouraged in the younger grades to
foster students’ relational thinking. One way to foster functional thinking is to lever-
age the power of students’ capabilities to reason with quantities and their relation-
ships. This paper explicates the ways in which reasoning directly with quantities can
support middle-school students’ understanding of linear and quadratic functions. It
explores how building quantitative relationships can support an initial function un-
derstanding from a covariation perspective, and later serve as a foundation to build
a more flexible view of function that includes the correspondence perspective.
Functions and relations comprise a critical aspect of algebra, and recommenda-

tions for supporting students’ algebraic reasoning emphasize an early introduc-
tion of functional relationships in middle school (NCTM 2000). Students’ diffi-
culties in acquiring the function concept is well documented (e.g., Carlson 1998;
Carlson et al. 2002; Cooney and Wilson 1996; Monk and Nemirovsky 1994), which
highlights the need to better support students’ emerging function concepts in ways
that are mathematically productive, setting a strong foundation for more formal
algebraic reasoning at the high school level. In this chapter I argue that reason-
ing directly with quantities and their relationships constitutes a powerful way to
help students build beginning conceptions of function at the middle-school level. In
particular, reasoning with quantities can directly support a covariation approach to
function, while also providing a foundation for reasoning more flexibly with func-
tional relationships later on.
A.B. Ellis ()

School of Education, University of Wisconsin-Madison, Madison, USA

216 A.B. Ellis
What Is Quantitative Reasoning?
Quantities “are attributes of objects or phenomena that are measurable; it is our ca-
pacity to measure them—whether we have carried out those measurements or not—
that makes them quantities” (Smith and Thompson 2007, p. 101, emphasis original).
A quantity is composed of one’s conception of an object, a quality of the object, an
appropriate unit or dimension, and a process for assigning a numerical value to the
quality (Kaput 1995); length, area, speed, and volume are all attributes that can be
measured in quantities. When students engage in quantitative reasoning, they op-
erate with quantities and their relationships; quantitative operations are therefore
conceptual operations by which one conceives a new quantity in relation to one or
more already-conceived quantities (Ellis 2007). For example, one might compare
quantities additively, by comparing how much taller one person is to another, or
multiplicatively, by asking how many times bigger one object is than another. The
associated arithmetic operations would be subtraction and division.
To illustrate the differences between a formal algebraic approach and a quan-
tities-based approach, consider two responses to the following problem about the
nature of quadratic growth:
Problem 1 Explain why the “second differences” for a quadratic function y = ax 2

are 2a for well-ordered tables in which the x-values increase by 1.
This problem emerged from an algebra II classroom in which the students’ in-
troduction to non-linear functions included an algorithm for determining the degree
of a function based on the finite differences rule (Ellis and Grinstead 2008). The
students easily remembered this algorithm, but it was unclear whether anybody un-
derstood its origins.
Justification #1: A typical algebraic argument involves relying on variables to
represent a general case and writing and manipulating expressions. For instance,
one can create a general table for y = ax 2 in which the x-values increase by 1:
Fig. 1 Table of x- and x | y

y-values for y = ax 2 |
1 | a
|
2 | 4a
|
3 | 9a
|
4 | 16a
Calculating the first differences reveals values of 3a, 5a, and 7a. Calculating sec-
ond differences reveals a constant second difference of 2a, and this approach can be
generalized to any three consecutive entries in the table in which x = n, x = (n + 1),
and x = (n + 2). Corresponding y-values will be y = an2 , y = a(n + 1)2 , and
y = a(n + 2)2 . Calculating the first differences reveals values of a(2n + 1) and
a(2n + 3), with the second difference therefore 2a. This approach lives entirely in
the world of symbolic expressions in a manner that is divorced from any realizable
Algebra in the Middle School: Developing Functional Relationships 217
situation and its constituent quantities. As a formal algebraic justification it provides

a valuable opportunity to generalize beyond specific numbers, but it may fail to sup-
port students’ understanding of the behavior of quadratic growth, what the second
differences can represent, and why they remain constant for quadratic functions.
Justification #2: One group of eighth-grade students created and analyzed tables
of quadratic data by exploring the relationships between the lengths, heights, and
areas of rectangles that grew while maintaining their length/height ratios. One stu-
dent attempted a justification by imagining an H × L rectangle that grew in discrete
increments by increasing H units in height and L units in length. The student con-
ceptualized the first differences, which he called the rate of growth (RoG), as the
growth of the area when the height increased by H units. He conceptualized the
second differences as the difference in the rate of growth (DiRoG), describing it as
the “rate that the increase in the area is increasing:”
Fig. 2 Eighth-grade
student’s justification
The student reasoned with the relationships between the quantities height, length,
and area, engaging in quantitative operations as he compared their differences. He
concluded that because he could calculate the difference in the rate of growth as
2H L each time the rectangle grew an additional H units in height and L units in
length, the second differences must represent twice the original area of the rectan-
gle.
The student’s justification contains some limitations, particularly because his
drawing only addresses a particular type of growth in which the height and length
increase by whole-unit increments of H and L. However, even though the student
did not reason about arbitrary increases of H and L, his justification represents
a meaningful attempt at a generalized argument. The student’s reliance on the re-
lationships between the quantities height, length, and area helped him develop an
218 A.B. Ellis
understanding of what the second differences represented, which provided a spring-

board for further investigation of why the second differences in well-ordered tables
are always constant for quadratic functions.
Steffe and Izsak (2002) argue that quantitative reasoning should be the basis for
algebraic reasoning. Focusing on relationships between quantities, rather than on
numbers disconnected from meaningful referents, can ground the study of algebra
in people’s conceptions of their experiential worlds (Chazan 2000). This provides a
meaningful starting point for mathematical inquiry, in contrast to taking numbers,
shapes, and relationships as givens in their own right (Thompson 1994). I propose
that adopting a quantitative reasoning approach can support students’ meaningful
engagement with algebra in general and with functions in particular. I will present
the results from two teaching experiments with middle-school students, the linear
functions teaching experiment and the quadratic functions teaching experiment. Ex-
cerpts from both teaching experiments demonstrate a number of ways in which stu-
dents’ reasoning with quantities fostered particular types of function understand-
ing.
The Importance of (and Difficulties with) Functional Thinking
The function is a central concept around which school algebra can be meaning-
fully organized (Kieran 1996; Yerushalmy 2000), and many researchers have argued
for the importance of a functional perspective in contrast to the more traditional
approach that focuses on algebra as symbolic manipulation (Bednarz et al. 1996;
Schliemann et al. 2007). Adopting an approach that places functional relationships
at the center of algebra allows us to couch algebraic thinking as the use of a vari-
ety of representations in order to make sense of quantitative situations relationally
(Kieran 1996). Beyond theoretical considerations, there are also practical reasons
for emphasizing a functional approach to algebra. Attaining a deep understanding
of function is critical for success in future mathematics courses (Carlson et al. 2003;
Romberg et al. 1993) and in courses on scientific inquiry (Farenga and Ness
2005). Many have argued that the function concept is foundational for under-
standing concepts in advanced mathematics (e.g. Kaput 1992; Rasmussen 2000;
Thompson 1994; Zandieh 2000), and as Romberg et al. (1993) argued, “there is
general consensus that functions are among the most powerful and useful notions in
all mathematics” (p. 1).
Given the widespread agreement on the importance of functions for algebraic
reasoning, the value of organizing algebra content around a functions approach, and
the need for a deep understanding of functions for further mathematical and scien-
tific inquiry, it is important that we develop ways of helping students successfully
understand functional relationships. However, these endeavors have proved diffi-
cult: Many studies conducted to investigate students’ function understanding sug-
gest that they demonstrate a limited view of the function concept (e.g. Carlson 1998;
Sfard and Linchevski 1994; Thompson 1994; Vinner and Dreyfus 1989). In general,
students emerge from middle school and high school algebra classes with a weak un-
derstanding of function (Carlson et al. 2002; Cooney and Wilson 1996; Monk 1992;
Monk and Nemirovsky 1994).
Two examples from my previous studies illustrate some of the common difficul-
ties students experience when encountering functional relationships. The first comes
from a problem presenting a direct-ratio situation within the context of a linear func-
tions unit (Ellis 2009):
Problem 2 Say you have a pile with 2 rolls of pennies and a pile with 5 rolls of
pennies. If you were to compare their weights, what might you notice?
Fig. 3 Picture accompanying

the penny-role problem
One eighth-grade student, Juanita, made both additive and multiplicative com-
parisons across the two piles, noting that the bigger pile had 3 more rolls, and would
weigh “2.5 times as much” as the smaller pile. When she investigated the pattern in
a tabular form, however, Juanita was unable to recognize the relationship as linear
and she could not develop an equation for the data:
Fig. 4 Table of number of # of Rolls | Weight

rolls and weight values |
2 | 9 oz
|
5 | 22.5 oz
|
12 | 54 oz
|
16 | 72 oz
AE: What does this table tell you?

J: It couldn’t be a straight line.
AE: How come?
J: (Calculating differences between successive x-values in the table): 3, 7, 4, and
that’s probably not an even spaced one. Wouldn’t be a straight line.
AE: And what’s your reason for that?
J: If you made your graph, it doesn’t look like it’d be a straight line because it
goes up (calculates differences between successive y-values in the table) by
13.5 and 31.5 and then 18.
Juanita created a rough sketch of the data to confirm her belief that the data were
not linear:
220 A.B. Ellis
Fig. 5 Juanita’s graph of the

rolls and their weight
Determined to find a pattern for the data in order to come up with an equation,
Juanita continued to search by taking the differences between the rolls and the
weight for each table entry, and then taking the differences of those results:
J: 7, 17.5, 42, 56. That’s what it goes up by. If you do in between them it’s 10.5,
24.5, and then 14. . . . There’s no patterns anywhere!
Juanita’s difficulty in recognizing the data as linear and her inability to create
an equation mirrors some of the documented difficulties students experience with
tables, patterns, and functions. Although students are adept at searching out pat-
terns in tabular representations, many struggle to perceive a functional relation-
ship (MacGregor and Stacey 1993; Mason 1996; Schliemann et al. 2001). Even
when students are able to detect patterns, they may not be able to formalize those
patterns correctly by writing appropriate equations or algebraic expressions (En-
glish and Warren 1995; Orton and Orton 1994; Stacey and MacGregor 1997).
Students struggle to correctly translate between tabular, graphical, and algebraic
representations of functional relationships, and can become overly dependent on
particular artifacts of representations, such as only recognizing a function as lin-
ear if its tabular representation has uniformly-increasing x-values (Lobato et al.
2003).
A second example illustrates some of the difficulties students can encounter when
approaching non-linear functions. High-school algebra II students encountered the
following graph and attempted to find an equation for the parabola (Ellis and Grin-
stead 2008):
Problem 3 Ravi has 120 meters of fence to make his rectangular rabbit pen. He
wants to enclose the largest possible area. Here is Ravi’s graph of the relationship
between the width and the area:
Fig. 6 Graph for the rabbit pen problem
Alexis, a tenth-grade student identified by her teacher as a high performer, deter-

mined that the equation for the parabola should have the form y = −ax 2 + 900,
because the maximum value of the parabola was at y = 900. In addition, Alexis
knew that the a-value should be negative, because the parabola was “upside down.”
In order to determine the value of “a”, Alexis explained, “you could do this, rise
over run.” She picked two points, (10, 500) and (20, 800), and then calculated the
rise and the run, ignoring the scales on the axes: “So it’s 3 over 1, which is ba-
sically 3.” Alexis concluded that the equation of the parabola should therefore be
y = −3x 2 + 900.
Alexis’ treatment of the graph and development of an equation reflects the

research demonstrating students’ difficulties understanding the value of “a” in
y = ax 2 + bx + c (Dreyfus and Halevi 1991; Zaslavsky 1997). The challenges
in connecting algebraic and graphical representations of quadratic functions can
further contribute to students’ struggles to describe the effects that changing the
parameters a, b, and c have on graphs of parabolas (Bussi and Mariotti 1999;
Leinhardt et al. 1990; Zazkis et al. 2003). In addition, Alexis’ inappropriate adop-
tion of the rise over run method for generating a “slope” mirrors many stu-
dents’ tendencies to generalize from linearity, regardless of the appropriateness
222 A.B. Ellis
of that generalization (Buck 1995; Chazan 2006; Schwarz and Hershkowitz 1999;
Zaslavsky 1997).
Given the widespread difficulties students experience as they learn about func-
tions, it is important to develop methods for helping students build a productive
understanding of functional relationships from the time that they first experience
them in the algebra classroom. Taking a quantities-based approach to informal (and
later formal) functional reasoning can support students’ initial approaches to func-
tional relationships as they explore coordinated changes between covarying quanti-
ties.
An Alternative Approach to Function: Quantities and

Covariation
Traditional approaches to function rely on a correspondence or stasis view (Smith

2003), in which one approaches a function as the fixed relationship between the
members of two sets. Farenga and Ness (2005) offer a typical correspondence def-
inition of function: “One quantity, y, is a function of another, x, if each value of x
has a unique value of y associated with it. We write this as y = f (x), where f is
the name of the function” (p. 62). This static view underlies much of school math-
ematics, particularly in the treatment of functions. Alexis’ approach reflects this
typical school experience, as she examined the graph and then attempted to build an
equation without imagining the two quantities changing together. Instead, Alexis’
treatment disconnected the properties of the graph and its associated equation from
the contextual situation that referenced the changing relationship between width and
area.
In contrast, Smith and Confrey (Smith 2003; Smith and Confrey 1994) de-
scribe the covariation approach to functional thinking. Under this approach, one
examines a function in terms of a coordinated change of x- and y-values. Con-
frey and Smith (1992, 1994, 1995) have found that students’ initial entry into a
problem is typically from the covariational perspective. In addition, they argue that
viewing a function as a way of representing the variation of quantities can be a
more powerful approach than the correspondence model, particularly in its abil-
ity to promote thinking about functions in terms of rates of change (Slavit 1997;
Smith and Confrey 1994). As Chazan (2000) argues, the covariation approach can
support a view of mathematics as a way of making sense of the phenomena of rela-
tionships of dependence, causation, interaction, and correlation between quantities.
Viewing a function as a relationship between covarying quantities is part of a
larger idea that acknowledges the importance of the mathematics of change. An
emphasis on the mathematics of change can encourage students to examine patterns
in relationship to the ways in which they grow or can be extended. Many have
suggested that this approach is a critical but overlooked element in the standard
U.S. curriculum (Nemirovsky et al. 1993; Mokros et al. 1995). Exploring function
as a way to measure change and variation is typically reserved for calculus, thus
effectively restricting access to these ideas to the 10% of students who will reach the
highest level of high-school mathematics (Roschelle et al. 2000). However, adopting
a rate-of-change perspective can be accessible even for beginning algebra students
in middle school. One way to foster students’ understanding of the mathematics of
change is through introducing rich situations that encourage students to construct
meaningful relationships between quantities.
For instance, one group of seventh-graders in a linear functions teaching exper-
iment explored constant rates of change by investigating two situations, gear ratios
and constant speed (Ellis 2007). The group consisted of 7 pre-algebra students who
had not yet studied linear functions or graphs in their mathematics classroom, and a
focus of the teaching experiment was to emphasize the activities of generalizing and
justifying through meaningful engagement with quantitative referents. The students
met for 15 sessions and during the first eight sessions they worked with physical
gears to examine different gear ratios. Early in the sessions, the students connected
a gear with 8 teeth to a gear with 12 teeth and then spun the gears together, trying to
identify ways to simultaneously keep track of the rotations of both gears. By putting
small pieces of masking tape on one of the teeth of each of the gears, the students
devised a counting system for keeping track of both gears’ rotations simultaneously,
and ultimately created tables of gear rotation pairs such as the following:
Fig. 7 Maria’s table of gear

rotations
By working with the physical gears, the students not only found ways to coordi-
nate the rotations of each of the gears, but also developed a covariation language for
discussing the nature of the coordinated quantities. For instance, in describing the
table in Fig. 7, Dora explained, “For a small turn, the big one goes a two-thirds turn.
For the big to turn once, the small one goes one and a half turn.”
Carlson and Oehrtman (2005) note that students need to be able to imagine how
one variable changes while imagining changes in the other. Relying on situations
that involve quantities that students can make sense of, manipulate, experiment with,
and investigate can foster their abilities to reason flexibly about dynamically chang-
ing events. These experiences were helpful when the linear functions students even-
tually encountered tables of data referencing multiple rotation pairs, such as the one
shown in Problem 4:
224 A.B. Ellis
Problem 4 The following table contains pairs of rotations for a small and a big
gear. Did all of these entries come from the same gear pair, or did some of them
come from different gears altogether? How can you tell?
Fig. 8 Table of gear pairs Small | Big

|
7 1/2 | 5
|
27 | 18
|
4 1/2 | 3
|
16 | 10 2/3
|
1/10 | 1/15
Dora explained her thinking about the problem by referencing the gears:
D: Think of a gear. When you spin it, the teeth on it pass through. One gear has 8
teeth, the other has 12. When you spin them, teeth pass through each other. For
every two-thirds of the teeth passed on the big one, that’s 8 teeth, so the small
one turns once. If the small one goes 3 turns, the big one will go 2. So if the
small one goes 7 and a half times, the big gear will go 5.
A covariation approach can also ultimately support students’ abilities to express
function relationships algebraically. After hearing Dora’s explanation, another stu-
dent, Larissa, expressed the gear ratio relationship by writing “s(2/3) = b”, which
represents the number of rotations between the small gears and the big gears. Larissa
explained, “s is the number of small rotations, the number of rotations that the small
gear does. And then b is the big rotations, the number of rotations that the big gear
makes.”
Carlson and Oehrtman identified a covariation framework (2005), in which they
decompose covariational reasoning into five mental actions. This decomposition
has proved useful for promoting covariational reasoning in students. Although the
framework evolved in the context of calculus students’ reasoning, the first four men-
tal actions described can also apply to algebra students. (The fifth mental action is
the coordination of instantaneous rate of change, which is not as applicable to be-
ginning algebra topics.) The gear rotation situation supported students’ abilities to
coordinate the change of both quantities simultaneously, fostering the first three
mental actions in the covariation framework: (1) coordinate the dependence of one
variable on another variable, (2) coordinate the direction of change of one variable
with changes in the other variable, and (3) coordinate the amount of change of one
variable with changes in the other variable. Because quantitative reasoning requires
the formation of relationships between quantities, students’ activity in constructing
these relationships can support the meaningful coordination of variables in function
relationships.
The fourth mental action is the coordination of the average rate-of-change of the
function with uniform increments of change in the input variable. Exploring phe-
nomena that are linearly related but not in direct proportion can prompt a shift from
direct multiplicative comparisons to the creation of ratios of change between co-

ordinated variables. In the gear context, the students examined scenarios in which
one gear spun a certain number of times on its own before a second gear was con-
nected to it, at which point they spun together. Although the situation is somewhat
contrived from an adult perspective, it was meaningful to students because it de-
scribed a familiar situation that they could directly imagine. The following table can
encourage students to coordinate the rates of change of each of the variables, both
because it is not well ordered and because it represents a situation that is not directly
proportional (the function described by the ordered pairs is y = (3/4)x + 5):
Problem 5 The following table contains pairs of rotations for a big and a small
gear. What is the relationship between the two gears?
Fig. 9 Table of gear pairs Small | Big

representing a y = mx + b |
1 | 5 3/4
situation |
4 | 8
|
12 | 14
|
25 | 23.75
One student, Timothy, identified the differences between successive table entries:
Fig. 10 Timothy’s
calculations with the gear pair
table
He explained, “The only thing I found out is that they go up by 3/4, because if you
subtract 1 from 4 and 5 and 3/4 from 8, you get 3 and 2.25, and 2.25 over 3 equals
3/4. And that’s how I found out that it works for all of them.” Pushed to explain why
this worked, Timothy said, “B goes up by 3/4 of what A goes up by.” When asked
to describe what was happening with the gears rotating, he noted, “B had already
turned 5 times. And B is like 3/4 the size of A. And so A times 3/4 means that it
only goes through 3/4 of its teeth.” When Timothy noted that B was 3/4 the size of
A, he spoke of the gear’s size but appeared to be thinking about the gear’s rotations
instead; this is consistent with the second half of his statement in which Timothy
said that B would only go through 3/4 of its teeth. Dora and several other students
expressed this relationship algebraically by writing “(3/4)a + 5 = b”, and could
explain each part of the equation in terms of the relationship between the gears’ ro-
tations and number of teeth. Ultimately the students were able to approach new data
by calculating the ratio of the change of one variable to the coordinated change in
the other variable in order to determine the appropriate relationship between mys-
tery gears.
Students’ first approaches to function are typically covariational in nature (Con-
frey and Smith 1992, 1994, 1995), but it is important to support these initial forays
226 A.B. Ellis
in a manner that supports a meaningful understanding of covarying phenomena, in

contrast to the common tendency to engage in recursive pattern seeking with naked
numbers. Although students are adept at creating multiple patterns, they can strug-
gle to identify patterns that are algebraically useful and generalizable. Embedding
these patterns in meaningful problem situations that require students to identify re-
lationships between covarying quantities can help circumvent the common pattern-
seeking traps that sometimes plague students. Quantity-based problem situations
can instead “serve as the true source and ground for the development of algebraic
methods” (Smith and Thompson 2007, pp. 96–97).
A Flexible Understanding of Functions

Coordinating Covariation and Correspondence Approaches
The prevalence of covariation approaches has been highlighted in the research lit-
erature, and this view provides a powerful mechanism for developing an under-
standing of function as a way of representing variation in coordinated quantities.
However, any complete understanding of functional relationships must ultimately
include a broader exploration of the relationships between two variables (Carraher
and Schliemann 2002). Carlson and Oehrtman (2005) argue that students must be
able to understand multiple views of function for success in mathematics: they must
develop an understanding of function as a process that accepts inputs and produces
outputs, as well as attend to the changing value of output and rate of change as the
independent variable is varied.
The shift from a covariation approach to the correspondence view can be diffi-
cult for students, but there is evidence that when working directly with quantities,
even young children can develop a flexible function understanding (e.g., Nunes et al.
1993; Schliemann et al. 1998, 2003). Working directly with accessible quantitative
relationships can aid in beginning algebra students’ investigations of functions from
multiple perspectives, as well as support their abilities to shift flexibly across dif-
ferent perspectives. The seventh-grade students’ experiences with gear ratios (and
later constant-speed situations) helped them create algebraic representations such
as “(3/4)a + 5 = b” that they could ultimately view in terms of both coordinated
changes in each gear and as a direct relationships between a and b. These experi-
ences helped the teaching-experiment students make meaningful sense of the pen-
nies problem (Fig. 3) that had caused such difficulty for Juanita, who was not in the
teaching experiment. Timothy’s response was typical of the teaching-experiment
students:
T: [Examining the table in Fig. 4]: Well, let’s see. 2 to 9 oz, so that’s 4.5 oz per
roll. For that. So multiply that by 5. Times 5, equals 22.5. So these (the first
two pairs in the table) are both from the same roll. Then multiply it by 12. 4.5
times 12 equals 54, so that’s from the same one. And then 16 times 4.5. 16
times 4.5 equals 72, so they’re all from the same thing because they all have
the same weight for 1 roll.
AE: Do you think the graph is going to be linear or non-linear?

T: It’s all going to go on the same line.
AE: Why do you think that would happen?
T: Because whatever the weight is, you can multiply it by 1 over 4.5 to get the
number of rolls.
Timothy’s reliance on his understanding of the relationship between the number
of rolls and the total weight in ounces supported a direct comparison across the x-
and y-columns of the table. He noted that whatever the weight is (the input variable),
you can multiply it by 1 over 4.5 to get the number of rolls (the output variable); even
though this is the reverse of how we might typically approach a table from an input-
output perspective, it is correct and enabled Timothy to successfully solve a number
of extrapolation and interpretation problems. Moreover, Timothy could move flex-
ibly between the correspondence and covariation approaches, as evidenced by his
predictions about the table’s graph: “It just looks like to me that all you’re doing is
going up by 4.5 oz and 1 roll. . . it’s going up by the exact same thing every time.”
Reasoning with quantitative relationships can support students’ flexible move-
ment between different function approaches for quadratic functions as well. In the
quadratic functions teaching experiment (consisting of 15 sessions with 7 eighth-
grade students), I introduced quadratic phenomena in terms of the relationships be-
tween the lengths, heights, and areas of rectangles that grew while maintaining their
length/height ratios. Although none of the students had yet experienced quadratic
functions in their normal classrooms, they had all experienced other functional re-
lationships and graphs (such as linear functions) in their algebra or pre-algebra
courses. The students worked with a script in Geometer’s Sketchpad to explore what
happened to the dimensions of a particular rectangle (for instance, a 3 cm by 2 cm
rectangle) as it grew and shrank. As predicted, the students made sense of these
phenomena from a covariation perspective, imagining what would happen to the
area as the length (or width) increased by a uniform amount. The students created
their own tables of data to represent the phenomena they observed; a typical table is
shown below:
Fig. 11 Student’s table of

data representing the growing
rectangle
228 A.B. Ellis
In this case, the student was able to coordinate the growth of the length and the area
of the rectangle as the width grew in 1-cm increments. He also identified the amount
by which the area increased for each additional centimeter in width, as well as their
differences.
I introduced a standard far-prediction problem to encourage a shift from the co-
variation approach to the development of a direct functional relationship between
height and area:
Problem 6 Here is a table for the height versus the area of a rectangle that is grow-
ing in proportion to itself. What will the area be when the rectangle is 82 units high?
Fig. 12 Table of height/area Height | Area

values for a growing rectangle |
2 | 18
|
3 | 40.5
|
4 | 72
|
5 | 112.5
|
6 | 162
The students’ initial entry into the problem was from a covariation perspective, in
which they coordinated the growth of three quantities: height, length, and area. Each
student introduced a third column, length, and noticed that the length increased by
4.5 units each time the height increased by 1 unit:
Fig. 13 Student’s table with

the added length column
One student, Ariel, stated that the area of a rectangle 82 units high would be 30,258
square units. Ariel explained that she found 30,258 by multiplying 82 by 4.5 units to
get the corresponding length of 369 units. The area would then be 82 units multiplied
by 369 units. Jim relied on his image of the rectangle and the way in which it grew
to explain Ariel’s reasoning to the class:
J: Well that was the length, the 369, so she has to do height times length equals
area. So she had to multiply [the 369 by 82] again.
AE: I see. So how did you know to multiply 82 by 4.5 units to get the length?
J: Because, that’s how much the length was going up by every time. So if you,
like, made a square, I mean, or a rectangle, and then you moved up 1 unit, it
would go over 4.5 for every time you go up the height 1.
At this stage, the students’ thinking relied on an image of the manner in which the
rectangle grew in order to coordinate the growth between the height and the length.
This supported their understanding from a covariational view, and they capitalized
on their understanding of the coordinated growth of the height and the length to de-
termine the area of the rectangle for a large height. However, the students’ images
of the nature of the rectangle’s growth were limited to cases in which the rectangles
grew in discrete whole-unit increments, typically increments in which the length or
the height increased by 1 unit. Simplifying the nature of the growth initially helped
the students coordinate the multiple quantities involved (length, width, area, and in-
creases in each of these quantities), but this was a strategy that would ultimately
need to be generalized to encompass the notion of non-unit increments and contin-
uous growth.
In an attempt to encourage the students to think about a direct relationship be-
tween the height and the area, I then asked them to compute the area when the height
was n units. They quickly produced the formula “area = 4.5n2 ”, and Jim explained
his reasoning to another student, Bianca:
J: I put n times 4.5 times n.
B: How did you figure it out?
J: Well, n can be any value. . .
B: Right.
J: Times 4.5 is your length. Times n again because I do height again, is your area.
Jim simply extended his previous reasoning to determine that the length of a rect-
angle n units high would be 4.5n, and thus the area must be the height times the
length, or n(4.5n).
The students continued to work with far prediction problems, and the introduc-
tion of tables that were not well ordered encouraged the students to conceptualize
the (unknown) length in terms of its relationship between the height and the area.
Unable to identify the rate of growth of the length, the students instead began to
develop third length columns by dividing the area by the height. The inclusion of
the length columns also encouraged students to make more explicit connections be-
tween the length/height ratio and the “a” in y = ax 2 , as seen below:
Fig. 14 Student’s third

length column for a table of
height/area values
230 A.B. Ellis
One student, Tai, explained, “I came up with this equation [area = 1.5h2 ]. It’s like,
the number in front of the height squared, is figured out by the area divided by the
height squared.” Daeshim added, “The number is what you have to multiply the
height by to get the length. And then height times length is the area, so that is why
it’s squared.” The norm that students must explain how their equations were related
to the quantities in the rectangle supported justifications such as Daeshim’s, and
encouraged additional connections between features of the equations (such as the
value of “a” in y = ax 2 ) and properties of the growing rectangle.
Although the shift from a covariation approach to a correspondence approach was
gradual, it was aided by the students’ abilities to make direct connections to their im-
ages of growing rectangles and their abilities to coordinate relationships between the
quantities length, width, and area. Moreover, their reliance on these constructed re-
lationships enabled the students to develop a flexible view of the quadratic function,
one in which they frequently shifted between the covariation and correspondence
views. In particular, these flexible views helped the students make connections be-
tween the value of “a” in y = ax 2 and the second differences for area, which the
students termed the “difference in the rate of growth of the area”, or the DiRoG for
area when the width increased in uniform amounts. The students created multiple
generalizations about the DiRoG of the area, including the notion that the DiRoG
(for tables in which the rectangle’s height increases by 1 unit) is twice the value of
“a” in y = ax 2 , the DiRoG is twice the area of the rectangle when the height is 1
unit, and the DiRoG is the value of the rectangle’s length when the height is 1 unit.
The students experienced little difficulty when they transitioned to tables that
were not well ordered for a number of reasons. First, they were accustomed to pic-
turing the rectangle that was represented in the table’s values, so every pattern they
developed was solidly grounded in the imagery of length, height, and area and their
relationships. This imagery supported the students’ abilities to create functional re-
lationships between height and area. In addition, the students had spent so much
time focusing on what the DiRoG meant for the rectangle’s area in relationship to
the equations they built, they became accustomed to moving seamlessly between re-
cursive and functional representations. Because they kept discussing what the values
represented in terms of length and area, the students were encouraged to represent
those relationships more generally in algebraic forms.
Flexibility Across Forms
Smith and Thompson (2007) remind us that one role of quantitative reasoning is to
support thinking that is flexible and general in character. Students in the linear and
quadratic functions teaching experiments created many tables and algebraic repre-
sentations to describe the same phenomena, and could move between them. But
what about graphical representations? In both cases, I deliberately refrained from
introducing graphs until the students had developed a meaningful understanding of
the relationships represented by the graphs. Once that foundation was in place, they
began to create their own graphs as a way to justify their conclusions about the
quantitative relationships they developed.
For instance, the linear functions students encountered a scenario in which a char-
acter walked 5 cm in 4 seconds. They created multiple equivalent ratios to represent
the character’s speed, and represented these ratios in tables of data. When asked to
explain why a speed of 15 cm in 12 seconds was the same as 5 cm in 4 seconds,
Timothy asked if he could create a graph:
Fig. 15 Timothy’s graph of same-speed values
Timothy’s partner, Dora, wrote the equation y = 4/5x and explained that the x-axis
represented centimeters and the y-axis represented seconds. Timothy explained that
he could put a line through the points, if they were appropriately exact:
T: You could put a line there. But it’s not a good graph so it’s not going to make
a straight line.
AE: Okay. You found that the slope of the line was 4/5. What does that mean?
T: Whatever x is, y is 4/5 of x. The slope means that whatever x goes up by, 4/5
of that is how much y goes up by.
AE: And what does 4/5 have to do with the speed of the clown?
T: It’s going basically 4/5 of a second per centimeter.
AE: Now why is the fact that the clown’s speed is 4/5 of a second per 1 centimeter,
why is that the same as the slope being 4/5? What’s the connection?
T: Because for every centimeter it goes, it’s going like 4, er, yeah, 4/5 of a second
I think. Every centimeter goes. . . yeah. Every centimeter it’s going 4/5 of a
232 A.B. Ellis
second. The slope is 4/5 because for every centimeter that you add, you add
4/5 seconds.
Timothy’s understanding of the speed situation, his familiarity with creating
same-ratio tables, and his ease with representing these phenomena algebraically all
supported his ability to create and make quantitative sense of a linear graph. In addi-
tion, Timothy was able to imagine the scenario from a correspondence perspective
(“Whatever x is, y is 4/5 of x”) as well as from a covariation perspective (“For
every centimeter you add, you add 4/5 seconds”). Each of these views, as well as
Timothy’s flexibility with moving across views, was enabled by his understanding
of the relationship between the quantities centimeters and seconds to create the phe-
nomenon of constant speed.
The quadratic functions students began to create graphs in the third week of
the teaching experiment and ultimately graphed both y = ax 2 and y = ax 2 + c
situations. Before they produced any graphs, they made predictions about what a
graph of the growing rectangle situation might look like:
AE: If you were to graph this one [comparing the height to the area of a square],
what do you think the graph would look like?
B: A curve.
S: I thought it would be straight because every time the area’s going up by 2.
AE: So what do you think about what Sara’s saying? She’s saying every time it
would go up by 2 so it would be straight.
B: Well, the area’s going up by 2 in between every time it’s going up by a different
number, so that makes me think it’s going up in a curve because it’s, like,
staired.
When the students ultimately created graphs, they showed the first and second dif-
ferences for the area in order to connect their prior emphasis on differences to the
graphical representation, and to explain why the graph must be curved instead of
straight. For instance, Daeshim’s graph identified the constant second differences
as 1.5 cm2 when the height increased in 0.5 cm increments, and he showed this
by calculating the increase in area for each 0.5-cm increase in the height, and then
showing the difference between each successive area increase to be 1.5:
Fig. 16 Daeshim’s graph of

A = 3h2
The students’ quantitative understanding of the rectangle situation enabled them

to make accurate predictions about the nature of graphs and interpret new graphs
by thinking about the value of each point in relationship to a hypothetical rectangle.
The students correctly predicted, for instance, that the parabola for y = 5x 2 would
be narrower than the parabola for y = 0.5x 2 , because the former represented a larger
rectangle that was adding much more area with each height increase than the latter.
They also made sense of graphs with non-zero y-intercepts by imagining rectangles
with a constant number of extra square units tacked on. While students’ later forays
into features of graphs and families of functions will likely rely less on quantitative
images, reasoning directly with the quantities can provide a critical sense-making
foundation for their initial investigations of graphical representations.
In both the linear and the quadratic case, the students made use of different rep-
resentations (tabular, algebraic, and graphical) to describe and make sense of the
quantitative situations involving gear ratios, speed, or growing rectangles. Since
each representation was a way of describing the quantitative phenomena, rather than
an instructor-introduced artifact divorced from any referents, the connections across
the representations were natural ones that enabled seamless transitions. Depending
234 A.B. Ellis
on the questions at hand, the students made use of the type of representation that
they found most helpful for describing quantitative phenomena.
Fostering a Focus on Quantities

The situations with gear ratios and growing rectangles were optimal contexts for
exploring linear and quadratic functions because the phenomena were precisely,
rather than approximately, linear and quadratic. Some problem situations involve
contexts in which the data are not exact; for instance, students may gather real-
world quadratic data from rolling balls down inclined planes, or explore contrived
problems presenting supposedly linear relationships between the number of surf
boards sold and the temperature for a given day. The contrived nature of some con-
texts may interfere with students’ natural sense making, and realistic situations with
messy data may prevent students from directly manipulating quantities in order to
form the necessary conceptual relationships that embody the phenomenon in ques-
tion (Ellis 2007). While approximate or messy data are fully appropriate data to
investigate, particularly in terms of highlighting the power of mathematical models
for making sense of real-world situations, these contexts may not be ideal for middle
school students who are exploring functional relationships for the first time.
Instead students will benefit from opportunities to explore the nature of linear (or
quadratic) relationships by directly manipulating quantities: for instance, examining
how changing time or distance independently affects the emergent quantity speed,
creating two-number ratios and then iterating them and partitioning them to form
equivalent ratios, and otherwise investigating how the constituent quantities affect
the functional relationship at hand. The students in the linear and quadratic func-
tions teaching experiments had opportunities to manipulate and explore physical
artifacts (for the gears) or run experiments with computer software (for the speed
situation and the growing rectangles situation). However, even in cases in which
physical artifacts or computer simulations are not available, students can investigate
how changing a particular quantity can affect the others related to it. Teachers may
have to take care to support students’ engagement with these problems, particularly
because the tendency to extract numbers and focus on pattern-seeking activities ap-
pears to be a strong pull for middle-school students. In these cases an instructor’s
intervention can draw students’ attention back to the quantitative referents of num-
bers and patterns. For instance, if a student describes a pattern in a table such as
“each time x goes up by 4, y goes up by 5”, a teacher could ask students to describe
what this means in terms of the gears rotating.
Students’ unique interactions with and interpretations of real-world situations
remind us that these contexts are not a panacea. Introducing a quantitatively-rich
situation does not guarantee that students will build quantitative relationships; a
quantity is, after all, a person’s conception of a measurable attribute, rather than the
attribute itself. Students may focus on any number of features in a problem situation,
and this focus may not always include productive relationships between quantities.
Therefore teachers play an important role in shaping a classroom discussion, posing
appropriate questions, inserting new information, and otherwise guiding students

to develop the quantitative operations that will support the formation of functional
relationships. A common refrain in the quadratic functions teaching experiment was
“what does this mean in terms of the rectangle?” because this reminder encouraged
the students to develop pattern generalizations that were meaningfully grounded
rather than arbitrary and unproductive.
Students’ initial learning of functions is particularly critical because it sets the
foundation for future work in algebra at the high school level and beyond. Support-
ing students’ abilities to make sense of functions from a quantitatively meaningful
stance can foster a function understanding that is productive, grounded, and flexible
in nature. A focus on numbers, relationships, and functional behaviors in absence
of quantitative referents is certainly appropriate for mathematics students and, in
the long term, necessary as students explore increasingly abstract ideas. However, I
argue that for middle school students’ first introduction to functional relationships,
a grounding in quantities, relationships, and meaningful situations can ultimately
support the eventual shift to more formal algebraic practices in high school.
References
Bednarz, N., Kieran, C., & Lee, L. (1996). Approaches to Algebra: Perspectives for Research and
Teaching. Boston, MA: Kluwer Academic Publishers.
Buck, J. C. (1995). Fostering connections between classes of polynomial functions. Paper pre-
sented at the Annual Meeting of the North American Chapter of the International Group for the
Psychology of Mathematics education (17th, Columbus, OH, October, 1995).
Bussi, M. G. B., & Mariotti, M. A. (1999). Semiotic mediation: From history to the mathematics
classroom. For the Learning of Mathematics, 19(2), 27–35.
Carlson, M. (1998). A cross-sectional investigation of the development of the function concept.
Research in Collegiate Mathematics Education III, Conference Board of the Mathematical Sci-
ences, Issues in Mathematics Education, 7, 114–163.
Carlson, M., & Oehrtman, M. (2005). Key aspects of knowing and learning the concept of function.
Mathematical Association of America Research Sampler.
Carlson, M., Jacobs, S., Coe, E., Larsen, S., & Hsu, E. (2002). Applying covariational reasoning
while modeling dynamic events: A framework and a study. Journal for Research in Mathemat-
ics Education, 33, 352–378.
Carlson, M. P., Smith, N., & Persson, J. (2003). Developing and connecting calculus students’
notions of rate of change and accumulation: The fundamental theorem of calculus. In N. A.
Pateman, B. J. Dougherty, & J. T. Zilliox (Eds.), Proceedings of the Joint Meeting of PME
and PMENA (Vol. 2, pp. 165–172). Honolulu, HI: CRDG, College of Education, University of
Hawai’i.
Carraher, D. W., & Schliemann, A. D. (2002, April). Designing and implementing early alge-
bra activities: From finding unknowns to representing variables. Paper presented at the annual
meeting of the National Council of Teachers of Mathematics, Las Vegas, NV.
Chazan, D. (2000). Beyond Formulas in Mathematics and Teaching: Dynamics of the High School
Algebra Classroom. New York, NY: Teachers College Press.
Chazan, D. (2006). “What if not?” and teachers’ mathematics. In F. Rosamund & L. Copes (Eds.),
Educational Transformations: Changing Our Lives Through Mathematics; A Tribute to Stephen
Ira Brown (pp. 3–20). Bloomington, Indiana: AuthorHouse.
Confrey, J., & Smith, E. (1992). Function probe: Multi-representational software for Learning
about functions. New York State Association for Computers and Technology in Education 6
(pp. 60–64).
236 A.B. Ellis
Confrey, J., & Smith, E. (1994). Exponential functions, rates of change, and the multiplicative unit.
In P. Cobb (Ed.), Learning Mathematics (pp. 31–60). Dordrech, Netherlands: Kluwer Academic
Publishers.
Confrey, J., & Smith, E. (1995). Splitting, covariation, and their role in the development of expo-
nential function. Journal for Research in Mathematics Education, 26(1), 66–86.
Cooney, T., & Wilson, M. (1996). Teachers’ thinking about functions: Historical and research
perspectives. In T. A. Romberg & E. Fennema (Eds.), Integrating Research on the Graphical
Representation of Functions (pp. 131–158). Hillside, NJ: Lawrence Erlbaum.
Dreyfus, T., & Halevi, T. (1991). QuadFun: A case study of pupil computer interaction. Journal of
Computers in Mathematics and Science Teaching, 10(2), 43–48.
Ellis, A. B. (2007). The influence of reasoning with emergent quantities on students’ generaliza-
tions. Cognition and Instruction, 25(4), 439–478.
Ellis, A. B. (2009, April). A quantitative understanding of quadratic growth: What is DiRoG and
why does it matter? Paper presented at the Annual Meeting of the American Educational Re-
search Association, Washington, DC.
Ellis, A. B., & Grinstead, P. (2008). Hidden lessons: How a focus on slope-like properties of
quadratic functions encouraged unexpected generalizations. Journal of Mathematical Behav-
ior, 27(4), 277–296.
English, L., & Warren, E. (1995). General reasoning processes and elementary algebraic under-
standing: Implications for instruction. Focus on Learning Problems in Mathematics, 17(4),
1–19.
Farenga, S. J., & Ness, D. (2005). Algebraic thinking part II: The use of functions in scientific
inquiry. Science Scope, 29(1), 62–64.
Kaput, J. (1992). Technology and mathematics education. In D. A. Grouws (Ed.), Research on
Mathematics Teaching and Learning (pp. 515–556). New York, NY: Macmillan Publishing.
Kaput, J. (1995). Long term algebra reform: Democratizing access to big ideas. In C. Lacampagne,
W. Blair, & J. Kaput (Eds.), The Algebra Initiative Colloquium (pp. 33–52). Washington, DC:
U.S. Department of Education.
Kieran, C. (1996). The changing face of school algebra. In 8th International Congress on Mathe-
matical Education, Selected Lectures (pp. 271–286). S.A.E.M. THALES.
Leinhardt, G., Zaslavsky, O., & Stein, M. K. (1990). Functions, graphs, and graphing: Tasks, learn-
ing, and teaching. Review of Educational Research, 60(1), 1–64.
Lobato, J., Ellis, A. B., & Muñoz, R. (2003). How “focusing phenomena” in the instructional
environment afford students’ generalizations. Mathematical Thinking and Learning, 5(3), 1–
36.
MacGregor, M., & Stacey, K. (1993). Seeing a pattern and writing a rule. In I. Hirabayashi, N.
Nohda, K. Shigematsu, & F. L. Lin (Eds.), Proceedings of the 17th International Conference
for the Psychology of Mathematics Education (Vol. 1, pp. 181–188). Tsukuba, Japan: PME
Program Committee.
Mason, J. (1996). Expressing generality and roots of algebra. In N. Bednarz, C. Kieran, & L. Lee
(Eds.), Approaches to Algebra (pp. 65–86). Dordrecht, the Netherlands: Kluwer.
Mokros, J., Economopoulos, K., & Russel, S. J. (1995). Beyond Arithmetic: Changing Mathemat-
ics in the Elementary Classroom. Palo Alto, CA: Dale Seymour.
Monk, S. (1992). Students’ understanding of a function given by a physical model. In G. Harel
& E. Dubinsky (Eds.), The Concept of Function: Aspects of Epistemology and Pedagogy (pp.
175–193).
Monk, G. S., & Nemirovsky, R. (1994). The case of Dan: Student construction of a functional
situation through visual attributes. In E. Dubinsky, A. H. Shoenfield, & J. J. Kaput (Eds.), Re-
search in Collegiate Mathematics Education I (Vol. 4, pp. 139–168). Providence, RI: American
Mathematical Society.
Nemirovsky, R., Tierney, C., & Ogonowski, M. (1993). Children, additive change, and calculus
(Working Paper 2-93). Cambridge: TERC.
Nunes, T., Schliemann, A. D., & Carraher, D. W. (1993). Mathematics in the Streets and in Schools.
Cambridge, UK: Cambridge University Press.
Orton, A., & Orton, J. (1994). Students’ perception and use of pattern and generalization. In J.
P. da Ponto & J. F. Matos (Eds.), Proceedings of the 18th International Conference for the
Psychology of Mathematics Education (Vol. III, pp. 407–414). Lisbon, Portugal: PME Program
Committee.
Rasmussen, C. L. (2000). New directions in differential equations: A framework for interpreting
students’ understandings and difficulties. Journal of Mathematical Behavior, 20, 55–87.
Romberg, T. A., Carpenter, T. P., & Fennema, E. (1993). Toward a common research perspective.
In T. A. Romberg, E. Fennema, & T. P. Carpenter (Eds.), Integrating Research on the Graphical
Representation of Functions (pp. 1–9). Hillsdale, NJ: Lawrence Erlbaum.
Roschelle, J., Kaput, J., & Stroup, W. (2000). SimCalc: Accelerating students’ engagement with
the mathematics of change. In M. J. Jacobson & R. B. Kozma (Eds.), Innovations in Science
and Mathematics Education: Advanced Designs for Technologies of Learning (pp. 47–75).
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Schliemann, A. D., Araujo, C., Cassundé, M. A., Macedo, S., & Nicéas, L. (1998). Multiplica-
tive commutativity in school children and street sellers. Journal for Research in Mathematics
Education, 29(4), 422–435.
Schliemann, A. D., Carraher, D. W., & Brizuela, B. M. (2001). When tables become function
tables. In Proceedings of the XXV Conference of the International Group for the Psychology of
Mathematics Education (pp. 145–152). Utrecht, The Netherlands: Kluwer.
Schliemann, A. D., Carraher, D. W., Brizuela, B. M., Earnest, D., Goodrow, A., Lara-Roth, S., &
Peled, I. (2003). Algebra in the elementary school. In N. Pateman, B. Dougherty, & J. Zilliox
(Eds.), Proceedings of the 2003 Joint Meeting of PME and PME-NA (Vol. 4, pp. 127–134).
Honolulu, HI: CRDG, College of Education, University of Hawai’i.
Schliemann, A. D., Carraher, D. W., & Brizuela, B. (2007). Bringing Out the Algebraic Charac-
ter of Arithmetic: From Children’s Ideas to Classroom Practice. Mahwah: Lawrence Erlbaum
Associates.
Schwarz, B. B., & Hershkowitz, R. (1999). Prototypes: Brakes or levers in learning the function
concept? The role of computer tools. Journal for Research in Mathematics Education, 30(4),
362–390.
Sfard, A., & Linchevski, L. (1994). The gains and the pitfalls of reification: The case of algebra.
Slavit, D. (1997). An alternative route to the reification of functions. Educational Studies in Math-
ematics, 33, 259–281.
Smith, E. (2003). Stasis and change: Integrating patterns, functions, and algebra throughout the
K-12 curriculum. In J. Kilpatrick, W. G. Martin, & D. Schifter (Eds.), A Research Companion
to Principles and Standards for School Mathematics (pp. 136–150).
Smith, E., & Confrey, J. (1994). Multiplicative structures and the development of logarithms: What
was lost by the invention of function. In G. Harel & J. Confrey (Eds.), The Development of
Multiplicative Reasoning in the Learning of Mathematics (pp. 333–364). Albany, NY: State
University of New York Press.
Smith, J., & Thompson, P. (2007). Quantitative reasoning and the development of algebraic rea-
soning. In J. J. Kaput, D. W. Carraher, & M. L. Blanton (Eds.), Algebra in the Early Grades
(pp. 133–160). New York: Erlbaum.
Stacey, K., & MacGregor, M. (1997). Building foundations for algebra. Mathematics Teaching in
the Middle School, 2(4), 253–260.
Steffe, L., & Izsak, A. (2002). Pre-service middle-school teachers’ construction of linear equation
concepts through quantitative reasoning. In D. Mewborn, P. Sztajn, D. White, H. Wiegel, R.
Bryant, & K. Noony (Eds.), Proceedings of the Twenty-Fourth Annual Meeting of the North
American Chapter of the International Group for the Psychology of Mathematics Education
(Vol. 4, pp. 1163–1172). Columbus, OH: ERIC Clearinghouse for Science, Mathematics, and
Environmental Education.
238 A.B. Ellis
Thompson, P. W. (1994). The development of the concept of speed and its relationship to concepts
of rate. In G. Harel & J. Confrey (Eds.), The Development of Multiplicative Reasoning in the
Learning of Mathematics (pp. 181–234). Albany, NY: SUNY Press.
Vinner, S., & Dreyfus, T. (1989). Images and definitions for the concept of function. Journal for
Research in Mathematics Education, 20(4), 356–366.
Yerushalmy, M. (2000). Problem solving strategies and mathematical resources: A longitudinal
view on problem solving in a function based approach to algebra. Educational Studies in Math-
ematics, 43(2), 125–147.
Zandieh, M. J. (2000). A theoretical framework for analyzing student understanding of the concept
of derivative. In E. Dubinsky, A. H. Schoenfeld, & J. J. Kaput (Eds.), CBMS Issues in Mathe-
matics: Research in Collegiate Mathematics Education IV (Vol. 8, pp. 103–127). Providence:
American Mathematical Society.
Zaslavsky, O. (1997). Conceptual obstacles in the Learning of quadratic functions. Focus on Learn-
ing Problems in Mathematics, 19(1), 20–44.
Zazkis, R., Liljedahl, P., & Gadowsky, K. (2003). Conceptions of function translation: Obstacles,
intuitions, and rerouting. Journal of Mathematical Behavior, 22(4), 435–448.
Representational Competence and Algebraic
Modeling
Andrew Izsák
Abstract This chapter reviews some key empirical results and theoretical perspec-
tives found in the past three decades of research on students’ capacities to reason
with algebraic and graphical representations of functions. It then discusses two re-
cent advances in our understanding of students’ developing capacities to use in-
scriptions for representing situations and solving problems. The first advance is the
insight that students have criteria that they use for evaluating external representa-
tions commonly found in algebra, such as algebraic and graphical representations.
Such criteria are important because they play a central role in learning. The sec-
ond advance has to do with recognizing the importance of adaptive interpretation,
which refers to ways in which students must coordinate shifts in their perspective
on external representations with corresponding shifts in their perspective on prob-
lem situations. The term adaptive highlights the context sensitive ways in which
students must learn to interpret external representations. The chapter concludes
with implications of these two advances for future research and algebra instruc-
tion.
Gaining insight into how students learn to reason with external representations, or
inscriptions, has been a central challenge in mathematics education research for
several decades. Research on algebra and functions has grappled with this challenge
extensively, perhaps more so than research on the teaching and learning of any other
The research reported in this chapter was supported by a Spencer Foundation Dissertation
Fellowship and by the National Science Foundation under Grant No. 9554564 and Grant No.
REC 0231879. The opinions expressed in this chapter are those of the author and do not
necessarily reflect the views of the Spencer Foundation or the National Science Foundation. This
chapter is a revised version of an article published in ZDM—International Reviews on
Mathematical Education, 37(1), 60–67. DOI 10.1007/BF2655898.
A. Izsák ()
Department of Mathematics and Science Education, University of Georgia, Athens, GA, USA

240 A. Izsák
mathematical content area. The importance of external representations to mathemat-

ical thinking, and to algebraic thinking in particular, has been highlighted in recent
influential documents. The National Council of Teachers of Mathematics’ Princi-
ples and Standards for School Mathematics (NCTM 2000) named representation as
a process standard, and the National Research Council’s Adding it Up (Kilpatrick et
al. 2001) characterized algebra in terms of three types of activity, one of which is
representational activity. More recently, Kieran (2007) included examples of reason-
ing with external representations when elaborating a framework that conceptualizes
algebraic activity in terms of generational activity, transformational activity, and
global/meta-level activity.
Modeling problem situations in the physical world is one important context for
generating and interpreting external representations. Furthermore, understanding
how students learn to represent situations and solve problems about those situa-
tions continues to be of practical and theoretical importance. From a practical point
of view, more traditional instructional materials often concentrate on procedures
for using prescribed representations—for instance, procedures for simplifying alge-
braic expressions or for plotting points on a Cartesian graph. In recent years, how-
ever, new reform-oriented instructional materials have been making their way into
schools. For instance, in the United States the National Science Foundation has sup-
ported the development of materials (e.g., Coxford et al. 1998; Lappan et al. 2002,
2006; TERC 2008; Wisconsin Center for Educational Research & Freudenthal In-
stitute 2006) that respond explicitly to standards developed by the National Council
of Teachers of Mathematics (NCTM; 1989, 2000). These newer instructional mate-
rials for algebra (and other mathematical content areas) are increasing the demands
placed on students to reason with various forms of representation. To illustrate, ask-
ing students to identify the advantages and disadvantages of tabular, graphical, and
algebraic representations is an example of a task that is not often found in more
traditional materials and that is more demanding than simplifying expressions or
plotting points. From a theoretical point of view, mathematics education researchers
are continuing to develop empirically grounded theory that provides new insight
into students’ cognition when generating, interpreting, and using external represen-
tations to solve problems about situations.
The goal of this chapter is to highlight two recent advances in our understand-
ing of students’ developing capacities for reasoning with external representations
of problem situations. Both advances are built on results from several studies that
sought insight into students’ internally held mental structures and processes. One
advance has to do with students’ capacities to generate external representations, and
the other has to do with students’ capacities interpret external representations. To
establish the context for these advances, I first trace some key empirical results and
theoretical perspectives found in the past three decades of research on students’
capacities to generate and interpret external representations commonly used in alge-
bra. (For comprehensive reviews of research on the teaching and learning of algebra,
see Kieran 1992, 2007.) I then discuss the two advances and examine their implica-
tions for future research and teaching.
Representational Competence and Algebraic Modeling 241
Early Results on Students’ Understandings of Standard

Representations in Algebra
During the 70s and 80s, researchers reported numerous difficulties that students en-
counter generating normative algebraic representations to model problem situations.
In one well-known study, Clement (1982) found that in a sample of science oriented
college students 37% failed to correctly answer the following problem, “Write an
equation using the variables S and P to represent the following statement: ‘There
are six times as many students as professors at this university’ ” (p. 17). In a sec-
ond example, Booth (1981) found that secondary-school students wrote expressions
such as ‘hhhht’ and ‘4ht’ to symbolize the perimeter of a pentagon in which four
sides were labeled ‘h’ and one side was labeled ‘t.’
Research conducted during these same years also uncovered a variety of diffi-
culties that students encounter interpreting algebraic notation appropriately. These
difficulties include interpreting letters (e.g., Harper 1987; Küchemann 1981) and the
equal sign (e.g., Kieran 1981). Küchemann analyzed written responses from approx-
imately 1000 14-year-old students to a set of test questions and found that students
either avoided working with letters altogether or thought of letters as standing for
specific unknown values. Rarely did students think of letters as standing for vari-
ables. Kieran (1981) reported that elementary and middle school students tend to
interpret the equal sign not as a relation between two expressions but rather as a
signal to compute the expression on the left-hand side and record the result on the
right-hand side.
Further research from this same era uncovered difficulties that students encounter
when generating and interpreting normative graphical representations of problem
situations (see Leinhardt et al. 1990, for a thorough review). In one oft cited study,
Kerslake (1981) analyzed written responses of nearly 1800 13-, 14-, and 15-year-
old students to a set of test questions. Among other things, she reported that many
students do not understand when it is and when is not appropriate to connect points,
indicating difficulties generating graphs. She also reported that students often mis-
interpret graphs as direct depictions of problem situations, a phenomenon referred
to as the graph-as-picture interpretation. To illustrate, when considering the motion
of an object, students interpret distance vs. time graphs as showing a trace of the
actual path of motion.
On the whole, research from this era emphasized constraints on what students
could do. Researchers prescribed particular normative external representations (e.g.,
conventional equations or graphs) with which students were to accomplish tasks
and reported errors that students made. Oftentimes, researchers characterized these
errors as misconceptions.
Theoretical Accounts of Reasoning with External

Representations
In addition to uncovering students’ numerous difficulties generating and interpret-
ing normative external representations used in algebra, researchers in mathematics
242 A. Izsák
education also sought theoretical accounts that could explain reasoning with exter-
nal representations (see Janvier 1987a, for one collection of theoretical perspectives
from this era). At least two developments in the field shaped these accounts signif-
icantly. The first development was the increasingly broad acceptance of construc-
tivist epistemology and tensions between that epistemology and statements such
as “X represents Y .” Although from a constructivist perspective Y does not exist
in any absolute sense, one can talk about individuals “re-presenting” their expe-
riences. As researchers began framing accounts in terms of interactions between
internal representations (mental structures of individuals) and external representa-
tions such as equations and graphs, they used the terms notations and inscriptions
to emphasize that an artifact, on its own, does not carry meaning. Rather, meaning
emerges when individuals generate and interpret tables, equations, graphs, and other
diagrams or notations. The second development was the increasing availability of
computers and software that linked tabular, algebraic, and graphical representations
of functions to each other (see Romberg et al. 1993, especially Chaps. 2–4) and that
linked graphs to physical phenomena such as motion (e.g., Mokros and Tinker 1987;
Monk and Nemirovsky 1994; Nemirovsky 1994). Researchers were interested in
theory that could explain students’ reasoning in these new environments.
Initial attempts within mathematics education to theorize about cognition around
forms of external representation, or inscriptions, described cognitive processes at
a coarse grain-size. One example that illustrates this grain-size is translation, the
process of moving among different external representations of the same situation
(e.g., Janvier 1987b; Lesh et al. 1987). A person might translate between verbal and
symbolic representations or between algebraic and graphical representations.
Kaput (1987, 1989, 1991) developed a perspective on reasoning with external
representations that underscores the distinction between internal and external rep-
resentations, is consistent with constructivist epistemology, and resolves coarse-
grained cognitive processes, like translation, into several components. He began by
defining a symbol system as a symbol scheme combined with a field of reference.
Briefly, a symbol scheme is a set of symbols and a set of rules for transforming those
symbols. Symbols for the algebra symbol scheme include numbers, letters, notations
for arithmetic operations, and the equal sign. Transformations include rules for ma-
nipulating the symbols in the scheme—for instance, rules for adding like terms and
for multiplying two binomials. A field of reference can be a problem situation in the
physical world or another symbol scheme.
Kaput (1987, 1991) then identified two types of cognitive activity associated with
symbol systems. The first type consists of encoding and reading, which correspond
to processes of generating and interpreting external representations. Encoding1 oc-
curs when “one has some conceptual structure or operation that one seeks to exter-
nalize for purposes of communication or testing” (Kaput 1991, p. 57). For instance,
one might generate an algebraic expression or equation to express understanding of
a problem situation. Reading occurs when “processes are based on an intent to use
1 Some researchers use the term encoding to describe internal representation of external stimuli.
Kaput’s definition is the reverse, and I use his sense here.
the physical material to assist one’s conceptual activity in traditional acts of reading
and interpreting” (ibid, p. 57). For instance, one might interpret an already written
expression or graph. The second type of cognitive activity has to do with how one
uses symbol systems. Syntactic elaboration refers to manipulating symbols using
the transformation rules. Semantic elaboration refers to elaborating the referents for
the symbols (see Kaput 1987, p. 177).
Kaput emphasized that the cognitive activities of encoding, reading, syntactic
elaboration, and semantic elaboration function in interaction with one another and
through cycles of reasoning. When compared to constructs like translation, his con-
structs were more fine-grained and pointed toward more nuanced analyses of the
complexities of generating and interpreting external representations to solve prob-
lems and of making connections between one representation and another. Results
presented in the next section have begun to shed light on those complexities.
Students’ Capacities to Reason with External Representations

Constructivist epistemology suggested particular limitations of the early empirical
results summarized above on students’ difficulties generating and interpreting nor-
mative algebraic and graphical representations. One limitation was that by focusing
on students’ errors, this generation of research did not examine how students might
use their existing knowledge productively to accomplish task. As a consequence, a
second limitation was that this generation of research provided few clues about the
transformation of more novice into more expert knowledge (see Smith et al. 1993,
for a constructivist critique of misconceptions research).
Several new lines of research have examined students’ capacities to use in-
scriptions productively and to construct knowledge about external representations.
One line has examined more closely the cognitive structures involved in build-
ing connections between different representations of functions (e.g., Knuth 2000;
Moschkovich 1998; Schoenfeld et al. 1993). Schoenfeld et al. and Moschkovich
have described a 3-slot schema to explain how for some students slopes, y-
intercepts, and x-intercepts of linear functions are salient features of graphs that
should all appear explicitly in equations. These researchers have described growth
and change of the multi-layered schema in terms of coordination, reorganization,
and refinement of multiple elements. The results underscore the limitations of
coarse-grained cognitive processes, such as translation, for explaining challenges
that students experience when connecting multiple representations.
A second line of research has examined students’ capacities to generate their
own representations and to use those for solving a range of problems related to
algebra. This approach has differed from the earlier generation of research sum-
marized above in which researchers examined constraints that students experienced
when using normative representations that researchers prescribed. Several studies
within this line have demonstrated that students are significantly more competent at
reasoning with external representations than previously reported. These studies have
looked at ways that students solve traditional word problems (e.g., Hall 1990; Hall et
244 A. Izsák
al. 1989), generate their own tabular and algebraic presentations of physical devices
(e.g., Izsák 2000, 2003, 2004; Meira 1995, 1998), and design graphical representa-
tions of motion (e.g., diSessa 2002; diSessa et al. 1991; diSessa and Sherin 2000;
Sherin 2000). Each of these studies has reported complex interplay between stu-
dents’ understandings of problem situations and the inscriptions with which they
work.
A third line of research has concentrated on students’ transition from arithmetic
to algebra. Within this line, several researchers have studied elementary and mid-
dle grades students’ capacities to generalize patterns that could be described with
linear functions. These researchers have emphasized the mathematical notion of
function (e.g., Carraher et al. 2006, 2008), have examined how students justify their
generalizations of patterns (e.g., Lannin 2005), and have used semiotics to exam-
ine how students learn to generate and interpret algebraic notation in particular
socio-cultural contexts (e.g., Radford 2000, 2003, 2008; Rivera and Becker 2008;
Warren and Cooper 2008).
This chapter concentrates on two results from the second more recent line of
research on students’ mental structures and processes. One result has to do with
students’ competencies for generating external representations, and one has to do
with students’ capacities for interpreting external representations in context sensi-
tive ways. Both results have emerged across several case studies that have examined
students’ competencies for generating and interpreting external representations of
problem situations. Furthermore, both results have emerged across different forms
of external representation, such as tables, equations, and graphs. Finally, both results
have emerged from studies that relied on interview settings and from studies situ-
ated in classrooms. Thus, the results are not tied to one particular set of students or
one form of representation, and they are directly relevant to classroom instruction.
First Result: Criteria for Evaluating External Representations
The first result I discuss can be traced back to one influential study by diSessa et al.
(1991). These authors reported on a sequence of lessons during which 8 sixth-grade
students (∼ 12 years of age) “invented” graphing. The students were participating
in an experimental after-school class. They were presented with a description of a
motorist driving through the desert and stopping for a drink of water and were asked
to invent static “motion pictures” that communicated key aspects of the described
motion such as going fast, going slow, and stopping. The students generated approx-
imately 10 different motion pictures, evaluated each other’s approaches, and refined
several.
diSessa et al. (1991) coined the term meta-representational competence to refer
to students’ capacities to invent and critique various graphical representations of mo-
tions. The result most relevant to the present article was that students marshaled ap-
proximately a dozen criteria when evaluating each other’s work. Examples included
transparency (the criterion that representations should need little or no explanation),
appropriate abstractness (the criterion that representations can omit nonessential as-
pects of problem situations), and consistency (the criterion that conventions should
not be adjusted to accommodate features unique to a particular problem situation).
In a follow-up study with a different sample of students, diSessa and colleagues
(e.g., diSessa 2002; diSessa and Sherin 2000; Sherin 2000) have investigated stu-
dents’ capacities to generate and critique graphical representations of motion and
spatially distributed data.
diSessa (2002) provided the most elaborate extension of the initial results on stu-
dents’ criteria for graphical representations. One main results was that different stu-
dents can make systematically different judgments about external representations.
That is, different students may apply different criteria when evaluating external rep-
resentations. He also articulated two hypotheses about criteria that have implications
for how they are studied. First, criteria are design-linked. This implies that students
will most likely activate criteria when engaged in designing representations. Sec-
ond, criteria are both reactive and implicit. This implies that students employ crite-
ria when reacting to particular representations but may not be able to articulate their
criteria clearly. Thus, to gain access to criteria, one should study contexts in which
students are building representations; examine all instances in which students judge
whether representations are good or bad and infer the basis for the judgments one
observes (see diSessa 2002, for details).
Izsák (2003, 2004) has extended results on students’ criteria from graphical rep-
resentations to algebraic representations. He studied how 12 pairs of eighth-grade
(∼ 14 years of age) U.S. students generated algebraic representations of a physical
device called a winch. His data came from series of semi-structured interviews con-
ducted with each pair. Nine pairs were taking an introductory algebra course, and
three were taking a pre-algebra course that included some work with variables and
functions. The winch (see Fig. 1) exemplifies situations that can be modeled by pairs
of simultaneous linear functions. The device stands 4 feet tall and at the top has a
rod with a handle for turning two spools, one 3 and one 5 inches in circumference.
Fishing line attaches one weight to each spool. Izsák referred to these as the 3- and
5-inch weights, respectively. Turning the handle moves the weights up and down a
yardstick, allowing measurements of heights, displacements, and distances between
the two weights.
Izsák (2003, 2004) configured the initial heights of the weights in various ways
and presented students with three types of tasks:
(1) Predict the distance between the weights after an arbitrary number of cranks.
(2) Determine whether and, if so, when one weight will ever be twice as high as the
other.
(3) Determine whether and, if so, when the weights will meet at the same height.
For each type of task, he first used initial conditions that allowed students to an-
swer questions simply by turning the crank. To illustrate with the third type of task,
students might turn the handle and observe that the two weights rise and meet at
28 inches. He then asked students to imagine a larger 100-inch winch and changed
the initial conditions so that students could not simply turn the handle. As an exam-
ple, students might need to imagine that, as they turned the handle, the two weights
246 A. Izsák
Fig. 1 The winch. From

“Inscribing the Winch:
Mechanisms by Which
Students Develop Knowledge
Structures for Representing
the Physical World with
Algebra” by A. Izsák (2000),
The Journal of the Learning
Sciences, 9, p. 33. Copyright
2000 by Lawrence Erlbaum
Associates, Inc. Reprinted
with permission,
http://www.informaworld.com
would rise and meet somewhere above the physical device in front of them. This
interview strategy focused students first on the physical phenomena and then on
ways of representing those phenomena that afforded solutions to the problems. For
the students in this study, generating algebraic representations of the winch was not
automatic. Rather, they had to use what they already knew about algebraic represen-
tations to design suitable equations, and students’ discussions about those equations
afforded access to their design-linked knowledge.
Izsák (2004) reported a detailed analysis of one pair of eighth-grade students,
Amy and Kate,2 who considered the winch set up so that the 3- and 5-inch weights
started by the 14- and 0-inch marks, respectively. Over 40 minutes, the students in-
troduced, evaluated, and refined a sequence of algebraic representations for vertical
distances between the weights. Three criteria that Amy and Kate applied were single
equation (the criterion that single equations are better than multiple ones), positive
distance (the criterion that expressions must generate positive values for distances),
and consistent interpretation (the criterion that letters in an equation should have
stable interpretations in the context of the problem).
Amy proposed the first equation: 14 − 2n = d. In this equation, she let 14 stand
for the initial distance between the two weights, 2 stand for the amount by which the
distance between the weights changed with each turn of the handle, n stand for the
number of turns of the handle, and d stand for the distance between the two weights.
The students tested the equation for fewer than 7 turns of the handle and more than
7 turns. They used the positive distance criterion to reject Amy’s equation because it
resulted in negative numbers when n was larger than 7. Amy then refined her initial
2 All names are pseudonyms.

approach and suggested using two equations, 14 − 2n = d to represent the distance

until the 5-inch weight caught up to the 3-inch weight and 2n = d to represent
the distance after the 5-inch weight passed the 3-inch weight. Amy still used d for
the distance between the two weights but now used n to mean both the number of
turns of the handle from the initial set-up and just the number of turns after the
two weights met. Kate understood that substituting values into Amy’s equations
produced positive numbers that matched distances measured on the winch, but she
still objected to Amy’s approach. Although Kate did not make explicit the basis
for her objection, she may have applied the single equation criterion (because Amy
used two equations) or the consistent interpretation criterion (because Amy used n
to count cranks both from the initial winch set-up and from the point when the two
weights were at the same height).
Ultimately, the students arrived at equation (1). Although this equation was un-
like any normative equation that Amy and Kate might have seen in their algebra
class or textbook (e.g., an equation of the form y = mx + b), it did model distances
on the winch correctly. In the equation, n stood for the total number of completed
turns of the handle, 7 stood for the number of turns that it took the 5-inch weight
to catch up to the 3-inch weight, 2 stood for the amount by which the distance
between the weights changed with each turn of the handle, and d stood for the
distance between the two weights. Izsák (2004) argued that Amy and Kate were
satisfied with (1) because it met the multiple criteria for equations that they had
been using—positive distance, single equation, and consistent interpretation. Thus,
Amy and Kate’s criteria fundamentally shaped which algebraic representations of
the winch made sense to them. Moreover, because the students used similar equa-
tions to model further winch problems (see Izsák 2004, for details), their criteria
played a key role in their learning.
|(n − 7)2| = d. (1)
More recently, Izsák et al. (2009) extended results on students’ criteria from ex-
perimental classroom and interview settings to more conventional classroom set-
tings. Their study came from Coordinating Students’ and Teachers’ Algebraic Rea-
soning, a project supported by the National Science Foundation. The project took
place in a U.S. middle school that uses reform-oriented curricular materials. Izsák
et al. reported on a sequence of lessons from Ms. Jennings’s eighth-grade classroom
during which students generated and evaluated alternative algebraic representations
for word problems. The lessons came from a unit on writing and solving linear
equations that is part of College Preparatory Mathematics (Sallee et al. 2002). The
materials instruct students first to solve word problems by developing guess-and-
check tables and then to write equations that express the resulting patterns. As was
the case for Amy and Kate, generating algebraic representations of linear patterns
was not automatic for students in Ms. Jennings’s class.
One lesson was particularly notable for the number of algebraic expressions and
equations that the students generated and evaluated, and students’ reactions to alter-
native equations afforded access to their design-linked knowledge. The problem on
which they worked was the following:
248 A. Izsák
Antony joined a book club in which he received 5 books for a penny. After that, he received
2 books per month, for which he had to pay $8.95 each. So far, he has paid the book club
$196.91. How many books has he received? (Unit 4, p. 10)
There are several equations that one could set up to solve this equation. Here is one
in which x stands for the number of books that Antony received for $8.95:
8.95x + 0.01 = 196.91. (2)
An aspect of the Book Club problem that challenged many students was that
the books could be described either in terms of the total amount of money that
Antony has spent or the total number of books that he has received. This challenge
afforded opportunities for students to generate alternative correct equations to model
the problem. Faced with alternative equations, several students used final units (the
criterion that terms in equations should be expressed in the same units as those of
final requested quantities). One student, Greg, challenged equations proposed by his
classmates in which 196.91 appeared alone on the right-hand side of the equal sign.
He argued that “you are equaling it up to 196.91, but see, and what you are trying to
do is find, figure out how many books you got, not the amount of money you have.”
He followed the instructions to use a guess-and-check table, determined the correct
answer of 27 books in all, and wrote the following:
(196.91 ÷ x) + 5 = 27. (3)
Greg’s explanation for his equation was unclear because his assignment for x was
unstable. Another student, Maria, offered two correct equations:
(x − 5) · 8.95 + .01 = 196.91 (4)
and
196.90 ÷ 8.95 + 5 = x. (5)
Notice that each term in equation (4) expresses an amount of money and that each
term in equation (5) expresses a number of books. When proposing equation (4),
Maria explained that x was the total number of books that Antony received, that she
subtracted the 5 books for one penny, and that she multiplied the result by 8.95. In
equation (5), Maria divided to determine the number of books that Antony received
for $8.95 each and then added the 5 books for penny. In a subsequent interview, she
reconsidered both of her equations and expressed a preference for equation (5). Like
Greg, Maria justified her choice by referring to the problem statement that asked for
the number of books that Antony has received.
As mentioned above, students’ tendency to interpret the equal sign as a com-
mand to compute an answer is well known (e.g., Kieran 1981; Knuth et al. 2006).
Although one might consider equation (3), equation (4), and equation (5) as further
examples of this phenomenon, the data on Greg and Maria suggested more complex
understandings of equations. Note that equation (3) and equation (4) demonstrate
that both students could generate equations that combine the letter x with other
numbers on left-hand side of equal sign. Neither in equation (3) nor in equation (4)
can one use the left-hand side alone to compute a number. Furthermore, both Greg
and Maria recognized that there was more than one way to construct an equation
with a number isolated on the right-hand side of the equal sign, and they had a
criterion for evaluating two alternatives.
Finally, in addition to documenting students’ use of criteria in an algebra I class-
room, Izsák et al. (2009) reported on criteria that Ms. Jennings used during the
sequence of lessons. This is the first study of which I am aware that has extended
results on criteria for representations from students to teachers. Describing these re-
sults are beyond the scope of the present chapter (see Izsák et al. 2009, for details).
Studies summarized in this section provide accumulating evidence for a com-
plex substrata of knowledge consisting of criteria, which are fine-grained knowl-
edge elements that support students’ capacities to evaluate external representations
of problem situations. This substrata is particularly visible when students are design-
ing external representations, or inscriptions, and are reacting to alternatives. Initial
results that emerged in conjunction with graphical representations have been ex-
tend to results on algebraic representations. Moreover, results that first emerged in
experimental contexts have provided insight into teaching and learning in more con-
ventional algebra classrooms.
Second Result: Adaptive Interpretation
The second result I discuss extends earlier reports that students have difficulty inter-
preting normative algebraic and graphical representations appropriately (e.g., Lein-
hardt et al. 1990; Matz 1982). As mentioned above, past results have tended to
concentrate on the errors that students make and on students’ prior experiences that
might be sources for those errors. The present result highlights the fact that often-
times, in the course of solving problems about situations, students must coordinate
shifts in their perspective on external representations with corresponding shifts in
their perspective on situations. I refer to this aspect of reasoning as adaptive inter-
pretation (see also Izsák and Findell 2005). The notion of adaptive interpretation
allows one to examine whether student errors arise not from misconceptions, but
rather from applying knowledge appropriate for one situation inappropriately in
another. As when discussing results about students’ criteria for external represen-
tations, I demonstrate that results about adaptive interpretation are not tied to one
particular group of students or form of external representation. I will also demon-
strate that adaptive interpretation can play an important role both in experimental
settings and in more conventional classrooms. For the first example, I return to the
winch (see Fig. 1).
Izsák (2003) reported further results on Amy and Kate’s reasoning when consid-
ering whether one weight on the winch would ever be twice as high as the other. The
students examined the winch set up so that the 3-inch weight started by the 28-inch
mark and the 5-inch weight started by the 0-inch mark. They described twice as high
correctly by equating the height of the lower weight with the distance between the
250 A. Izsák
two weights, and they generated three equations:
0 + 5n = h
28 − 2n = d (6)
d = h.
In this system, the students let n stand for the total number of turns of the handle,
h stand for the height of the weight attached to the 5-inch spool, and d stand for
the distance between the two weights. The first equation expressed the height of
the 5-inch weight for any number of turns of the handle, the second expressed the
distance between the two weights for any number of turns of the handle (the initial
distance decreased by 2 inches with each turn), and the third expressed the specific
moment when the lower height would equal the distance between the two weights.
Initially, Amy resisted equating the expressions 0 + 5n and 28 − 2n because each
expression was true for any number of turns of the handle, but the resulting equation
was not. At one point, she objected that “The height and the distance between the
two is not equal” and went on to say “We want a statement that is always going to
be true.” Because, in some cases, appropriate equations do hold for all values of the
independent variable, Amy’s criterion was reasonable: It was her application of this
knowledge to the present situation that created tension. Kate was more confident
in equation (6), but her understanding was still emerging. Thus, the students had to
develop adaptive interpretations that would allow them to examine algebraic repre-
sentations for all n in some contexts and for unique n in others. The students finally
accepted 0 + 5n = 28 − 2n after discovering they could solve for n and answer
the problem. Matz (1982) also reported students who had trouble distinguishing be-
tween equations that are true for any value of the independent variable from those
that constrain the independent variable to a unique value.
The second example of students struggling with adaptive interpretation comes
from a second case study conducted as part of the Coordinating Students’ and
Teachers’ Algebraic Reasoning project (see above). The example extends results
on adaptive interpretation from algebraic representations to tabular and graphical
representations and from experimental settings to more conventional classroom set-
tings. The seventh-grade teacher, Ms. Bishop, and her students (∼ 13 years of age)
were studying a unit called Variables and Patterns that focuses on using tables and
graphs to solve problems about situations that contain covarying quantities. The unit
is part of the Connected Mathematics Program materials (CMP, Lappan et al. 2002),
one set of curriculum materials developed with support from the National Science
Foundation in response to NCTM standards (see above).
Students in Ms. Bishop’s class were working on the Popcorn Problem (abbrevi-
ated in Fig. 2). Prior to working on this problem, they had completed an activity
in which they performed jumping jacks and recorded the total every 10 seconds.
(Jumping jacks are a form of exercise performed from a standing position by jump-
ing to a position with legs spread and arms raised above the head, jumping to the
original position, and repeating.) As the students tired, the total increased more
slowly. The Popcorn Problem was a second task in which students were asked to
Fig. 2 From Connected Mathematics: Variable and Patterns Teacher Edition Page 10. Copyright
© 2004 by Michigan State University, Glenda Lappan, James T. Fey, William M. Fitzgerald, Susan
N. Friel, and Elizabeth D. Phillips. Adapted by permission of Pearson Education, Inc. All rights
reserved
reason about non-linear data. Although the word “change” in part b could refer ei-
ther to changes each hour or to increases over the course of the day, the sample
answer in the teacher’s guide makes the intended meaning clear:
Very few bags were sold before 7 a.m., perhaps because many people do not eat popcorn so
early in the morning. But the number jumped by 12 bags between 7 a.m. and 8 a.m., when
perhaps people were stopping for a snack on their way to school. The number goes up at a
rate of about 5 bags per hour between 8 a.m. and 11 a.m. From 11 a.m. until noon it jumps
to 15 bags, and 13 bags from noon to 1 p.m.; during these two hours, perhaps people are
buying lunch. (Lappan et al. 2002, p. 10)
The intended task requires adaptive interpretation because students must be able
to look at the fourth line of the table, for example, and interpret it as 20 bags sold
by 9 a.m. At the same time, when turning their attention to sales each hour, students
must be able to determine by subtraction that 5 bags were sold between 8 a.m. and
9 a.m., and so on. Thus, students must coordinate a shift in attention from total sales
to sales each hour with a shift in attention from rows in the table to differences in
those rows.
Adaptive interpretation may be unproblematic for those with experience using
tables and graphs to reason about covarying quantities, but the classroom and in-
terview data made clear that coordinating shifting perspectives on situations with
shifting interpretations of representations was challenging for many students. Dur-
ing the initial class discussion of part b to the Popcorn Problem, some students gave
answers that did not make clear distinctions between the total number of bags sold
for the entire day and the number of bags sold each hour. For instance, Ashley may
have focused on increases in both total bags sold and bags sold each hour when she
offered, “The changes increased probably because as the day went on more people
wanted popcorn and most people don’t want popcorn in the morning.” Other stu-
dents apparently misinterpreted the table in Fig. 2 as bags sold per hour: Rachel
added the entries for total bags sold and said, “I put from 6 a.m. to noon they only
sold 139 bags.”
As the class discussion continued, students attended to both sales each hour and
total sales but many seemed not to coordinate the situation and the table appro-
priately. To make clearer the distinction between total sales and sales each hour,
252 A. Izsák
Fig. 3 Revisiting the

Popcorn Problem
Ms. Bishop introduced the phrase “cumulative graph” to describe the popcorn data
and defined cumulative as “successive addition” or “continuously adding to.” Later
during the same lesson, she moved to the next problem in the materials which used
discrete dots to show soda sales each hour over a day. Ms. Bishop called this a “rate
graph.” Absent from the lesson was an explicit explanation of how the table in Fig. 2
conveyed both cumulative and rate data.
Over the next several lessons, the students often began work on a problem by
considering whether the included table or graph presented cumulative or rate data
but they continued to have difficulty with the distinction. To address the persistent
confusion, Ms. Bishop returned to the Popcorn Problem 10 days later and added a
third column to record differences between successive rows. Figure 3 reproduces her
written work. Although Ms. Bishop had verbally explained the relationship between
total sales and sales each hour during previous lessons, this was the first time that
she used inscriptions to explicitly demonstrate how to see both either in a table or in
a graph. She did not explain the significance, if any, of the ovals or the rectangles.
Interviews with students from Ms. Bishop’s class revealed that adaptive interpre-
tation of the popcorn table remained problematic, even after Ms. Bishop revisited
the problem. I present data on one pair of students, Nikki and Jennifer. Nikki was
a mid-achieving student, and Jennifer was a low- to mid-achieving student. I con-
ducted the interview with these students 4 days after Ms. Bishop generated the table
reproduced in Fig. 3. After watching a video clip of the class discussion, Nikki and
Jennifer could describe the subtraction that Ms. Bishop used to generate each of the
numbers in the augmented table. Nikki also commented, “I really understood what
she did that day for the first time ‘cause she actually broke it down from hour to
hour.” That the students continued to have difficulty with adaptive interpretation be-
came clear in the following exchange (Nikki’s language did not distinguish between
graphs and tables):
Int: So what do you think about the 12 and the 5 (referring to the right hand column in
Fig. 3)? Are they, are those cumulative data? Are they rate data? Are they some other kind
of data?
Nikki: Now, with those two numbers, I’m not exactly sure because she subtracted them.
Jennifer: I think rate.
Nikki: Maybe. But it’s a cumulative graph. So why would it have two different sets of
information on there if it’s cumulative?
As the interview progressed, Nikki asked her question about “two different sets
of information” in the context of graphs as well. The interviewer gave the students
a graph showing new popcorn sales data as discrete dots, similar to other discrete
graphs with which the students had worked during class. The interviewer empha-
sized that each point on the graph indicated how many bags had been sold all
together and then asked the students how many were sold each hour. Nikki and
Jennifer answered by calculating differences. The interviewer then pointed out that
calculated differences were “rate data” in the sense discussed in class because they
represented how many bags sold in one hour. The interviewer was trying to get the
students to see that, similar to the table in Fig. 3, one graph could convey both cu-
mulative and rate data. He then reminded Nikki of her comment that a graph can
show only one type of information:
Nikki: Yeah. That’s what I figured. Since it’s being a cumulative graph, why would it be
one kind of graph and have another type of information on it?
Int: Okay. What do you think now?
Nikki: I’m still unsure.
Int: Really?
Nikki: I guess I won’t be clear about it unless I ask Ms. Bishop personally. That’s definitely
something I can’t answer myself. Because it’s confusing: You thinking about a graph being
cumulative and having rate information, or you think about a rate graph having cumulative
information, or you think about another graph having another kind of information. It doesn’t
ring a bell to me.
In further discussion, Nikki continued to demonstrate understanding of the rela-
tionship between total sales and sales each hour and was apparently learning that a
given graph or table could convey both types of information. That Nikki struggled
with her question about “two different sets of information” when working with both
tables and graphs indicates that she was not asking about one particular form of rep-
resentation, tables or graphs, but about representations more generally. It is possible
that classroom discussions in which Ms. Bishop and her students identified partic-
ular tables and graphs as cumulative or rate contributed unintentionally to Nikki’s
difficulties.
The two studies summarized in this section provide evidence that interpreting ex-
ternal representations commonly used in algebra can be particularly challenging for
students because appropriate interpretations can be context sensitive. These chal-
lenges arise when students are generating and then interpreting their own inscrip-
tions, as in the first example from Amy and Kate’s work, and when students are
working with external representations given to them, as in the example from Ms.
Bishop’s classroom. These results imply that part of developing representational
competence involves learning how to recognize when particular interpretations of
a given representation are, and are not, appropriate. Finally, as in the case of cri-
teria, initial results that emerged in experimental settings have proven germane to
teaching and learning in more conventional algebra classrooms.
Conclusion
An earlier generation of research, conducted during the 70s and 80s, examined stu-
dents’ understandings of external representations commonly found in algebra by
254 A. Izsák
prescribing the forms of representation with which students were to accomplish

given tasks. Results emphasized students’ difficulties generating and interpreting
normative algebraic and graphical representations appropriately. A more recent gen-
eration of research has uncovered competencies that students demonstrate when
they are allowed to generate their own external representations, or inscriptions, to
solve problems about situations. Furthermore, this recent line of research has pro-
duced an accumulating body of evidence that students have knowledge specifically
for reasoning with external representations of problem situations. I conclude with
several implications of this insight for future research and instruction related to al-
gebra.
The first implication for future research has to do with the grain size of knowl-
edge that students employ when generating and interpreting external representa-
tions to solve problems about situations. The most generalized finding that cuts
across the studies summarized in the present article is that capturing students’ sense
making when using external representations, or inscriptions, requires analyses of
fine-grained knowledge. This finding underscores the limitations of course-grained
accounts of knowledge for gaining insight into students’ experiences. As discussed
in a preceding section of the chapter, the 3-slot schema (e.g., Moschkovich 1998;
Schoenfeld et al. 1993) exposes limitations of translation for describing the process
by which students connect algebraic and graphical representations. Results about
criteria for algebraic and graphical representations and about adaptive interpretation
provide insight into how students might self-regulate broader cognitive processes
for reasoning with external representations, including the processes identified by
Kaput (1987, 1989, 1991) of encoding and reading and of syntactic and semantic
elaboration. Future research on students’ sense making should be directly informed
by this general result—for instance, by explicitly identifying and justifying the grain
size at which knowledge is described.
The second implication for future research builds on the first. There is much
still to learn about the forms and contents of students’ knowledge specific to rea-
soning with external representations of problem situations, but results to date sug-
gest a substratum of diverse elements that includes complex structures like the 3-
slot schema reported by Schoenfeld et al. (1993) and by Moschkovich (1998) and
also criteria for various forms of external representations like those reported by
diSessa and colleagues (diSessa 2002; diSessa et al. 1991; diSessa and Sherin 2000;
Sherin 2000) and by Izsák and colleagues (Izsák 2003, 2004; Izsák et al. 2009). The
literature contains still other reports of fine-grained knowledge structures for reason-
ing with algebraic representations (e.g., Sherin 2001) and graphical representations
(Clement 1989). Research consistent with broad tenets of constructivist epistemol-
ogy should continue to seek access to this substratum of knowledge because it is
often at this level that one can “see” students’ prior knowledge supporting and con-
straining their reasoning.
The third implication for future research is to study growth and change of stu-
dents’ knowledge for reasoning with external representations. Several studies dis-
cussed above (e.g., Izsák 2003, 2004; Moschkovich 1998; Schoenfeld et al. 1993)
have reported cases in which students refined and reorganized their knowledge as
they worked through a series of challenging tasks. Nevertheless, we still do not
have a good picture of how less expert knowledge for reasoning with external rep-
resentations, or inscriptions, is gradually transformed into more expert knowledge.
For instance, we know only a little about how students’ experiences generating and
interpreting external representations, or inscriptions, in arithmetic support and con-
strain their capacities to generate and interpret algebraic and graphical represen-
tations. One well-known example is students’ difficulties interpreting equal signs
(e.g., Kieran 1981), where researchers have conjectured that students’ experiences
with the equal sign in arithmetic lead them to understand the sign as a command to
compute an answer. A more recent study (Knuth et al. 2006) demonstrates that mid-
dle school students today continue to have similar difficulties, even when they study
with reform-oriented instructional materials. It is plausible that as students study
arithmetic, they develop further understandings about how to use external repre-
sentations, or inscriptions, as tools for solving problems. For instance, they may
develop certain criteria for “good” representations of problem situations. Insights
into the growth and change of such knowledge could ultimately inform curricular
trajectories that better support students’ capacities to use external representations as
tools for reasoning about problem situations.
I close with implications for instruction. The implications for research discussed
above have implications for instruction because the more we understand about stu-
dents’ prior knowledge and how it might transform into more expert knowledge
the better able we will be to design appropriate learning experiences. I also empha-
size that although results about students’ criteria and adaptive interpretation were
first reported in studies that relied on experimental instruction and interviews, this
line of research has begun to provide insight into how students reason with more
conventional forms of representation—including normative tables, equations, and
graphs—in classrooms using commercially available curricular materials. Thus, it
is becoming increasingly apparent that the advances discussed in the present arti-
cle have broad implications for classroom instruction related to algebra. One way to
think about this development is that vivid results from studies based on experimental
instruction and interviews can help sensitize us to aspects of students’ reasoning that
are still important, even if they are harder to notice, in more conventional classroom
settings.
Finally, if students have a complex substrata of knowledge for reasoning with
external representations, or inscriptions, when solving problems about situations,
then teachers should elicit that knowledge. For instance, teachers could have con-
versations with students that concentrate on how external representations might be
generated and interpreted in order to solve problems about situations. These con-
versations could include explicit comparisons of different approaches. The litera-
ture already suggests a few examples of questions teachers might ask—for instance,
does a letter such as n stand for a single unknown value or for all possible values?
Other examples suggested by results about students’ criteria and adaptive interpre-
tation include the following: (1) If a letter sometimes stands for all possible values
and sometimes stands for a subset of possible values, how can you tell when to use
which interpretation? (2) Do all salient features of a situation have to be captured
when generating a representation, or is just a subset sufficient for solving a problem?
256 A. Izsák
(3) When generating an equation to solve a problem, do the terms have to express
directly the final requested quantity, or can one start by expressing constraints in
terms of different quantities? (4) What range of information about a problem situa-
tion can be determined from a single external representation? With opportunities to
think about questions such as these, students might develop greater competence at
reasoning with representations to model and solve problems with algebra.
References
Booth, L. R. (1981). Child-methods in secondary mathematics. Educational Studies in Mathemat-
ics, 12, 29–41.
Carraher, D., Schliemann, A., Brizuela, B., & Earnest, D. (2006). Arithmetic and algebra in early
mathematics education. Journal for Research in Mathematics Education, 37(2), 87–115.
Carraher, D., Martinez, M., & Schliemann, A. (2008). Early algebra and mathematical generaliza-
tion. ZDM—The International Journal of Mathematics Education, 40(1), 3–22.
Clement, J. (1982). Algebra word problem solutions: Thought processes underlying a common
misconception. Journal for Research in Mathematics Education, 13(1), 16–30.
Clement, J. (1989). The concept of variation and misconceptions in Cartesian graphing. Focus on
Learning Problems in Mathematics, 11(2), 77–87.
Coxford, A., Fey, J., Hirsch, C., Schoen, H., Burrill, G., Hart, E., & Watkins, A. (with Messen-
ger, M., & Ritsema, B.). (1998). Contemporary Mathematics in Context: A Unified Approach.
Chicago: Everyday Learning.
diSessa, A. A. (2002). Students’ criteria for representational adequacy. In K. Gravemeijer,
R. Lehrer, B. van Oers, & L. Verschaffel (Eds.), Symbolizing, Modeling and Tool Use in Math-
ematics Education (pp. 105–129). Dordrecht, The Netherlands: Kluwer Academic.
diSessa, A. A., & Sherin, B. (2000). Meta-representation: An introduction. Journal of Mathemati-
cal Behavior, 19, 385–398.
diSessa, A., Hammer, D., Sherin, B., & Kolpakowski, T. (1991). Inventing graphing: Meta-
representational expertise in children. Journal of Mathematical Behavior, 10(2), 117–160.
Hall, R. (1990). Making mathematics on paper: Constructing representations of stories about re-
lated linear functions (Technical Report No. 90-17). University of California, Irvine.
Hall, R., Kibler, D., Wenger, E., & Truxaw, C. (1989). Exploring the episodic structure of algebra
story problem solving. Cognition and Instruction, 6(3), 223–283.
Harper, E. (1987). Ghosts of diophantus. Educational Studies in Mathematics, 18, 75–90.
Izsák, A. (2000). Inscribing the winch: Mechanisms by which students develop knowledge struc-
tures for representing the physical world with algebra. The Journal of the Learning Sciences,
9(1), 31–74.
Izsák, A. (2003). “We want a statement that is always true”: Criteria for good algebraic repre-
sentations and the development of modeling knowledge. Journal for Research in Mathematics
Education, 34(3), 191–227.
Izsák, A. (2004). Students’ coordination of knowledge when learning to model physical situations.
Cognition and Instruction, 22(1), 81–128.
Izsák, A., & Findell, B. (2005). Adaptive interpretation: Building continuity between students’
experiences solving problems in arithmetic and in algebra. Zentralblatt für Didaktik der Math-
ematik, 37(1), 60–67.
Izsák, A., Çağlayan, G., & Olive, J. (2009). Teaching and learning to model word problems with
algebraic equations. The Journal of the Learning Sciences, 18, 1–39.
Janvier, C. (Ed.) (1987a). Problems of Representation in the Teaching and Learning of Mathemat-
ics. Hillsdale, NJ: Lawrence Erlbaum Associates.
Janvier, C. (1987b). Translation processes in mathematics education. In C. Janvier (Ed.), Problems
of Representation in the Teaching and Learning of Mathematics (pp. 27–32). Hillsdale, NJ:
Kaput, J. (1987). Towards a theory of symbol use in mathematics. In C. Janvier (Ed.), Problems
of Representation in the Teaching and Learning of Mathematics (pp. 159–195). Hillsdale, NJ:
Kaput, J. (1989). Linking representations in the symbol systems of algebra. In S. Wagner & C.
Kieran (Eds.), Research Issues in the Learning and Teaching of Algebra (pp. 167–194). Reston,
VA: National Council of Teachers of Mathematics; Hillsdale, NJ: Lawrence Erlbaum.
Kaput, J. (1991). Notations and representations as mediators of constructive processes. In E. von
Glasersfeld (Ed.), Radical Constructivism in Mathematics Education (pp. 53–74). Dordrecht:
Kluwer Academic.
Kerslake, D. (1981). Graphs. In K. M. Hart (Ed.), Children’s Understanding of Mathematics: 11–
16 (pp. 120–136). London: John Murray.
matics, 12(3), 317–326.
F. Lester (Ed.), Second Handbook of Research on Mathematics Teaching and Learning (pp.
707–762). Charlotte, NC: Information Age Publishing.
Kilpatrick, J., Swafford, J., & Findell, B. (Eds.) (2001). Adding It Up: Helping Children Learn
Mathematics. Washington, DC: National Academy Press.
Knuth, E. (2000). Student understanding of the Cartesian connection: An exploratory study. Jour-
nal for Research in Mathematics Education, 31(4), 500–507.
Knuth, E., Stephens, A., McNeil, N., & Alibali, M. (2006). Does understanding the equal sign
matter? Evidence from solving equations. Journal for Research in Mathematics Education,
37(4), 297–312.
Küchemann, D. (1981). Algebra. In K. M. Hart (Ed.), Children’s Understanding of Mathematics:
11–16 (pp. 102–119). London: John Murray.
Lappan, G., Fey, J. T., Fitzgerald, W., Friel, S. N., & Phillips, E. D. (2002). Connected Mathematics
Series. Glenview, IL: Prentice Hall.
Lappan, G., Fey, J. T., Fitzgerald, W., Friel, S. N., & Phillips, E. D. (2006). Connected Mathematics
2. Boston, MA: Prentice Hall.
Leinhardt, G., Zaslavsky, O., & Stein, M. K. (1990). Functions, graphs, and graphing: Tasks, learn-
ing, and teaching. Review of Educational Research, 60(1), 1–64.
Lesh, R., Post, T., & Behr, M. (1987). Representations and translations among representations in
mathematics learning and problem solving. In C. Janvier (Ed.), Problems of Representation
in the Teaching and Learning of Mathematics (pp. 33–40). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Matz, M. (1982). Towards a process model for high school algebra errors. In D. Sleeman & J. S.
Brown (Eds.), Intelligent Tutoring Systems (pp. 25–50). New York: Academic Press.
Meira, L. (1995). The microevolution of mathematical representations in children’s activities. Cog-
nition and Instruction, 13, 269–313.
Meira, L. (1998). Making sense of instructional devices: The emergence of transparency in math-
ematical activity. Journal for Research in Mathematics Education, 29, 121–142.
Mokros, J. R., & Tinker, R. F. (1987). The impact of microcomputer-based labs on children’s
ability to interpret graphs. Journal of Research in Science Teaching, 24(4), 369–383.
Monk, S., & Nemirovsky, R. (1994). The case of Dan: Student construction of a functional situation
through visual attributes. In A. Schoenfeld, E. Dubinsky, & J. Kaput (Eds.), Research in Colle-
giate Mathematics Education (Vol. 1, pp. 139–168). Washington, DC: American Mathematics
Association.
Moschkovich, J. (1998). Resources for refining mathematical conceptions: Case studies in learning
about linear functions. The Journal of the Learning Sciences, 7(2), 209–237.
National Council of Teachers of Mathematics (1989). Curriculum and Evaluation Standards for
School Mathematics. Reston, VA: Author.
258 A. Izsák
Nemirovsky, R. (1994). On ways of symbolizing: The case of Laura and the velocity sign. Journal
of Mathematical Behavior, 13, 389–422.
Radford, L. (2000). Signs and meaning in students’ emergent algebraic thinking: A semiotic anal-
ysis. Educational Studies in Mathematics, 42(3), 237–268.
Radford, L. (2003). Gestures, speech, and the sprouting of signs: A semiotic cultural approach to
Radford, L. (2008). Iconicity and contraction: A semiotic investigation of forms of algebraic gen-
eralizations of patterns in different contexts. The International Journal of Mathematics Educa-
tion, 40(1), 82–96.
Rivera, F. D., & Becker, J. R. (2008). Middle school children’s cognitive perceptions of construc-
tive and deconstructive generalizations involving linear figural patterns. ZDM—The Interna-
tional Journal of Mathematics Education, 40(1), 65–82.
Romberg, T. A., Fennema, E., & Carpenter, T. P. (Eds.) (1993). Integrating Research on the Graph-
ical Representation of Functions. Hillsdale, NJ: Erlbaum.
Sallee, T., Kysh, J., Kasimatis, E., & Hoey, B. (2002). College Preparatory Mathematics 1 (Alge-
bra 1) (2nd ed.). Sacramento, CA: CPM Educational Program. Teacher Edition: Version 6.1.
Schoenfeld, A. H., Smith, J., & Arcavi, A. (1993). Learning: The microgenetic analysis of one
student’s evolving understanding of a complex subject matter domain. In R. Glaser (Ed.), Ad-
vances in Instructional Psychology (Vol. 4, pp. 55–175). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Sherin, B. (2000). How students invent representations of motion: A genetic account. Journal of
Mathematical Behavior, 19(4), 399–441.
Sherin, S. (2001). How students understand physics equations. Cognition and Instruction, 19(4),
479–541.
Smith, J. P., diSessa, A. A., & Roschelle, J. (1993). Misconceptions reconceived: A constructivist
analysis of knowledge in transition. The Journal of the Learning Sciences, 3(2), 115–163.
TERC (2008). Investigations in Number, Data, and Space (2nd ed.) Glenview, IL: Pearson Scott
Foresman.
Warren, E., & Cooper, T. (2008). Generalising the pattern rule for visual growth patterns: Actions
that support 8 year olds’ thinking. Educational Studies in Mathematics, 67(2), 171–185.
Wisconsin Center for Educational Research & Freudenthal Institute (Eds.) (2006). Mathematics in
Context. Chicago: Encyclopedia Britannica, Inc.
Middle School Students’ Understanding of Core
Algebraic Concepts: Equivalence & Variable
Eric J. Knuth, Martha W. Alibali, Nicole M. McNeil, Aaron Weinberg,

and Ana C. Stephens
Abstract Algebra is a focal point of reform efforts in mathematics education, with

many mathematics educators advocating that algebraic reasoning should be inte-
grated at all grade levels K-12. Recent research has begun to investigate algebra
reform in the context of elementary school (grades K-5) mathematics, focusing in
particular on the development of algebraic reasoning. Yet, to date, little research has
focused on the development of algebraic reasoning in middle school (grades 6–8).
This article focuses on middle school students’ understanding of two core alge-
braic ideas—equivalence and variable—and the relationship of their understanding
This research is supported in part by the National Science Foundation under grant No.
REC-0115661. The opinions expressed herein are those of the authors and do not necessarily
reflect the views of the National Science Foundation, the Department of Education, or the
National Institute of Child Health and Human Development. This chapter is a reprint of an article
published in ZDM—International Reviews on Mathematical Education, 37(1), 68–76.
DOI 10.1007/BF0255899.
E.J. Knuth ()
M.W. Alibali
Department of Psychology, University of Wisconsin-Madison, Madison, USA
N.M. McNeil
Department of Psychology, University of Notre Dame, Notre Dame, USA
A. Weinberg
Department of Mathematics, Ithaca College, Ithaca, USA
A.C. Stephens
Wisconsin Center for Education Research, University of Wisconsin-Madison, Madison, USA

260 E.J. Knuth et al.
to performance on problems that require use of these two ideas. The data suggest
that students’ understanding of these core ideas influences their success in solving
problems, the strategies they use in their solution processes, and the justifications
they provide for their solutions. Implications for instruction and curricular design
are discussed.
Introduction
Algebra is considered by many to be a “gatekeeper” in school mathematics, critical
to further study in mathematics as well as to future educational and employment op-
portunities (Ladson-Billings 1998; National Research Council [NRC] 1998). Unfor-
tunately, many students experience difficulty learning algebra (Kieran 1992), a fact
that has led to first-year algebra courses in the United States being characterized
as “an unmitigated disaster for most students”. (NRC, p. 1) In response to grow-
ing concern about students’ inadequate understandings and preparation in algebra,
and in recognition of the role algebra plays as a gatekeeper, recent reform efforts
in mathematics education have made algebra curricula and instruction a focal point
(e.g., Bednarz et al. 1996; Lacampagne et al. 1995; National Council of Teachers of
Mathematics 1997, 2000; NRC 1998; RAND Mathematics Study Panel 2003). In
fact, Kaput (1998) has argued that algebra is the keystone of mathematics reform,
and that teachers’ abilities to facilitate the development of students’ algebraic rea-
soning is the most critical factor in algebra reform. Moreover, he contended that the
“key to algebra reform is integrating algebraic reasoning across all grades and all
topics—to ‘algebrafy’ school mathematics”. (p. 1)
Underlying this call to ‘algebrafy’ school mathematics is a belief that the tradi-
tional separation of arithmetic and algebra deprives students of powerful schemes
for thinking about mathematics in the early grades and makes it more difficult for
them to learn algebra in the later grades (Kieran 1992). Algebrafying school mathe-
matics, however, means more than moving the traditional first-year algebra curricu-
lum down to the lower grades. There is a growing consensus that algebra reform
requires a reconceptualization of the nature of algebra and algebraic reasoning as
well as a reexamination of when children are capable of reasoning algebraically and
when ideas that require algebraic reasoning should be introduced into the curricu-
lum (Carpenter and Levi 1999). Recent research has begun to investigate algebra
reform in the context of elementary school mathematics, focusing in particular on
the development of algebraic reasoning (e.g., Bastable and Schifter 2008; Carpenter
et al. 2003; Carpenter and Levi 1999; Kaput et al. 2008). Yet, to date, little research
has focused on the development of algebraic reasoning in the Middle Grades—the
time period linking students’ arithmetic and early algebraic reasoning and their de-
velopment of increasingly complex, abstract algebraic reasoning. In this chapter, we
present results from a multi-year research project that seeks to understand the de-
velopment of middle school students’ algebraic reasoning. In particular, the chapter
focuses on students’ understanding of two core algebraic ideas—equivalence and
variable—and the relationship of their understanding to performance on problems
that require use of these two ideas.
Middle School Students’ Understanding of Core Algebraic Concepts 261
Student Understanding of Equivalence & Variable

Algebraic reasoning depends on an understanding of a number of key ideas, of
which equivalence and variable are, arguably, two of the most fundamental. In this
section we briefly describe research that has examined students’ understandings of
these two ideas; this description will serve to situate the present study in the larger
body of research as well as to highlight the contribution of the present study.
Equivalence
The ubiquitous presence of the equal sign symbol in mathematics at all levels high-
lights its important role in mathematics, in general, and in algebra, in particular.
Within the domain of algebra, Kieran (1992) contended that “one of the require-
ments for generating and adequately interpreting structural representations such as
equations is a conception of the symmetric and transitive character of equality—
sometimes referred to as the ‘left-right equivalence’ of the equal sign”. (p. 398)
Yet, there is abundant literature that suggests students do not view the equal sign
as a symbol of equivalence (i.e., a symbol that denotes a relationship between
two quantities), but rather as an announcement of the result or answer of an arith-
metic operation (e.g., Falkner et al. 1999; Kieran 1981; McNeil and Alibali 2005;
Rittle-Johnson and Alibali 1999). For example, Kieran (1981) found that 12- and
13-year old students described the equal sign in terms of the answer and provided
examples of its use that included an operation on the left-hand side of the symbol
and the result on the right-hand side (e.g., 3 + 4 = 7). McNeil and Alibali (2005)
found similar conceptions of the equal sign in definitions generated by third- through
fifth-grade students.
While such a (mis)conception concerning the meaning of the equal sign may not
be problematic in elementary school, where students are typically asked to solve
equations of the form a + b = 1, it does not serve students well in terms of their
preparation for algebra and algebraic ways of thinking. In algebra, students must
view the equal sign as a relational symbol (i.e., “the same as”) rather than as an
operational symbol (i.e., “do something”). The relational view of the equal sign
becomes particularly important as students encounter and learn to solve algebraic
equations with operations on both sides of the symbol (e.g., 3x − 5 = 2x + 1).
A relational view of the equal sign is essential to understanding that the transfor-
mations performed in the process of solving an equation preserve the equivalence
relation (i.e., the transformed equations are equivalent)—an idea that many students
find difficult, and that is not an explicit focus of typical instruction. Steinberg et al.
(1990) concluded that many eighth- and ninth-grade students do not have a good
understanding of equivalent equations. They found that many students knew how
to use transformations in solving equations, however, many of these same students
did not seem to utilize such knowledge in determining whether two given equations
were equivalent. Although not examined in their study, it seems reasonable to con-
clude that many of these latter students may have had inadequate conceptions of
mathematical equivalence.
Variable
Algebra has been called the study of the 24th letter of the alphabet. Although this
characterization is somewhat facetious, it underscores the importance of develop-
ing a meaningful conception of variable in learning and using algebra. The idea
of variable, not surprisingly, has also received considerable attention in the math-
ematics educational research community (e.g., Küchemann 1978; MacGregor and
Stacey 1997; Philipp 1992; Usiskin 1988), and the results of such work suggest that
the use of literal symbols in algebra presents a difficult challenge for students. In
Küchemann’s frequently cited study, for example, he found that most 13-, 14-, and
15-year-old students considered literal symbols as objects (i.e., the literal symbol is
interpreted as a label for an object or as an object itself). Few students considered
them as specific unknowns (i.e., the literal symbol is interpreted as an unknown
number with a fixed value), and fewer still as generalized numbers (i.e., the literal
symbol is taken to represent multiple values, although it is only necessary to think of
the symbol taking on these values one at a time) or variables (i.e., the literal symbol
represents, at once, a range of numbers). Further, his study showed that students’
misunderstandings of literal symbols seem to be reflected in their approaches to
symbolizing relationships in problem solutions—an essential aspect of algebra and
algebraic ways of thinking.
In sum, developing an understanding of equivalence and variable is essential to
algebra and the ability to use it, yet they are ideas about which many students have
inadequate understandings. In this article, we examine the meanings middle school
students ascribe to the equal sign and variable, their performance on problems that
require use of these ideas, and the relationship between the meanings they ascribe
to each and their performance on the corresponding problems.
Method
Participants
Participants were 373 middle-school (6th through 8th grade) students drawn from
an ethnically diverse middle school in the American Midwest. The demographic
breakdown of the school’s student population is as follows: 25% African American,
5% Hispanic, 7% Asian, and 62% White. The middle school had recently adopted a
reform-based curricular program, Connected Mathematics, and, with the exception
of one section of 8th grade algebra, the classes were not tracked (e.g., all 6th grade
students were in the same mathematics course). The school was selected as the site
for this research based upon the recommendation of the school district’s mathemat-
ics resource teacher, who felt that the principal and teachers would be interested in
participating.
The following questions are about this statement:

3+4=7
↑
a) The arrow above points to a symbol. What is the name of the symbol?
b) What does the symbol mean?
c) Can the symbol mean anything else? If yes, please explain.
Fig. 1 Interpreting the equal sign
Is the number that goes in the 1 the same number in

the following two equations? Explain your reasoning.
2× 1+15 = 31 2× 1+15 − 9 = 31 − 9
Fig. 2 Using the concept of mathematical equivalence
Data Collection
The data that are the focus of this article consist of students’ responses to a subset
of items from a written assessment that targeted their understandings of various
aspects of algebra. In particular, the focus is on four items that were designed to
assess students’ understanding of the ideas of equal sign (1 item) and of variable (1
item) as well as their performance on two Problem solving items that (potentially)
required the use of these ideas. The assessment consisted of three forms with some
overlap of items; all 373 students were administered the equal sign understanding
and variable understanding items, 251 students were administered the equal sign
performance item, and 122 students were administered the variable performance
item. The assessment was administered near the beginning of the school year.
Equal Sign Items
In the first item (shown in Fig. 1), students were asked to define the equal sign. The
rationale for the first prompt (What is the name of the symbol?) was to preempt
students from using the name of the symbol in their response to the second prompt
(What does the symbol mean?). The rationale for the third prompt (Can the symbol
mean anything else?) was to provide students the opportunity to give an alternative
interpretation; in previous work, we have found that students often offer more than
one interpretation when given the opportunity. The second item (shown in Fig. 2),
the equivalent equations problem, was designed to assess students’ understanding
of the fact that the transformations performed in the process of solving an equation
preserve the equivalence relation. We expected that students who viewed the equal
sign as representing a relationship between quantities would conclude that the num-
ber that goes in the box is the same in both equations because the transformation
The following question is about this expression:

2n + 3
↑
The arrow above points to a symbol. What does the symbol stand for?
Fig. 3 Interpreting a literal symbol used as a variable
Can you tell which is larger, 3n or n + 6? Please explain your answer.

Fig. 4 Using the concept of variable
performed on the second of the two equations preserved the quantitative relation-
ship expressed in the first equation.1
Variable Items
Item 3 (shown in Fig. 3) was designed to assess students’ interpretations of literal

symbols. The fourth item (shown in Fig. 4), the which is larger problem, was de-
signed to assess students’ abilities to use the concept of variable to make a judgment
about two varying quantities. In particular, to be successful on the final item, stu-
dents must recognize that the values of 3n and n + 6 are dynamic and depend on the
value of n, that is, they must view n as a variable—a literal symbol that represents,
at once, a range of numbers.
Coding
In this section we provide details regarding the coding of each item; in the re-
sults section, we provide sample student responses. For all items, responses that
students left blank or for which they wrote “I don’t know” were grouped in a no
response/don’t know category, and responses for which students’ reasoning could
not be determined and responses that were not sufficiently frequent to warrant their
own codes were grouped in an other category.
Coding Equal Sign Understanding
Student responses to parts b) and c) of Item 1 were coded as relational, operational,

other, or no response/don’t know, with the majority of responses falling into the
1 There were two versions of the equivalent equations problem, one which used a box (as in Fig. 2)
and one which used n to represent the missing values. Performance did not differ across versions
(60% correct box, 61% correct n), and the distribution of strategies used to solve the problem did
not differ across versions (χ 2 (4, N = 252) = 1.729, ns), so we collapse across versions in the
analyses presented in this chapter.
first two categories. A response was coded as relational if a student expressed the
general idea that the equal sign means “the same as” and as operational if the stu-
dent expressed the general idea that the equal sign means “add the numbers” or “the
answer”. In addition to coding the responses to parts b) and c) separately, students
were also assigned an overall code indicating their “best” interpretation. Many stu-
dents provided two interpretations, often one relational and one operational; in such
cases, the responses were assigned an overall code of relational.
Coding Performance on the Equivalent Equations Problem
Students’ responses to Item 2 were coded for correctness as well as strategy use.
Responses were coded as correct if students responded that the two equations have
the same solution. Students’ strategies for solving the problem were classified into
one of five categories: answer after equal sign, recognize equivalence, solve and
compare, other, or no response/don’t know. In the answer after equal sign category,
students’ rationale for their conclusion was that each equation had the same “an-
swer” (in this case, 31) to the immediate right of the equal sign and the equations
were therefore equivalent (an incorrect strategy). In the solve and compare category,
students’ rationale for their conclusion was based on either (1) determining the so-
lution to the first equation, substituting that solution into the second equation, and
noting that the value satisfied both equations, or (2) determining the solutions to
both equations and comparing them. Finally, in the recognize equivalence category,
students’ rationale for their conclusion was based on recognizing that the transfor-
mation performed on the second equation preserved the equivalence relation. Note
that only the recognize equivalence strategy appears to explicitly require a relational
understanding of the equal sign.
Coding Variable Understanding
Students’ responses to the literal symbol interpretation item were classified into
five categories, multiple values, specific number, object, other, or no response/don’t
know. A response was coded as multiple values if the student expressed the general
idea that the literal symbol could represent more than one value; as specific number
if the student indicated that the literal symbol represents a particular number; and
as object if the student suggested that the literal symbol represents a label for a
physical object (such as stating that n represents newspapers).
Coding Performance on the Which Is Larger Problem
Students’ responses to the which is larger problem were coded both in terms of
the judgment about which quantity was larger (3n, n + 6, or can’t tell) and for the
reasoning underlying that judgment. Students’ explanations of their reasoning were
classified into five categories: variable explanations, single-value explanations, op-

eration explanations, other, or no response/don’t know. variable explanations ex-
pressed the general idea that one cannot determine which quantity is larger because
the variable can take on multiple values. Single-value explanations tested a single
value and drew a conclusion on that basis; thus, students’ conclusions varied de-
pending on the value tested. Operation explanations expressed the general idea that
one type of operation leads to larger values than the other (for example, multiplica-
tion produces larger values than addition).
Coding Reliability
To assess reliability of the coding procedures, a second coder rescored approxi-

mately 20% of the data. Agreement between coders was 90% for coding students’
interpretations of the equal sign, 91% for coding students’ strategies on the equiva-
lent equations problem, 91% for coding students’ interpretations of literal symbols,
and 95% for coding students’ explanations on the which is larger problem.
Results
We focus first on students’ interpretations of the equal sign symbol, and how these
interpretations relate to performance on the equivalent equations problem. We then
turn to students’ interpretations of a literal symbol (n) used as a variable, and how
these interpretations relate to performance on the which is larger problem. Repre-
sentative excerpts from students’ written responses are provided to illustrate partic-
ular findings. In reporting the results, we describe (and illustrate) only those coding
categories that are most germane to the focus of the article. Finally, the statistical
analysis of the data was performed using logistic regression because the outcome
variables of interest were categorical. All reported statistics are significant with al-
pha set at .05.
Interpretation of the Equal Sign
We first examined the relationship between grade level and interpretation of the
equal sign symbol. The following responses are typical of those coded as opera-
tional:
It means the total of the numbers before it. (6th grade student)
It means whatever is after it is the answer. (8th grade student)
The following responses are typical of those coded as relational:

Fig. 5 Equal sign interpretations of sixth-, seventh-, and eighth-grade students
It means the number(s) on its left are equivalent to the number(s) on its right. (6th grade
student)
The things on both sides of it are of the same value. (7th grade student)
Students were classified as providing a relational interpretation if they provided

one on either their first or second response. As seen in Fig. 5, the proportion of
students providing a relational interpretation for the equal sign differed across the
grades, Wald(2, N = 373) = 7.80, and this difference was accounted for by a sig-
nificant linear trend, β̂ = 0.52, z = 2.78, Wald(1, N = 373) = 7.72. Despite this
improvement across grades, however, the overall level of performance was strik-
ingly low. Even at grade 8, only 46% of students provided a relational interpretation
of the equal sign.
Performance on the Equivalent Equations Problem
The proportion of students who correctly judged that the two equations had the same
solution differed across the grades, Wald(2, N = 251) = 10.21, p = .006, and was
accounted for by a significant linear trend, β̂ = 0.72, z = 3.18, Wald(1, N = 251) =
10.08, p = .002. Students’ strategies for solving the equivalent equations problem
are displayed in Table 1. The majority of students’ strategies were categorized into
one of the following categories: recognize equivalence, solve and compare, or an-
swer after equal sign. Typical responses in each of these three categories included
the following:
Yes because you’re doing the same equation but just minusing 9 from both sides in the
second one. (recognize equivalence, 8th grade student)
Table 1 Proportion of
students at each grade level Grade Level
who used each strategy use Strategy 6th 7th 8th
on the equivalent equations
problem Recognize equivalence 0.12 0.17 0.34
Solve and compare 0.39 0.33 0.25
Answer after equal sign 0.11 0.11 0.11
Other 0.31 0.25 0.27
No response/Don’t know 0.08 0.14 0.02
Yes if you substitute 8 for n, each answer will be equal and make sense [student shows
computations for determining the value of n and for checking that the value satisfies the
second equation]. (solve and compare, 6th grade student)
Yes because it has to be to get a 31 in both answers. (answer after equal sign, 6th grade
student)
As seen in Table 1, there was also a substantial number of students who left their
answer sheets blank, simply wrote that they did not know, or used idiosyncratic
strategies (i.e., strategies that could not be determined or that were insufficiently
frequent to warrant their own codes, both of which were classified as other strate-
gies). To some extent, the large proportion of strategies in the other category may
not be too surprising: with the exception of one 8th grade algebra class, the stu-
dents’ exposure to algebra, in general, and equivalent equations, in particular, had
been minimal at best. (Alternatively, it is encouraging to see that so many students—
students who used the recognize equivalence or solve and compare strategies—were
able to engage with the problem in mathematically appropriate ways prior to formal
instruction in “algebra”.)
Is interpretation of the equal sign associated with performance on problems that
involve equations? More specifically, do students who hold a relational view of this
symbol perform better than their peers who do not hold such a view on a problem for
which they must judge the equivalence of two equations? To find out, we examined
the relationships among grade level (6, 7 or 8), equal sign interpretation (relational
or not), and performance on the equivalent equations problem. We first consider
students’ judgments about whether the two equations had the same solutions or not,
and then we consider their strategies for arriving at those judgments.
As seen in Fig. 6, students who provided a relational interpretation of the
equal sign were more likely to judge that the two equations had the same solu-
tions than were students who did not provide a relational interpretation. The ef-
fect of equal sign interpretation was significant when controlling for grade level,
β̂ = −1.24, z = −4.15, Wald(1, N = 251) = 17.23. In addition, the effect of grade
level was significant when controlling for equal sign interpretation, Wald(2, N =
251) = 9.00, and was accounted for by a significant linear trend, β̂ = 0.70, z = 2.98,
Wald(1, N = 251) = 8.89.
As seen in Fig. 7, students who provided a relational interpretation for the equal
sign were also more likely to use the recognize equivalence strategy than were stu-
dents who did not provide a relational interpretation. The effect of equal sign in-
Fig. 6 Proportion of sixth-, seventh-, and eighth-grade students in each equal sign understanding
category who answered the equivalent equations problem correctly
Fig. 7 Proportion of sixth-, seventh-, and eighth-grade students in each equal sign understanding
category who used the recognize equivalence strategy on the equivalent equations problem
terpretation was significant when controlling for grade, β̂ = −1.33, z = −4.01,

Wald(1, N = 251) = 16.10. In addition, the effect of grade level was significant
when controlling for equal sign interpretation, Wald(2, N = 251) = 12.11, and was
accounted for by a significant linear trend, β̂ = 0.93, z = 3.17, Wald(1, N = 251) =
10.05. It is worth noting that a subset of students who used the recognize equiv-
alence strategy (24%) displayed an operational view of equality on the equal sign
interpretation item. Thus, different problem contexts appear to activate or draw on
different aspects of students’ knowledge.
In sum, students’ understanding of the equal sign was associated with their per-
formance on the equivalent equations problem, both in terms of their judgments for
the problem and the strategies they used to arrive at those judgments. Thus, students
Fig. 8 Students’ interpretations of a literal symbol
who demonstrated a relational understanding of the equal sign appeared to use this
understanding in determining that the two equations had the same solutions.
Interpretation of a Literal Symbol
We turn now to students’ interpretations of a literal symbol (n) used as a variable in

a mathematical expression. The most common meaning students at all three grade
levels provided was that of a variable—the literal symbol could represent more than
one value. The following responses are representative of the multiple values code:
The symbol is a variable, it can stand for anything. (6th grade student)
A number, it could be 7, 59, or even 363.0285. (7th grade student)
That symbol stands for x which stands for a number that goes there. (8th grade student)
In the final example above, it is interesting that the student apparently felt the need
to replace n with x, the latter of which represents a number; this response may
be an artifact of school mathematics in which the prototypical literal symbol is x.
Not surprisingly, as seen in Fig. 8, the proportion of students providing a correct
interpretation (i.e., multiple values) differed across the grades, Wald(2, N = 372) =
22.58, increasing from fewer than 50% of students in grade 6 to more than 75%
of students in grade 8. This improvement across grades was accounted for by a
significant linear trend, β̂ = 0.91, z = 4.71, Wald(1, N = 372) = 22.27.
It is also worth noting the relatively large proportion of 6th grade students whose
responses were categorized as either other or no response/don’t know. One possible
explanation for the nature of these students’ responses relates again to the curricu-
lum: the first formal introduction of the concept of variable does not occur until the
7th grade (in the Connected Mathematics curriculum), thus the 6th grade students
may lack experience with literal symbols used as variables in algebraic expressions.
Fig. 9 Students’ judgments on the which is larger problem
Performance on the which Is Larger Problem
Figure 9 displays students’ judgments to the question prompt (i.e., Can you tell
which is larger, 3n or n + 6?), and Table 2 displays the justifications students pro-
vided for their judgments. In some cases, students provided only a judgment and not
a justification for their judgment (justifications in these cases were assigned to the
no response/don’t know category).
In the 6th grade, the majority of students appeared either unable to provide a
justification or to provide an idiosyncratic justification (see Table 2). Relative to the
6th grade students, the 7th and 8th grade students were more likely to respond with
a correct justification that focused on the fact that the literal symbol could take on
multiple values. The following justifications are typical of variable responses:
No because you don’t know what n is. (6th grade student)
No, because n is not a definite number. If n was 1, 3n would be 3 and n + 6 would be 7,
but if n was 100, 3n would be 300 and n + 6 would be 106. This proves that you cannot tell
which is larger unless you know the value of n. (8th grade student)
Although the coding category of single value appeared in fewer than 5% of the
responses at each grade level, it is worth noting, because these students at least
seemed to recognize that the literal symbol represents a number. In such cases,
the students tested a specific number and based their judgments on the results of
their computations. Likewise, the coding category of operation was also rare. Based
on prior work (e.g., Greer 1992), we expected that some students would focus on
the operation (for example, one seventh-grade student stated, “Yes, n + 6 is bigger
because they +”.). Although such responses did occur in 7th and 8th grades, they
represented fewer than 10% of the responses at each grade level.
Is holding a multiple-values interpretation of n associated with performance on
the which is larger problem? That is, were students who interpreted the literal sym-
bol (item 3) as a variable more likely to answer “can’t tell” and to provide a correct
Table 2 Proportion of
students at each grade level Grade Level
who provided each type of Justification 6th 7th 8th
justification for the which is
larger problem Variable 0.11 0.51 0.60
Single Value 0.03 0.05 0.04
Operation 0.00 0.05 0.09
Other 0.42 0.15 0.16
No response/Don’t know 0.45 0.23 0.11
Fig. 10 Proportion of sixth-, seventh-, and eighth-grade students in each literal symbol interpre-
tation category who provided a correct judgment for the which is larger problem
justification than were students who did not provide a multiple-values interpreta-
tion? To find out, we examined the relationships among grade level (6, 7 or 8),
literal symbol interpretation (multiple values or not), and performance on the which
is larger problem. We first consider students’ judgments about whether 3n or n + 6
is larger, and then we consider their justifications.
As seen in Fig. 10, students who provided a multiple-values interpretation were
indeed more likely than their peers who did not provide a multiple-values interpre-
tation to answer “can’t tell” on the which is larger problem. The effect of having
a multiple-values interpretation was significant when controlling for grade level,
β̂ = −0.97, z = 2.22, Wald(1, N = 122) = 4.90. In addition, the proportion of stu-
dents who correctly answered “can’t tell” increased across the grade levels. The
effect of grade level on performance was significant when controlling for literal
symbol interpretation, Wald(2, N = 122) = 11.54, and was accounted for by a sig-
nificant linear trend, β̂ = 1.27, z = 3.34, Wald(1, N = 122) = 11.13.
Lastly, as seen in Fig. 11, students who provided a multiple-values interpretation
of the literal symbol were also more likely than were students who did not provide a
multiple-values interpretation to provide correct justifications on the which is larger
problem. The effect of having a multiple-values interpretation was significant when
Fig. 11 Proportion of sixth-,

seventh-, and eighth-grade
students in each literal
symbol interpretation
category who provided a
correct justification for the
which is larger problem
controlling for grade, β̂ = −0.95, z = −2.05, Wald(1, N = 122) = 4.21. Further,

the proportion of students who provided a correct justification increased across the
grade levels. The overall effect of grade was significant when controlling for lit-
eral symbol interpretation, Wald(2, N = 122) = 13.79, and was accounted for by a
significant linear trend, β̂ = 1.61, z = 3.65, Wald(1, N = 122) = 13.32. It is worth
noting that a subset of students who provided a correct justification on the which is
larger item (20%) did not provide a multiple values interpretation on the variable
understanding item. Thus, as for the equal sign items, different problem contexts
appear to activate or draw on different aspects of students’ knowledge.
In sum, understanding of variable was associated with performance on the which
is larger problem, in terms of both students’ judgments about which quantity was
larger and their justifications for their judgments. Thus, students who had a multiple-
values interpretation of a literal symbol appeared to use this understanding in deter-
mining that one cannot tell whether 3n or n + 6 is larger.
Discussion
The focus of this chapter was on middle school students’ understandings of the ideas
of equivalence and variable, their performance on problems that require use of these
ideas, and the relationship of their understanding to performance. In this section, we
briefly discuss the results and their implications for mathematics education.
Equivalence Results
The finding that many students hold an operational view of the equal sign is not par-
ticularly surprising, given that similar results have been found in previous research
(e.g., Falkner et al. 1999; Kieran 1981; McNeil and Alibali 2005; Rittle-Johnson
and Alibali 1999). Although our results suggest that students’ views of the symbol
become more mathematically sophisticated (i.e., view the equal sign as a relation
between two quantities) as they progress through middle school, the majority of
students at each grade level continued to exhibit less sophisticated views of the
equal sign (e.g., as a “do something” symbol). This result is troublesome in light of
our finding that students who have a relational view of the equal sign outperformed
their peers who hold alternative views on a problem that requires use of the idea of
mathematical equivalence. We report elsewhere that middle school students’ views
of the equal sign also play a role in their success in solving algebraic equations and
simple algebra word problems (Knuth et al. 2006). Taken together, such results sug-
gest that an understanding of equivalence is a pivotal aspect of algebraic reasoning
and development. Consequently, students’ preparation for and eventual success in
algebra may be dependent on efforts to enhance their understanding of mathematical
equivalence and the meaning of the equal sign.
Yet, equivalence is a concept traditionally introduced during students’ early ele-
mentary school education, with little instructional time explicitly spent on the con-
cept in the later grades. In fact, teachers generally assume that once students have
been introduced to the concept during their elementary school education, little or no
review is needed. Some previous work at the elementary school level has focused on
promoting a relational view of the equal sign (e.g, Carpenter et al. 2003); however,
there is little explicit attention to this concept in the later grades. This lack of at-
tention may explain, in large part, why many students continue to show inadequate
understandings of the meaning of the equal sign in secondary school and even into
college (e.g., McNeil and Alibali 2005; Mevarech and Yitschak 1983). Further exac-
erbating students’ opportunities to develop their understanding of equivalence is the
fact that very little attention is paid to the concept in curricular materials—despite
the ubiquitous presence of the equal sign. Moreover, analyses of middle school cur-
ricular materials suggest that relational uses of the equal sign are less common than
operational uses (McNeil et al. 2004). This pattern of exposure may actually condi-
tion students to favor less sophisticated and generalized uses of equivalence (such
as “operations equals answer”).
Variable Results
The findings regarding students’ views of literal symbols are, in general, more pos-
itive than the results of previous research (cf. Küchemann 1978). In particular, a
substantial proportion of students interpreted a literal symbol as representing more
than one value, increasing from approximately 50% of the 6th grade students to
more than 75% of the 8th grade students. Students’ use of this knowledge suggests,
however, that knowledge of the concept of variable may be somewhat fragile, par-
ticularly among 6th -grade students, who were largely unable to correctly answer the
which is larger problem. Yet, those students who did provide a multiple-values in-
terpretation of a literal symbol were more likely than their peers to not only use this
understanding to determine that one could not tell whether 3n or n + 6 is larger,
but also to provide a correct justification for why one could not tell which is larger.
These latter results highlight the importance of fostering a multiple-values interpre-
tation of literal symbols and suggest that efforts to foster such an interpretation will
likely contribute to students’ preparation for algebra and algebraic ways of thinking.
In contrast to the treatment of equivalence, the concept of variable is one that
receives explicit instructional and curricular attention in middle school (7th grade in
the Connected Mathematics curriculum). It may be the case, however, that providing
students with opportunities to meaningfully encounter literal symbols in ways that
support the development of a multiple-values understanding at an earlier age may
be beneficial in terms of their preparation for and eventual success in algebra (a
perspective shared by others, e.g., Carraher et al. 2000). Students often encounter
literal symbols during their elementary school education (e.g., 8 + 3 = 1, 3+? = 7),
however, the nature of such exposure may lead students to consider literal symbols
in less sophisticated and mathematically powerful ways (e.g., as specific numbers).
Concluding Remarks
If a goal of mathematics education reform is to better prepare all students for suc-
cess in algebra, then the nature of students’ “pre-algebraic” mathematical experi-
ences must lay the foundation for more formal study of algebra. Much of this foun-
dation can be laid as well as strengthened in the middle school grades—the time
period linking students’ arithmetic and early algebraic reasoning and their devel-
opment of increasingly complex, abstract algebraic reasoning. In this chapter we
presented results concerning students’ understanding of two fundamental algebraic
ideas—equivalence and variable—and the relationship of their understanding to per-
formance on problems that require use of these two ideas. It is our hope that these
results will inform the work of both teachers and curriculum developers, so that they
can each provide more opportunities for students to develop their understanding of
these core concepts.
References
Bastable, V., & Schifter, D. (2008). Classroom stories: Examples of elementary students engaged
in early algebra. In J. Kaput, D. Carraher, & M. Blanton (Eds.), Algebra in the Early Grades
(pp. 165–184). Mahwah, NJ: Erlbaum Associates.
Bednarz, N., Kieran, C., & Lee, L. (1996). Approaches to Algebra: Perspectives for Research and
Teaching. Dordrecht: Kluwer Academic.
Carpenter, T., & Levi, L. (1999). Developing conceptions of algebraic reasoning in the primary
grades. Paper presented at the annual meeting of the American Educational Research Associa-
tion, Montreal (Canada), April, 1999.
Carpenter, T., Franke, M., & Levi, L. (2003). Thinking Mathematically: Integrating Arithmetic &
Algebra in Elementary School. Portsmouth, NH: Heinemann.
Carraher, D., Brizuela, B., & Schliemann, A. (2000). Bringing out the algebraic character of arith-
metic: Instantiating variables in addition and subtraction. In T. Nakahara & M. Koyama (Eds.),
Proceedings of the Twenty-Fourth International Conference for the Psychology of Mathematics

Education, Hiroshima (Japan), July, 2000 (pp. 145–152).
Falkner, K. P., Levi, L., & Carpenter, T. P. (1999). Children’s understanding of equality: A foun-
dation for algebra. Teaching Children Mathematics, 6(4), 56–60.
Greer, B. (1992). Multiplication and division as models of situations. In D. Grouws (Ed.), Hand-
book of Research on Mathematics Teaching and Learning (pp. 276–295). New York: Macmil-
lan.
Kaput, J. (1998). Transforming algebra from an engine of inequity to an engine of mathematical
power by “algebrafying” the K-12 curriculum. Paper presented at the Algebra Symposium,
Washington, DC, May, 1998.
Kaput, J., Carraher, D., & Blanton, M. (2008). Algebra in the Early Grades. Mahwah, NJ: Erlbaum
Associates.
matics, 12(3), 317–326.
Knuth, E., Stephens, A., McNeil, N., & Alibali, M. (2006). Does understanding the equal sign
matter? Evidence from solving equations. Journal for Research in Mathematics Education,
37(4), 297–312.
Küchemann, D. (1978). Children’s understanding of numerical variables. Mathematics in School,
7(4), 23–26.
Lacampagne, C., Blair, W., & Kaput, J. (1995). The Algebra Initiative Colloquium. Washington,
DC: U.S. Department of Education & Office of Educational Research and Improvement.
Ladson-Billings, G. (1998). It doesn’t add up: African American students’ mathematics achieve-
ment. In C. E. Malloy & L. Brader-Araje (Eds.), Challenges in the Mathematics Education
of African American Children: Proceedings of the Benjamin Banneker Association Leadership
Conference (pp. 7–14). Reston, VA: NCTM.
MacGregor, M., & Stacey, K. (1997). Students’ understanding of algebraic notation: 11–15. Edu-
cational Studies in Mathematics, 33, 1–19.
McNeil, N., Grandau, L., Stephens, A., Krill, D., Alibali, M. W., & Knuth, E. (2004). Middle-
school students’ experience with the equal sign: Saxon Math does not equal Connected Math-
ematics. In D. McDougall (Ed.), Proceedings of the Twenty-Sixth Annual Meeting of the North
American Chapter of the International Group for the Psychology of Mathematics Education.
Toronto, October, 2004 (pp. 271–275).
McNeil, N. M., & Alibali, M. W. (2005). Knowledge change as a function of mathematics experi-
ence: All contexts are not created equal. Journal of Cognition and Development, 6, 285–306.
Mevarech, Z., & Yitschak, D. (1983). Students’ misconceptions of the equivalence relationship. In
R. Hershkowitz (Ed.), Proceedings of the Seventh International Conference for the Psychology
of Mathematics Education. Rehovot (Israel), July, 1983 (pp. 313–318).
National Council of Teachers of Mathematics (1997). A framework for constructing a vision of
algebra: A discussion document. Unpublished manuscript.
matics. Reston, VA: NCTM.
National Research Council (1998). The Nature and Role of Algebra in the K-14 Curriculum. Wash-
ington, DC: National Academy Press.
Philipp, R. (1992). The many uses of algebraic variables. Mathematics Teacher, 85, 557–561.
RAND Mathematics Study Panel (2003). Mathematical Proficiency for All Students: Toward a
Strategic Research and Development Program in Mathematics Education. Santa Monica, CA:
RAND.
Rittle-Johnson, B., & Alibali, M. W. (1999). Conceptual and Procedural Knowledge of mathemat-
ics: Does one lead to the other? Journal of Educational Psychology, 91, 175–189.
Steinberg, R., Sleeman, D., & Ktorza, D. (1990). Algebra students’ knowledge of equivalence of
equations. Journal for Research in Mathematics Education, 22(2), 112–121.
Usiskin, Z. (1988). Conceptions of school algebra and uses of variables. In A. F. Coxford (Ed.),
The Ideas of Algebra, K-12 (pp. 8–19). Reston, VA: NCTM.
An Approach to Geometric and Numeric
Patterning that Fosters Second Grade Students’
Reasoning and Generalizing about Functions
and Co-variation
Joan Moss and Susan London McNab
Abstract In this chapter, we present illustrations of second grade students’ rea-

soning about patterns and two-part function rules in the context of an early algebra
research project that we have been conducting in elementary schools in Toronto and
New York City. While the study of patterns is mandated in many countries as part of
initiatives to include algebra from K-12, there is a plethora of evidence that suggests
that the route from patterns to algebra can be challenging even for older students.
Our teaching intervention was designed to foster in students an understanding of lin-
ear function and co-variation through the integration of geometric and numeric rep-
resentations of growing patterns. Six classrooms from diverse urban settings partici-
pated in a 10–14-week intervention. Results revealed that the intervention supported
students to engage in functional reasoning and to identify and express two-part rules
for geometric and numeric patterns. Furthermore, the students, who had not had for-
mal instruction in multiplication prior to the intervention, invented mathematically
sound strategies to deconstruct multiplication operations to solve problems. Finally,
the results revealed that the experimental curriculum supported students to transfer
their understanding of two-part function rules to novel settings.
Introduction
The study of patterns is now commonplace in elementary school curricula in many

countries, arising out of initiatives to include algebra from Kindergarten through
J. Moss ()
Department of Human Development and Applied Psychology, Ontario Institute for Studies
in Education, University of Toronto, Toronto, Canada
S. London McNab
Ontario Institute for Studies in Education, University of Toronto, Toronto, Canada

278 J. Moss and S. London McNab
Grade 12 (e.g., Noss et al. 1997; Ontario Ministry of Training and Education 2005;
Sasman et al. 1999; Warren 2000). The National Council of Teachers of Mathemat-
ics (NCTM) advocates that patterns should be taught from the first years of school-
ing with the expectation that students, as early as second grade, should be able to
“analyze how both repeating and growing patterns are generated”, and by the end of
fifth grade should be able to “represent patterns and functions in words, tables and
graphs” (NCTM 2000). It has been suggested that patterns can: (1) support students
to understand the dependent relations among quantities that underlie mathemati-
cal functions (e.g. Carraher et al. 2008; Ferrini-Mundy et al. 1997; Mason 1996;
Lee 1996); (2) serve as a concrete and transparent way for young students to begin
to grapple with abstraction and generalization (Watson 2000; Noss and Hoyles 1996;
Kieran 1992); and (3) support students to develop the language of conjecture and
proof in communicating their reasoning about pattern rules (Kuchemann 2008;
Moss and Beatty 2006b).
Developmentally, the inclusion of patterns seems to fit well with mathematics
learning in the early years. We know that young children are fascinated with patterns
(Ginsburg and Seo 1999) and are capable not only of noticing patterns but also of
using this skill naturally to make sense of their world (Greenes et al. 2001).
However, the potential of pattern work to support algebra learning has not been
substantially realized (e.g. Carraher et al. 2008; Dorfler 2008). An extensive lit-
erature review on patterning research—conducted primarily with older students—
reveals that without specific targeted pedagogical supports even older students have
significant difficulty finding algebraic rules for patterns, strongly suggesting that
the route from perceiving patterns to finding useful rules and algebraic represen-
tations is difficult (English and Warren 1998; Kieran 1992; Lannin et al. 2006;
Lee and Wheeler 1987; Orton and Orton 1999; Stacey and MacGregor 1999;
Steele and Johanning 2004).
One challenge in moving from pattern study to algebra is the tendency of students
to use additive strategies for identifying and describing patterns—that is, to focus
on the variation within a single data set rather than on the relationship between two
data sets (e.g., Orton et al. 1999; Rivera 2006; Rivera and Becker 2007; Warren
2006). While this recursive approach allows students to predict the “next” position
of a pattern, it does not support co-variational thinking about a relationship across
data sets to find the underlying function rule. As well, even when students begin
to grasp two-part pattern rules, they often use incorrect proportional thinking or
“whole object reasoning” to make predictions about the number of elements in a
far position of a sequence (e.g., English and Warren 1998; Lee 1996; Orton 1997;
Stacey 1989).
Over the last several years, however, there has been an increasing number of
accounts of middle school students who, as part of dedicated instructional inter-
ventions, demonstrated the ability to work constructively with patterns. For exam-
ple, in 2008, the journal ZDM, Mathematics education published a special issue
that focused on patterns and generalizing problems (Becker and Rivera 2008). Re-
search reported in this special issue by Amit and Neria, Radford, Rivera and Becker,
and Steele analyzed strategies that middle school students employed in their pattern
An Approach to Geometric and Numeric Patterning 279
work. These studies had in common the use of an analytic framework (e.g. Stacey
1989; Lannin 2005; English and Warren 1998) that captured the progression of stu-
dents’ reasoning. They noted that students working on generalizing problems began
by using an additive or recursive approach, then, if they were able, switched to ex-
plicit or functional reasoning to find far positions and general rules. While there
were many promising results reported in these articles, there were also indications
of limited strategy use. Amit and Neria (2008), who studied the patterning prob-
lem solving strategies of 139 gifted 11- to 13-year-olds, went so far as to conclude
that it is only advanced mathematics students who are able to learn to generalize. In
their words, “Because of the higher-order thinking involved in generalization, such
as abstraction, holistic thinking, visualization, flexibility and reasoning, the ability
to generalize is a feature that characterizes capable students and differentiates them
from others.”
However, we along with others believe that it is not patterns per se, but the ways
that patterns are presented that may limit students’ ability to engage in the higher
order thinking that characterizes generalization. While there has been less research
conducted with very young children, our present study with Grade 2 students in
diverse urban settings joins the work of other researchers (e.g. Carraher et al. 2006,
2008; Carraher and Earnest 2003; Cooper and Warren 2008; Mulligan et al. 2004;
Mulligan and Mitchelmore 2009) to examine the potential of pattern work to support
algebraic thinking in the early elementary school years.
Our Project
Over the last five years, we have been investigating new approaches to pattern teach-
ing and learning that support students to forge connections amongst different repre-
sentations of pattern. Our goal is to promote multiple ways of working with patterns
(Mason 1996), and to foster what Lee (1996) has termed “perceptual agility—the
ability to see multiple patterns coupled with a willingness to abandon those that do
not prove useful for rule making” (p. 95). Our project to date has included interven-
tion studies in 20 inner city elementary school classrooms (e.g., Beatty and Moss
2006a, 2006b; Beatty et al. 2006; London McNab and Moss 2004; Moss 2005; Moss
and Beatty 2006a, 2006b; Moss et al. 2008). Further, it has served as a basis for a
school district-wide professional development intervention (Beatty et al. 2008). This
research began with a series of studies in second grade classrooms; it is the methods
and data from these Grade 2 studies that we present in this chapter.
Our Approach: Theoretical
The predominant theoretical inspiration for our research on patterning emanated

from the theories about mathematical development of Case and colleagues (Case
and Okamoto 1966). Case and his colleagues’ previous work in mathematics devel-
opment for number sense in the domains of whole number (e.g. Griffin and Case
1997) and rational number (e.g. Moss and Case 1999; Moss 2004) offered a model
for the integration of numeric and spatial schemes that we paralleled in linking nu-
meric and geometric representations of patterns. A central tenet of the instructional
design of Case et al. is the focus on the development of students’ visual/spatial
schemes. The theoretical proposal is that the merging of the numerical and the vi-
sual provides the students with a new set of powerful insights that can underpin not
only the early Learning of a new mathematical domain but subsequent Learning as
well (Case and Okamoto 1966; Kalchman et al. 2001). As we elaborate below, our
experimental patterning curriculum was designed to support students to forge con-
nections between visual/spatial patterns in the form of geometric growth sequences
and numeric patterns embedded in “Guess my rule” games. Our conjecture was that
the merging of these two types of patterns would serve as a foundation to support
students to gain an initial understanding of linear functions. To test this conjecture,
we designed a lesson sequence that was pilot-tested, revised and refined over a two-
year period, and implemented in 6 different Grade 2 classrooms.
Context and Students
The 7- and 8-year old students in our study were from intact classrooms of between
20 and 22 students each, in urban settings in Toronto and New York City. These stu-
dents represented diverse populations and a range of math competency. The class-
rooms were chosen because of the teachers’ interest in learning more about this
new approach and in involving their children in this study. Overall, the classrooms
seemed to have in common an invitational sense of welcoming student contribution;
the students were all accustomed to expressing their thoughts and reasoning in math,
as in all subjects.
All of the students had previous experience with repeating patterns as part of the
early years math curricula; however, none of the students had worked with growing
patterns. Importantly, there had been no formal instruction in multiplication in any
of the classrooms prior to the intervention. Although the activities in the intervention
could be approached through multiplication, at no time was multiplication formally
taught.
The length of the intervention varied from 10 to 14 lessons of about forty minutes
each. In four of these classrooms, the interventions were taught by the classroom
teachers with the help of research assistants; in the other classrooms, the interven-
tions were taught by the first or second author with the assistance of the classroom
teacher. It is important to note that there were research assistants in the classroom,
who were able to work with Small groups or individual children, and that in some
of the classrooms math was taught to only half the class at a time.
Instructional Sequence
Visual Representation: Geometric Growing Patterns
The lessons began by presenting students with the first three positions in a geometric
growing pattern. These patterns were made of square tiles set out in arrays that grew
by a given coefficient. To enable students to keep track of the ordinal position num-
ber of these tile patterns, position number cards were placed below the geometric
arrays that represented that position of the pattern.
This clarified for students the functional relationship between the position num-
ber (independent variable) and the number of elements in each position (dependent
variable). So, for example, for a pattern representing the relationship described by
the equation y = 3x (please see Fig. 1), students could easily connect the position
number card “1” to the single row of three square tiles, the position number “2” to
the 2 rows of three square tiles (6 tiles), the position number “3” to the 3 rows of
three square tiles (9 tiles), and so on.
The initial challenges that the teacher posed were designed to focus students’ at-
tention on the relationship between the position number and the number of elements
in each position, through the geometric configurations of the tile arrays. Referring
once more to the pattern in Fig. 1, in the first lessons, the teacher’s questions to
the students followed a specific sequence: If this pattern keeps growing in the same
way, what would the next position look like? How many blocks would there be in
the next position? What would the 10th position look like? How many blocks in
the 10th ? What about the 100th position? In subsequent lessons after the students
had experience with the function machine activity “Guess my rule?” (see below),
the teacher would go on to ask, What if you had any position? What could the rule
be?
Next, two-part functions were introduced geometrically. To demonstrate the con-
stant, a fixed number of tiles was placed at the top of the array; this configuration
of tiles remained the same from position to position, while the array grew multi-
plicatively by one row for each new position. Because of the spatial representation
of the constant as tiles that jutted out from the array (please see Fig. 2), the students
began to refer to the constant as the “bump”, and we made a deliberate decision to
encourage their use of this natural language.
Fig. 1 Position number cards

Fig. 2 Composite functions:

the “bump”
As the lessons progressed, students built their own patterns and challenged class-
mates to make conjectures for general rules.
Numeric Representations: Function Machine
We interspersed these visually based pattern lessons with numeric-based lessons

that incorporated function machine activities (Carraher and Earnest 2003; Ruben-
stein 2002; Willoughby 1997). Please see Fig. 3a for an example of a function ma-
chine. As in the geometric lessons, in the first series of function machine activities,
we focused on one-step multiplicative rules. To begin, the teacher led the activi-
ties; then the students took turns creating functional rules (e.g. “double the number
and add 3 more”) to challenge their classmates in the “Guess my rule” game. The
teacher modeled the use of a T-table to record the non-sequential pairs of input and
output numbers; please see Fig. 3b for an example of the Function machine T-table.
Pairs of students generated between 3 and 5 examples of non-sequential pairs of
input and output numbers, as clues to their rule. The children who were solving the
challenges given to them by their classmates used T-tables to record the input and
output numbers, and their iterative conjectures for what the rule might be. It was
important that the numeric clues were non-sequential to allow students to focus on
the “across” (on a T-table) or functional rule, rather than on the “down” pattern or
“what comes next” strategy identified as interfering with functional generalizations.
Further, the T-tables were used only to record the non-sequential clues in the “Guess
my rule” game, but not to generate further pairs of values as is often done in many
classrooms.
Because the children had not yet been taught multiplication, the coefficients we
initially presented were confined to what we determined to be arithmetically man-
ageable numbers—2, 3, 5 and 10—that they would have practiced as skip-counting
in first grade. However, there were no such constraints on the numbers the students
could choose when they were creating their own rules for the function machine,
giving them the opportunity to experiment with even difficult or tricky arithmetic if
they chose.
Fig. 3 (a) Function machine.

(b) Function machine T-table
(a)
(b)
Integration Activities: Pattern Sidewalk
Both the geometric and the numeric activities offered students a chance to con-
sider the idea of co-variation and function rules. The geometric activities specifically
highlighted the direct connection between the position number and the structure of
the corresponding arrays and number of tiles in each position. The function ma-
chine activities illuminated the idea of explicit rather than recursive rules. In order
to integrate these two complementary approaches within a non-sequential presenta-
tion and exploration of geometric patterns, we designed what became known as a
“pattern sidewalk”. This is a large counting line placed on the classroom floor, with
ordinal position numbers on each section of the sidewalk, from 1 to 10.
Fig. 4 Pattern sidewalk
The pattern sidewalk activities were geometric, but paralleled those of the nu-
meric function machine game. First, the teacher would build one position of a pat-
tern that conformed to a secret rule on, for example, the third section of the side-
walk, and ask the children to consider a possible pattern rule that would fit with
that configuration in that position. Then the teacher would build another position of
the same rule on any other non-adjacent section of the sidewalk, for example, the
seventh pattern position on the seventh section of the sidewalk. At this point the
students were given the opportunity to revise their initial rule conjecture. Finally,
another pattern position would be shown to the students, on its appropriate section
of the sidewalk and students were allowed to guess the rule, and asked to show their
guess by building another position of the pattern correctly on a new section of side-
walk. The students were then invited to take on the role of teacher and present their
own challenges to their classmates.
These three components—the geometric pattern building, the function machine
and the pattern sidewalk—comprised the main elements of the lesson sequence in
all cases, and were introduced to the students in the specific order described above.
However, as the students gained experience of increasingly complex patterns, the
teacher/researcher would revisit these different components, adding new elements
such as moving from one-step to two-step functions.
Role of the Teacher
Another specific feature of this instruction that needs to be highlighted is the very
particular ways in which the teachers engaged with the students, and the prompts
and foci they adopted in their interactions with the children. Aligned with Radford’s
“theory of knowledge objectification” (Radford 2008), we shaped both the order
of activities and the teachers’ prompts to help students notice and make sense of
the ways in which patterns stay the same and the ways in which they change, thus
moving students towards generalization and algebraic reasoning. Specifically, the
teachers focussed the students’ attention and grounded their perceptions of change
in the curriculum’s sequenced arrays of tiles that grew by a multiplicative factor;
the teacher also grounded the students’ perception of what stays the same by draw-
ing their attention to those configurations of tiles outside the arrays which remained
constant for each position of a pattern. The teachers also carefully supported the
children’s learning by helping them draw connections between the idea of rules, as
they are experienced in “Guess my rule” function machine activities, and the pos-
sibility that geometric pattern growth can also be predicted by a rule. Further, the
teachers focused the children’s attention on the relationship between the position
number cards and the number of elements in, and structure of, the corresponding ar-
ray, thus supporting these young students’ emerging understandings of co-variation.
Procedures and Measures: Grade 2 Interventions
In order to assess the potential of the intervention, we collected data from many
different sources. Our major analyses were qualitative and descriptive, based on
classroom artifacts, field notes, transcripts of videotaped classroom lessons and ad
hoc interviews with students during the lessons. In addition, we gave each of the
students in all of the research classrooms a short pre-test interview, that was ad-
ministered again at the end of the intervention as a post-test, consisting of patterns
in different representations. In keeping with the literature on patterning discussed
above, this pre-/post-test was designed to assess changes in students’ abilities to
find “near” and “far” positions of patterns (e.g., Lannin 2003), to identify whether
students relied on recursive strategies or functional reasoning and finally to assess
students’ abilities to find and express general pattern rules. At the end of the second
year, we introduced an additional assessment in which we interviewed students to
look specifically for transfer in their reasoning to a novel context.
Results
The results presented here focus on four general areas: the way students developed
their reasoning about pattern rules, the explicit aim of our research; students’ con-
structed understandings of multiplication, an implicit research question; the use of
zero as both co-efficient and position number, an unexpected result; and the transfer
of understandings to a novel context.
Finding Rules for Patterns and Generating Patterns Based on

Given Rules
Our overall analyses for each year of the study revealed that students made signif-
icant gains in their ability to discern function rules for geometric growth patterns
and reciprocally/conversely could also build patterns based on given rules. In con-
trast to findings from other studies that reveal the pervasiveness of recursive reason-
ing, the students in our research classrooms used a functional approach, which was
evidenced in the way that they talked about the position number in relation to the
number of elements in a position.
Constructing a Pattern from a Rule: “A ‘number times two, plus

one’ pattern?”
The following transcript of a conversation between Ricardo and the classroom

teacher was initiated by the teacher as she walked around the class during a por-
tion of a lesson where students were building patterns based on given rules. She
asked Ricardo to build a “number times two, plus one” pattern. Ricardo, using the
square tiles from the pattern block set, built the first four positions of a pattern (that
could be described in the informal notation of this classroom as “n × 2 + 1”, placing
position number cards below each position. Each position was comprised of a row
of increasing numbers of square tiles, with one tile on top of the row. (Please see
Fig. 5.)
Our transcript begins when the teacher asks Ricardo to explain his pattern:
Ricardo: See, this is the first position. [Ricardo points to the first position of the
pattern he built and then picks up the position card that he had placed
Fig. 5 Building a composite

function
under the first position]. So [touching the row of 2 blocks], so one times
two is two, plus one [points to the block on top] is three. [Keeping this
same speech rhythm he picks up the second position card] So, two times
two is four [points to the row of four blocks] and ONE [said emphatically]
makes five.
In his explanation of the next (third) position, Ricardo makes an error. As he is
about to make the same mistake again for the fourth position, he catches himself
and corrects his explanation:
[He points to the position card on which is written a 3.] So three times
three [sic] is six and one [points to the single tile] is 7. [He repeats the
same set of actions for the fourth position—the final one he has built]) So,
four times four. . . I mean four PLUS four. . . or, two TIMES four is eight
and one makes nine.
Teacher: Well done. How many blocks would there be in the 10th position?
R: [putting his hand over his eyes in thinking position and then rapidly drop-
ping it and saying with a smile] Twenty-one.
What was notable to us about Ricardo’s explanation was his clear understand-
ing of the co-variation of the position number and the number of tiles. What also
was revealed in the interaction was the way that Ricardo constructed the pattern to
clearly reflect his ability to distinguish the coefficient from the constant. Finally,
this exchange also demonstrates this young student’s fluency in being able to use
his understanding of the function rule not only to build sequential pattern positions
from 1 to 4 but also to predict quickly, easily and accurately the number of tiles that
would be required for the 10th position—a far position.
Finding a Rule for a Given Pattern: “Position number times three,

plus one”
In this transcript, two students, Zoya and Marie, are sitting at a table examining the
first three positions of a geometric pattern (y = 3x + 1) to determine the pattern rule.
The figures each consisted of a planar column of yellow hexagonal pattern blocks
with a single green triangle placed on the top, that grew by three yellow blocks each
time (Fig. 6).
The students stare at the first three positions of the pattern:
Marie: It’s a times 3 pattern, right?
Zoya: [touches the blocks in the second pattern position] Because this is a
GROUP of 3 [separates and points to one group of 3 in the second
position, and then moves her finger in a circle around it], and this is a
GROUP of 3 [points to the remaining group of hexagons in the second
position; then she looks at the triangle and exclaims:] Wait, oh but it
can’t be, because [indicating the green triangle] it’s a whole block. So,
Fig. 6 Caterpillar pattern
never mind! [pushes groups back together, and throws up her hands;
the students seem stumped and continue to look at the pattern]
Marie: I’ve got it. [long pause as she stares at the pattern] It is number times
3. . . No, position number times 3, plus 1.
Researcher: How do you know that?
Marie: Because, here is a group of 3, so that is times 3 and 1 makes 4.
Here we see the flexibility of students who were able to discard an initial
conjecture of a pattern rule when the evidence (the built pattern) did not sup-
port their rule. We contrast the flexible reasoning of these very young students
with findings of other researchers (e.g. Lee and Wheeler 1987; Stacey 1989;
Stacey and MacGregor 1999) who all document older students’ reluctance to change
incorrect conjectures of rules in the face of contradictory evidence.
Marie offered an initial rule; Zoya immediately tried to support this conjecture
by referring to the structure of the pattern. However, in trying to “prove” this rule,
Zoya realized that the built pattern did not fit the rule, so they abandoned their initial
conjecture. Eventually, Marie does figure out the correct rule, which she expresses
in informal algebraic language as position number times 3, plus 1. Notable as well
is the “groups of” language that illustrates one of the ways in which students con-
structed their multiplicative reasoning.
Students’ Invention of Multiplication
As mentioned previously, none of the classes had been taught multiplication prior
to the patterning lessons. Perhaps amongst the most salient of our findings was the
way that the pattern activities worked to support students to construct a robust un-
derstanding of multiplication, revealed in the diversity of approaches the students
had “invented” and the deep conceptual orientation to multiplication they had con-
structed through continual experience with arrays of tiles. It appears that the arrays
in the geometric growth patterns provided the students with a visual representation
of multiplication as a set of relationships that they could construct and deconstruct.
Schliemann et al. (2001) have suggested that operations such as multiplication may
in fact be more effectively understood as functions.
As shown in the following transcripts, even arithmetically lower-achieving stu-
dents who struggled to perform some of the required calculations were nonetheless
able to use multiplication to explain their reasoning about pattern rules and the num-
ber of elements in pattern positions.
Deconstructing Multiplication: “Double the position, plus the

position”
One way in which students constructed their understanding of multiplication in the

context of patterning was through a deconstruction of the operation. In this excerpt,
taken in the context of a classroom lesson, Moni is presented with tile arrays rep-
resenting the first four positions of a y = 3x pattern in which the first position is a
row of 3 tiles, the second position is two rows of 3 tiles, and so on. She explains her
thinking about how the number of tiles in the fourth position (an array of 4 rows of
3 tiles each, or 3 columns of 4 tiles each) conforms to a general rule:
Moni: [running her finger up and down one column of 4 tiles] Here it would be
4 doubles; that would be 8. . . So when you put these two lines together
it’s 8. And here’s [indicates position number card] 4. So you double the
position, with the number.
Teacher: So, it’s the number. . . ?
Moni It’s the position [number], plus the double of the position.
It is noteworthy that Moni chose to reason in a structural way rather than counting
out the full number of tiles in her attempt to calculate the total. This type of reason-
ing typified the approach taken by many of the children. The geometric configura-
tion (array) that Moni’s explanation relied on clearly supported her deconstruction
of multiplication; 3n is decomposed into both n + 2n (It’s the position, plus the
double of the position) and 2n + n (Double the position, with the number). This
visual/spatial reference that anchors her understanding further allows her to demon-
strate correctly both the commutative property of addition and the distribution of
multiplication over addition.
Using a Structural Understanding of Multiplication to Predict Far

Positions: “It’s 40 up, and 3 to the side”
In this next transcript, Oscar was shown the first two positions of a pattern (repre-
senting the functional relationship y = 3x + 5) built with square and triangular tiles,
Fig. 7 Photo of rocket ship

pattern
designed to look like a “rocket ship” for astronauts. Each position included an array
of a number of rows of 3 square tiles each, one row of 3 tiles for each astronaut.
Below this array were two triangular tiles, positioned one on each side to look like
rocket boosters; three triangular tiles above the array formed the nose of the rocket
(Fig. 7).
After Oscar was shown the first two positions of this pattern, the researcher won-
dered if Oscar could use his understanding of the structure of the pattern to de-
termine how many blocks there would be in the 40th (a far) position (there are not
enough blocks on the table for Oscar to build it), so asks him what the pattern would
look like for 40 astronauts:
Researcher: Do 40, first.
Oscar: Forty [thinks]. Forty would be 40 up, 40 up [moving his finger along
an imaginary column on the table] and 3 to the side [moves his finger
sideways] ‘cause one astronaut is 3 blocks long.
Researcher: Oh, it’s 40 up and 3 to the side. Can you figure out how many blocks
that would be in all?
Oscar: That would mean 3 rows [sic] of 40. And 40 [counts with one finger
held up] plus 40 [counts on another finger] is 80. And another 40 is. . .
another 40 is. . . . [turns to his partner] What’s 80 plus 40? [the partner
replies, “120.”] A hundred and twenty. . . So that’s a hundred twenty.
[He now begins to put 5 triangle blocks down one at a time, very delib-
erately. He first places 2 at the base of a smaller array he has already
built, then 3 triangles far above this configuration, apparently at the
top of the imaginary much larger array that he is describing.] Then
a hundred and twenty one, a hundred and twenty two, a hundred and
twenty three, a hundred and twenty four, a hundred and twenty five.
That’s the answer. . . Can I write that answer down before I lose it?
The fact that Oscar could predict the 40th position (from only two examples)
revealed his growing understanding of multiplication as an array and the structure of
the linear functions we had been working with Further, Oscar’s reasoning illustrates
the understanding students developed of the co-variation of the position number and
of the number of elements in a given position. Finally, this short excerpt is indicative
of the kind of excitement this work generated in the students, the “big numbers” they
were willing to engage with and the kind of effort they were willing to make.
The Discovery of Zero
Our sequence of lessons required the children to move back and forth between nu-
meric and geometric expressions of function rules; the lessons also moved back and
forth between teacher and student generation of patterns and rules. In the course of
creating their own rules within both the numeric and the visual/spatial investigations
of pattern, students in different research classrooms independently made the discov-
ery of zero as a powerful mathematical idea and arithmetic tool. Below we present
examples taken from different classrooms revealing how students used zero in their
pattern designs, first as a coefficient and then as a position number.
Zero as a Coefficient: “Zero groups of 4 million is zero”
The first example comes from a classroom lesson at a time when pairs of students
were working independently to create function machine challenges for their fellow
classmates. The researcher approached two students, Clarice and Emma who had
already invented a rule and had created written pairs of input and output numbers.
Researcher: Okay, I’m going to give you an input number you don’t already have,
and can you give me an output? Ready, my input is 2.
Clarice: It’s going to be 5.
Researcher: Okay, input number is 17.
Clarice: 5, 5, 5!
Researcher: Input number is 672!
Clarice & Emma: 5, 5, 5, 5, 5! [laughing]:
Researcher: Wow. Can I ask you a question—what does that mean? How many
groups of the input number are there?
Clarice: If it’s times zero, it would always be zero. Zero groups of 4 million is
zero!
Researcher: Zero—there are just no groups of them.
Clarice: And then plus 5, so it’s always 5.
We were surprised, as there had been no prior discussion of zero in any con-
text. As this lesson progressed, Emma and Clarice had the opportunity to sit at
the function machine and present their challenge to their classmates. While many
were stumped, one student asserted, “It’s the number, minus itself, plus 5”. We
were interested to note that Emma and Clarice were flexible in being able to
accept this different rule as an expression of the same relationship, something
older students have difficulty doing (e.g., Lannin 2003; Lee 1996; Mason 1996;
Stacey 1989.)
Fig. 8 The zero-th position
Zero as a Position Number: “the zero-th position”
Whereas the two girls in the example above were excited to use zero as a trick, in the
next excerpt we see Anna, pictured in Fig. 8 with her geometric pattern that offered
no clear visual organizational clues for her classmates to identify the constant. Anna
suggested that the “zero-th” position was a “help” to them in discovering her pattern
rule:
Researcher: Can you show the class the pattern that you have built?
Anna: [holding a Small piece of paper in her hand, with a rule eon it written
by a researcher] So what our rule was, was—this is our second one—
it’s times 5 plus 3. [gestures to the pattern in general and then points
to a position card she has made with a zero and indicates the three
tiles above it] And the zero-th position helps you a lot, it gives you a
big clue. [she splays her 3 fingers as she tries to indicate the 3 tiles
that are over the “zero-th position” card]. It’s 3 [touching all 3 at
once]—this is the bump, cause the bump stays the same, but there’s
no GROUPS of 3 [according to the pattern rule]. So it’s the bump. It
helps you a lot because it identifies what the bump would look like.
So it’s like 5 [pointing to the first position configuration], and then
plus the bump. [pointing to the second position] Five and then 5, plus
the bump. [pointing to the third position] Five, and then 5, and then 5
[holding her hands over each group of 5], plus the bump.
The inclusion of zero was Anna’s and her partner’s idea. There were no position
cards with zero written on them. As she prepared the challenge for her classmates,
she had requested a blank card to make a zero-th position card. Anna was aware
that the position number always indicated how many groups there were, regardless
of how many tiles were in each group; so she determined that if she used zero as
a position number, then there would be no groups at all, isolating the constant and
making it easy to identify. Thus, she was giving her classmates what she determined
to be a big hint to finding the rule for her visually complicated pattern.
Transfer of Structure
As we observed the students over the course of the research lessons, we could see
that they were gaining fluency with geometric patterns and becoming increasingly
successful with function machine games, and integrating these features within their
work in the pattern sidewalk. However, what we could not tell through observation
was the robustness of the students’ acquired understandings of the two-part function
structure (represented formally as y = mx + b) and whether they could transfer
their new understandings to other mathematical contexts. Accordingly, at the end
of the second year of our Grade 2 interventions, we interviewed pairs of students
using a word problem—a novel context—that was a narrative representation of a
two-part function. The problem was presented only orally; there were no visual
representations, and the students were not given the opportunity to write or draw
to find the solution. There was nothing in the word problem that resembled what
they had done on patterning in the research lessons, and no verbal cues that linked
the word problem to what we had done in the classroom. The word problem is as
follows:
Charlotte really wants to buy a scooter. But she doesn’t have enough money. The
scooter she wants costs $100! From the tooth fairy, Charlotte already has saved
$10. But she decides to earn more money by walking her neighbour’s dog, Sparky.
For each day that Charlotte walks Sparky, her neighbour will pay her $5.
Circumventing Whole Object Reasoning
The first transcript comes from a post-intervention interview in which two students,
although not asked to state a function rule, clearly demonstrate their ability to dis-
cern and use a rule:
Researcher: Okay, kids. Now you really have to listen hard. I have a question for
you. I don’t have any paper or anything. [Reads the problem out loud.]
How much money will Charlotte have altogether at the end of the first
day of walking Sparky?
Mai: She. . . .
Juanita: 15.
Researcher: How do you know?
Juanita: She already has 10, and then she gets 5.
Researcher: How much money will she have altogether at the end of Day 2?
How about on the second day?
Mai: 20.
Juanita: Hey!
Researcher: Why, is that what you were going to say, Juanita?! How much money
will she have altogether at the end of Day 5?
Juanita: 20. . . 25. . . . Just a second. [counting by fives on her fingers]
Mai: 35.
Researcher: How did you get 35?
Mai: Because 5 × 5 is 25, plus 10 is 35.
Researcher. Is that how you did it, Juanita?
Juanita: Yeah.
Researcher: How much money will she have altogether at the end of Day 10?
Mai: 60.
Researcher How did you get 60?
Juanita: Well, on the 5th day is 25, and 25 and 25 is 50, plus 10 is 60.
Juanita. Same with me.
Researcher: What day would it be if Charlotte has $70 altogether?
Juanita: I think the. . . twelfth.
The inappropriate use of proportional reasoning or “whole object reasoning” in
the context of patterning problems is well documented. For this reason, the sequence
of questions in our interview progressed from asking the students how much money
Charlotte would have in 5 days, to how much she would have in 10 days. A whole
object strategy, which could be anticipated, would produce an incorrect answer of
70. That is, if 5 days equals $35, then 10 days would be double that, or $70. How-
ever, Juanita, like the majority of students in the Grade 2 research classrooms, gave
the correct answer of 60.
While Juanita and Mai had not been asked to express a rule explicitly, they ap-
peared to understand what the rule was and how to use it to predict positions of the
pattern, as evidenced in their responses.
Further, the students’ ability to correctly answer the final question in the narrative
problem (how many days would it be if Charlotte has $70) is a further indication of
their agility and robust understanding. They could use their explicit understanding
of the coefficient and constant to reason backwards, i.e. to begin with the number of
elements (money) in an unknown position and find the position (day).
Informal Algebraic Expressions of Rules in the Sparky Problem
In the excerpts below, from another classroom, the researcher gave students the
opportunity to explain their thinking and to propose a general rule for the Sparky
problem. The responses below are impressive in that these students were able to
extract the functional relationships inherent in the Sparky problem, and express them
in syncopated language (Sfard 1995). The three examples below of rules offered by
Luca, Tomas and Stella show increasing levels of abstraction of rules:
Luca: It’s counting by 5s with a 10 bump.
Tomas: Oh, I get it—it’s a groups of 5 pattern with a 10 bump because she
[Charlotte] already had 10 dollars from the tooth fairy.
Another student, Stella went on to notice that the constant was larger than the
coefficient, which had not been the Case in the geometric problems that the students
engaged with as part of the intervention. Her references to the geometric, as well as
those of the previous students, in talking about “groups of” and the “bump”, illus-
trate the crucial role of the visual/spatial representation in their ability to transfer.
Stella: It’s always the day [ordinal position number] times 5, plus 10. So there’s 10
bumps and 5 normal things, more bumps than normal things—that’s weird!
Taken together, in the order that they are presented, these three excerpts reveal
the increasing degree of formalization of the students’ explanations of rules. Our
conjecture is that the role of spatial-inspired terms like “bumps” and “normal things”
was fundamental in ensuring the abstraction required to tackle the purely numeric
Sparky pattern. We see this generalization as related to what Radford (2003) has
called algebraic contextual generalizations. Further, we concur with Radford that
adherence to conventions is not necessarily an indicator of algebraic thinking: “It
is not notations which make thinking algebraic; it is rather the way the general is
thought about” (2008, p. 84).
Discussion
The Grade 2 students in our research classrooms did not rely on recursive reasoning
in their solutions to patterning problems, nor did they use inappropriate proportional
(“whole object”) strategies, both of which have been indicated in the literature as
common problems even among older students. Rather, they appeared to develop a
fairly robust understanding of two-part function rules through their engagement with
the curriculum: they could predict how a pattern would grow, find general rules for
geometric and numeric patterns, and construct patterns based on given rules. As
well, our results revealed that the students were able to transfer their understanding
of rules to a new (narrative) context, both finding and applying a rule.
The invention of multiplication has been noted in other studies of young students
engaging with patterns; however, the diversity and quality of the approaches these
students invented seemed noteworthy. Not only did the students find mathematically
sound ways to deconstruct multiplication operations to solve problems, but some
students also, at their own initiative, experimented with the effect of using zero as
either the position number or the coefficient.
Finally, the students appeared to enjoy the lessons and seemed intrigued by the
geometric presentation of patterns. They were interested in making, justifying and
testing conjectures, were flexible in their general approach, and were excited to ex-
plore different ways of creating difficult challenges for their classmates. This con-
trasts with the concerns expressed by scholars, such as Mason (1996) and Hewitt
(1992), who noted that when geometric sequences are introduced, often students
produce a table of values from which they extract a closed form formula which they
check with only one or two figural examples. The question arises: what are the char-
acteristics of our program that may have supported these very young students in
their productive and flexible approach to patterning activities?
Analyses of our findings over the many iterations of our studies suggest a number
of distinct but overlapping factors that may have contributed to our students’ ability
to work with patterns. We draw your attention to three factors in particular. First, is
the design of the curriculum with its deliberate movement back and forth between,
and then bridging of, geometric and numeric representations of growing patterns
through the idea of “rules”. Second, and related, is the primacy given to the visual
and to the way that the curriculum design deliberately focuses students’ attention
on the spatial/geometric pattern formations. And third, inherent in the design of
the instructional sequence is the emphasis on student invention. In the sections that
follow we briefly elaborate on these factors.
The Curriculum with Its Focus on Integration
The theoretical framework that underpinned this research, and served as a heuristic
for the design of the curriculum, came from previous work of Case and colleagues
on children’s mathematical development. Specifically, we were guided by the pro-
posal of Case et al. that children’s development in a domain of mathematics (e.g.,
whole number, rational number) is underpinned by the integration of the children’s
visual schemas on the one hand, and their numeric understandings on the other, for
the mathematics domain in question (Please see Kalchman et al. 2001 for details of
this theory). Further, as we mentioned earlier, this theoretical framework helped to
establish a developmentally grounded sequence for our intervention. Students first
worked with geometric (tile array) representations and then moved on to numeric
(function machine) patterns. This sequence served to help the students to consol-
idate and extend their separate understandings in both the geometric and numeric
domains. These separate understandings were bridged for the students by the con-
cept of (function) rule which enabled the students to begin to move between the vi-
sual and the numeric with increasing flexibility. The subsequent introduction of the
pattern sidewalk, in its use of a non-sequential geometric representation of pattern,
also fostered this integration and flexibility. Our preliminary conjecture is that it was
the specific movement back and forth between these two representations, geometric
and numeric, that ultimately supported the students to gain not only flexibility with,
but also a structural sense of, two-part linear functions, thus supporting/enabling
the students to discern and understand pattern rules in contexts that were new to
them.
Prioritizing Visual Representations of Pattern
While we believe that the back and forth movement was critical to the flexible rea-
soning that students ultimately were able to demonstrate, we also suggest that the
students’ initial grounding in the visual geometric context was also significant in the
effectiveness of the curriculum. All through the lessons and interviews with students
we were made aware of how their reasoning was underpinned by their interpreta-
tions and analyses that were based on geometric figures. When probed for explana-
tions of rules, the students focused on how a pattern grew in relation to the position
number; how the addition of the constant or “bump” was related to the coefficient,
or multiplicative; how parts of the pattern changed and parts stayed the same based
on their visual configurations. Finally, even in their post-intervention explanations
of the “Sparky” narrative word problem, many students referred to the two elements
of the two-part function in geometric terms: “Oh, I get it—it’s a groups of 5 pattern
with a 10 bump because she already had 10 dollars from the tooth fairy.”
Indeed, a number of researchers have reported on the support provided by figu-
ral representations for students working with generalizing problems (e.g., Carraher
et al. 2008; Healy and Hoyles 2000; Lannin 2005; Noss et al. 1997; Rivera and
Becker 2008; Sasman et al. 1999; Stacey 1989). When visual representations are
prioritized, and students are supported to focus on the figural patterns as a way of
discerning general rules, they are better able to find, express and justify functional
rules.
However, research has also shown that, overwhelmingly, students and adults have
a strong tendency to ignore the geometric properties of figural patterns, and to focus
instead on the number of elements in the given pattern. The focus on the visual in
our program appears to have had a double advantage for students: providing a rich
context in which to analyze growth and change, and also supporting students to be
aware of covariation.
Pedagogy and Student Inventions
In the opening sections of this article we discussed the particular ways that in which
the teachers interacted with the students and how they focussed the children’s at-
tention on salient features of the instructional sequence to support the children’s
learning. It is also our proposal that another significant contribution to the success
of the intervention was the ongoing insertion into the learning sequence of the chil-
dren’s own inventions: specifically the geometric patterns they designed and also
the challenges they created for their classmates with the function machine.
As we mention in earlier sections of this chapter, inasmuch as there was a con-
tinuous movement back and forth between geometric and numeric representations
of patterns, so too was there movement back and forth from the standpoint of the
pedagogy: for example, in iterative fashion teachers modeled geometric one-step
patterns, and then students designed and presented their own patterns; teachers pre-
sented challenges with the function machine, and then students in pairs followed
their lead and designed their own challenges for their classmates. The evidence,
based on the transcripts of classroom lessons, is clear in revealing that these student-
invented challenges created excitement, interest and motivation among the children.
They also may have served other important purposes. First was the opportunity to
practice. Students in Grade 2 had little or no experience with growing patterns prior
to the intervention, and many held firmly to the belief that patterns could only repeat.
Creating their own patterns gave the grade 2 students the opportunity to discover and
experience how linear growing patterns worked or did not work. In addition, by cre-
ating their own geometric and numeric patterns, students had the time and space to
invent and then practice multiplication. Also, in the course of developing challenges
for their classmates, the students had the opportunity to take on an additional per-
spective in anticipating how their classmates might respond. In our view, this kind
of anticipation and planning added an extra dimension (metacognitive) to students’
thinking, thus enriching the learning potential of the lessons.
Concluding Thoughts
Typically, patterns are taught in the early years as repeating, with children asked to
find “what comes next”. As Blanton and Kaput point out, this limited view does not
capitalize on the potential of patterns to support later mathematics learning (Blan-
ton and Kaput 2004). A number of researchers have included a focus on young
children and patterns (e.g. Carraher et al.; Mitchelmore and Mulligan; Mulligan,
Prescott & Mitchelmore; Warren & Cooper), investigating ways of promoting alge-
braic thinking, generalizing and awareness of structure through the use of pattern-
ing. We join with these researchers in trying to illuminate the potential of pattern
work for young children. Our findings suggest that, with appropriate instruction,
the study of patterns can support students of all levels of mathematics abilities to
foster the kinds of mathematical thinking that Kieran suggests is fundamental to al-
gebraic reasoning: “analyzing relationships between quantities, noticing structure,
studying change, generalizing, problem solving, modeling, justifying, proving, and
predicting” (Cai and Knuth 2005, pg. 1)
References
Amit, M., & Neria, D. (2008). “Rising to challenge”: Using generalization in pattern problems to
unearth the algebraic skills of talented pre-algebra students. ZDM Mathematics Education, 40,
111–129.
Beatty, R., & Moss, J. (2006a). Grade 4 generalizing through online collaborative problem solving.
Paper presented at the annual meeting of the American Educational Research Association, April
2006.
Beatty, R., & Moss, J. (2006b). Connecting grade 4 students from diverse urban classrooms: virtual
collaboration to solve generalizing problems. In C. Hoyles & J. Lagrange (Eds.), Proceedings
of the Seventeenth ICMI Study: Technology Revisited. Hanoi, Vietnam.
Beatty, R., Moss, J., & Scardamalia (2006). Grade 4 generalizing through online collaborative
problem solving. Paper presented at the annual meeting of the American Educational Research
(April 2006), San Francisco, CA.
Beatty, R., Moss, J., & Bruce, C. (2008). Bridging the research/practice gap: Planning for large-
scale dissemination of a new curriculum for patterning and algebra. Paper presented at NCTM
Research Presession, Salt Lake City.
Blanton, M., & Kaput, J. (2004). Elementary grades students’ capacity for functional thinking. In
M. Jonsen Hoines & A. Hoines (Eds.), Proceedings of the 28th Conference of the International
Group for the Psychology of Mathematics Education.
Cai, J., & Knuth, E. (2005). The development of students’ algebraic thinking in earlier grades from
curricular, instructional and learning perspectives. ZDM, 37(1), 1–3.
Carraher, D. W., & Earnest, D. (2003). Guess my rule revisited. In N. Pateman, B. Dougherty,
& J. Zilliox (Eds.), Proceedings of the 27th Conference of the International Group for the
Psychology of Mathematics Education (Vol. 2, pp. 173–180). Honolulu: University of Hawaii.
Carraher, D., Schlieman, A., Bruzuella, B., & Earnest, D. (2006). Arithmetic and algebra in early
mathematics education. Journal for Research in Mathematics education, 37(2), 87–115.
Carraher, D., Martinez, M., & Schlieman, A. (2008). Early algebra and mathematical generaliza-
tion. ZDM Mathematics Education, 40, 3–22.
Case, R., & Okamoto, Y. (1966). Monographs of the Society for Research in Child Development:
Vol. 61. The Role of Central Conceptual Structures in the Development of Children’s Thought.
(1-2, Serial No. 246).
Cooper, T., & Warren, E. (2008). The effect of different representations on Years 3 to 5 students’
ability to generalise. ZDM Mathematics Education, 40, 23–37.
Dorfler, W. (2008). On route from patterns to algebra: comments and reflections. ZDM Mathemat-
ics Education, 40, 143–160.
English, L. D., & Warren, E. (1998). Introducing the variable through pattern exploration. The
Mathematics Teacher, 91(2), 166–171.
Ferrini-Mundy, J., Lappan, G., & Phillips, E. (1997). Experiences with patterning. In Teaching
Children Mathematics, Algebraic Thinking Focus Issue (pp. 282–288). Reston, VA: National
Ginsburg, & Seo (1999). Mathematics in children’s thinking. Journal Mathematical Thinking and
Learning, 1(2), 113–129.
Greenes, C., Cavanagh, M., Dacey, L., Findell, C., & Small, M. (2001). Navigating through Alge-
bra in Prekindergarten—Grade 2. Reston, VA: National Council of Teachers of Mathematics.
Griffin, S., & Case, R. (1997). Re-thinking the primary school math curriculum an approach based
on cognitive science. Issues in Education, 3(1), 1–49.
Healy, L., & Hoyles, C. (2000). A study of proof conceptions in algebra. Journal for Research in
Mathematics Education, 31(4), 396–428.
Hewitt, D. (1992). Train spotters’ paradise. Mathematics Teaching, 140, 6–8.
Kalchman, M., Moss, J., & Case, R. (2001). Psychological models for the development of mathe-
matical understanding: Rational numbers and functions. In S. Carver & D. Klahr (Eds.), Cog-
nition and Instruction: 25 Years of Progress.
Kieran, C. (1992). The learning and teaching of school algebra. In D. A. Grouws (Ed.), The Hand-
book of Research on Mathematics Teaching and Learning (pp. 390–419). New York: Macmil-
lan.
Kuchemann, D. (2008). Looking for structure: A report of the proof materials project. BookPrintin-
gUK.com Peterborough PE29JX UK.
Lannin, J. K. (2003). Developing algebraic reasoning through generalization. Mathematics Teach-
ing in the Middle School, 8(7), 343–348.
Lannin, J. K. (2005). Generalization and justification: The challenge of introducing algebraic rea-
soning through pattern activities. Mathematical Thinking and Learning, 7(3), 231–258.
Lannin, J., Barker, D., & Townsend, B. (2006). Recursive and explicit rules: How can we build
student algebraic understanding. Journal of Mathematical Behavior, 25, 299–317.
Lee, L. (1996). An initiation into algebraic culture through generalization activities. In N. Bednarz,
C. Kieran, & L. Lee (Eds.), Approaches to Algebra: Perspectives for Research and Teaching
(pp. 65–86). Dordrecht: Kluwer Academic Publishers.
Lee, L., & Wheeler, D. (1987). Algebraic Thinking in High School Students: Their Conceptions of
Generalisation and Justification (Research Report). Concordia University, Montreal, Canada.
London McNab, S., & Moss, J. (2004). Mathematical modeling through knowledge building: Ex-
plorations of algebra, numeric functions and geometric patterns in grade 2 classrooms. In D.
McDougall & J.A. Ross (Eds.), Proceedings of the Twenty-Sixth Annual Meeting North Ameri-
can Chapter of the International Group for the Psychology of Mathematics Education, Toronto,
Canada, Vol. 2.
(Eds.), Approaches to Algebra: Perspectives for Research and Teaching (pp. 65–86). Dordrecht,
Netherlands: Kluwer.
Moss, J. (2004). Pipes, tubes, and beakers: Teaching rational number. In J. Bransford & S. Donovan
(Eds.), How People Learn: A Targeted Report for Teachers. Washington: National Academy
Press.
Moss, J. (2005). Integrating numeric and geometric patterns: A developmental approach to young
students’ learning of patterns and functions. Paper presented at the Canadian mathematics ed-
ucation study group 29 Annual Meeting at the University of Ottawa, May 27–31.
Moss, J., & Beatty, R. (2006a). Knowledge building in mathematics: Supporting collaborative
learning in pattern problems. International Journal of Computer Supported Collaborative
Learning, 1(4), 441–465.
Moss, J., & Beatty, R. (2006b). Multiple vs numeric approaches to developing functional under-
standing through patterns—affordances and limitations for grade 4 students. To be included in
the Proceedings of the Twenty-Eight Annual Meeting of the Psychology of Mathematics Edu-
cation, North American Chapter. Merida, Mexico.
Moss, J., & Case, R. (1999). Developing children’s understanding of rational numbers: A new
model and experimental curriculum. Journal for Research in Mathematics Education, 30(2),
122–147 119.
Moss, J., Beatty, R., Barkin, S., & Shillolo, G. (2008). What is your theory? What is your rule?
Fourth graders build their understanding of patterns. In C. Greenes (Ed.), NCTM Yearbook for
Algebraic Reasoning. Reston, VA: National Council of Teachers of Mathematics.
Mulligan, J., & Mitchelmore, M. (2009). Awareness of pattern and structure in early mathematical
development. Mathematics Education Research Journal, 21(2), 33–49.
Mulligan, J. T., Prescott, A., & Mitchelmore, M. C. (2004). Children’s development of structure
in early mathematics. In M. Høines & A. Hoines (Eds.), Proceedings of the 28th Annual Con-
ference of the International Group for the Psychology of Mathematics Education (Vol. 3, pp.
393–401). Bergen, Norway: Bergen University College.
matics. Reston, VA: National Council of Teachers of Mathematics.
Noss, R., & Hoyles, C. (1996). Windows on Mathematical Meanings. Dordrecht, The Netherlands:
Kluwer.
Noss, R., Healy, L., & Hoyles, C. (1997). The construction of mathematical meanings: Connecting
the visual with the symbolic. Educational Studies in Mathematics, 33(2), 203–233.
Ontario Ministry of Training and Education (2005). The Ontario Curriculum, Grades 1–8: Math-
ematics, Revised. Queen’s Printer for Ontario.
Orton, J. (1997). Matchsticks, pattern and generalisation. Education, 3–13.
Orton, A., & Orton, J. (1999). Pattern and the approach to algebra. In A. Orton (Ed.), Pattern in
Orton, J., Orton, A., & Roper, T. (1999). Pictorial and practical contexts and the perception of
pattern. In A. Orton (Ed.), Patterns in the Teaching and Learning of Mathematics (pp. 121–
136). London: Cassell.
Radford, L. (2008). Iconicity and contraction: a semiotic investigation of forms of algebraic gen-
eralizations of patterns in different contexts. ZDM Mathematics Education, 40, 83–96.
Rivera, F. (2006). Sixth Graders’ ability to generalize patterns in algebra: issues, insights. In J.
Novotná, H. Moraová, M. Kratká, & N. Stehlková (Eds.), Proceedings of the 30th Conference of
the International Group for the Psychology of Mathematics Education (Vol. 1, p. 320). Prague,
Czech Republic.
Rivera, F., & Becker, J. R. (2007). Abduction-induction (generalization) processes of elementary
majors on figural patterns in algebra. Journal of Mathematical Behavior, 140–155.
Rubenstein, R. (2002). Building explicit and recursive forms of patterns with the function game.
Mathematics Teaching in the Middle School, 7(8), 426–431.
Sasman, M., Olivier, A., & Linchevski, L. (1999). Factors influencing students’ generalization
thinking processes. In O. Zaslavski (Ed.), Proceedings of the 23th International Conference for
Psychology of Mathematics Education (Vol. 4, pp. 161–168). Haifa, Israel.
Schliemann, A. D., Carraher, D. W., & Brizuela, B. M. (2001). When tables become function
tables. In Proceedings of the XXV Conference of the International Group for the Psychology of
Mathematics Education (Vol. 4, pp. 145–152). Utrecht, The Netherlands.
tive. Journal of Mathematical Behavior, 12, 353–383.
Stacey, K. (1989). Finding and using patterns in linear generalizing problems. Educational Studies
in Mathematics, 20, 147–164.
Stacey, K., & MacGregor, M. (1999). Taking the algebraic thinking out of algebra. Mathematics
Education Research Journal, 11(1), 25–38.
Steele, D., & Johanning, D. (2004). A schematic-theoretic view of problem solving and develop-
ment of algebraic thinking. Educational Studies in Mathematics, 57(1), 65–90.
Warren, E. (2000). Visualisation and the development of early understanding in algebra. In T.
Nakahara & M. Koyama (Eds.), Proceedings of the 24th Conference of the International Group
for the Psychology of Mathematics Education (Vol. 4, pp. 273–280). Hiroshima.
Warren, E. (2006). Learning comparative mathematical language in the elementary school: A lon-
gitudinal study. Educational Studies in Mathematics, 62(2), 169–189.
Watson, A. (2000). Going across the grain: Generalisation in a group of low attainers. Nordisk
Matematikk Didaktikk (Nordic Studies in Mathematics Education), 8(1), 7–22.
Willoughby, S. (1997). Functions from kindergarten through sixth grade. Teaching Children Math-
ematics, 3, 314–318.
Grade 2 Students’ Non-Symbolic Algebraic
Thinking
Luis Radford
Abstract The learning of arithmetic, it has recently been argued, need not be
a prerequisite for the learning of algebra. From this viewpoint, it is claimed that
young students can be introduced to some elementary algebraic concepts in primary
school. However, despite the increasing amount of experimental evidence, the idea
of introducing algebra in the early years remains clouded by the lack of clear dis-
tinctions between what is arithmetic and what is algebraic. The goal of this chapter
is twofold. First, at an epistemological level, it seeks to contribute to a better un-
derstanding of the relationship between arithmetic and algebraic thinking. Second,
at a developmental level, it explores 7–8-years old students’ first encounter with
some elementary algebraic concepts and inquires about the limits and possibilities
of introducing algebra in primary school.
Introduction
This chapter is about the journey of a group of Grade 2 students (7–8-years old)
into algebra. It explores the students’ first encounter with some elementary alge-
braic concepts and inquires about the limits and possibilities of introducing algebra
in primary school. The chapter is inspired by a recent idea according to which the
learning of arithmetic need not be a prerequisite for the learning of algebra (e.g.,
Blanton and Kaput 2000; Carraher et al. 2006; Dougherty 2003). It is also inspired
by the particular context in which I work—a context in which algebra is introduced
in the early years. A progressive introduction to algebra in the early grades, it has
This chapter is a result of a research program funded by The Social Sciences and Humanities
Research Council of Canada/Le Conseil de recherches en sciences humaines du Canada
(SSHRC/CRSH).
L. Radford ()
École des science de l’education, Laurentian University, Sudbury, Ontario, Canada

304 L. Radford
been suggested, may facilitate the students’ access to more advanced algebraic con-
cepts in later grades.
Despite the increasing amount of experimental evidence that supports it (e.g.,
Brizuela and Schliemann 2004; Moss and Beatty 2006; Becker and Rivera 2006a,
2006b; Warren 2006; Warren and Cooper 2008), the idea of introducing algebra in
the early years remains clouded by the lack of clear distinctions between what is
arithmetic and what is algebraic. To make such distinctions would require making
explicit our own assumptions about the nature of arithmetic and algebraic thinking
and their interrelationships. And, of course, this is not an easy matter, neither for
research nor for practice. Let me give a concrete example.
In 1997 there was an important curriculum re-orientation in Ontario. New goals
and expectations were established concerning the content and aims of the teaching
and learning of mathematics. Since then a specific domain, called “Patterning and
Algebra”, has listed the overall and specific expectations from Grade 1 to 8 (see On-
tario Ministry of Education 1997). For instance, a specific expectation in Grades 1,
2 and 3 states that students identify, describe, and extend patterns. Another expecta-
tion states that the students demonstrate an understanding of the concept of equality
by partitioning whole numbers in a variety of ways (e.g., 5 is equal to 1 + 4, 2 + 3,
etc.). By Grade 5 the students are expected to use letters to represent relations (e.g.,
C = 3 × n). As these few examples show, there is a “progression” in terms of con-
ceptual content. However, the progression raises several questions. For example, are
the activities of extending patterns and partitioning numbers part of algebra or arith-
metic? Is the use of letters specific to algebra or could it also be part of arithmetic?
Let us assume for a moment that extending patterns and partitioning numbers are
really part of algebra. What would it be that makes them algebraic? What would be
the algebraic concepts required in those kinds of tasks? Unfortunately the official
document does not supply elements to answer those questions.
Of course, this problem is not specific to the Ontario curriculum. It is a token, I
think, of the difficulties that curriculum designers and we, mathematics educators,
still have in making a clear distinction between arithmetic and algebraic thinking,
and in elucidating the relationships between them.
I should hasten to say that the journey this article is about does not answer those
difficult questions. It is rather a modest attempt to reflect on them both from a prac-
tical and a theoretical perspective. What I shall present and discuss in this article
comes from what I learned from the students of a Grade 2 class and their teacher.
It is a journey in which I also embarked. Before I go into more specific details, it
may be necessary to say something about the protagonists of the journey and its
dynamics.
The journey is part of a still ongoing longitudinal research program in which
a Grade 2 class of 25 students and their teacher were invited to participate. Part
of a progressive French School Board in Ontario, the school is open to projects
and partnerships with different sectors of its community. The superintendent, the
principal and the teacher were interested to participate in a study to explore the
problems and challenges surrounding the introduction of algebra in the early years.
The teacher, the research assistants and I met regularly, either at the school or at the
Grade 2 Students’ Non-Symbolic Algebraic Thinking 305
Fig. A The first four figures of a sequence given to the students in a Grade 2 class
university, to discuss the choice of mathematical tasks and forms of interaction, and
how these fitted with our philosophy of teaching and learning and the curriculum
expectations.
Throughout the study, the classroom activities were carried out by the teacher,
while we videotaped. After videotaping, the research assistants and I made tran-
scriptions. Interpretations of our data were done with the teacher. Also, after each
mathematics lesson of the study, the teacher and I used a voice recorder to record our
impressions. As data was generated and interpreted, new hypotheses were formu-
lated and incorporated into the research, thereby ensuring a dialectical loop between
theory generation and data interpretation.
Even if our journey also included an investigation into the realm of equations,
in this chapter I shall limit myself to commenting on what we learned about the
generalization of patterns. In the next two sections I discuss some conceptual as-
pects involved in extending a sequence and the kind of abstraction it entails. These
sections pave the way for a discussion of the boundaries of arithmetic and algebraic
thinking in generalizing tasks while the last sections are devoted to a discussion of
embodied forms of algebraic thinking.
Extending Sequences
In this section I make a brief analysis of the students’ procedures in extending a

sequence. The analysis of the students’ procedures aims at shedding some light
on the type of thinking that is required to accomplish this type of task. The idea
is to make available some concrete information to reflect on the question of the
relationship between extending sequences and algebraic thinking.
The students worked on four activities about pattern generalization. These activi-
ties lasted five days in total (about 60 minutes each day). The first activity was based
on the sequence shown in Fig. A. The students were divided into groups of two or
three and were asked first to draw Figures 5 and 6 and then, after answering other
questions, to come up with a procedure or formula to find the number of squares in
some “big” figures, like Figure 25.
To continue the sequence the students had to grasp a commonality noticed in the
four given figures and generalize it to other terms of the sequence. Since there are
many ways in which to see things in front of us, the commonality that the students
noticed was not always the same. The commonality appeared in fact progressively
in the course of the students’ spatial-temporal experience of grasping it.
306 L. Radford
Fig. B While drawing Figure 5, Erica goes back to Figure 4 to count the squares on the top row.
The finger helps her to see and count
For instance, in one of the groups, formed by Erica, Cindy and Carl, Erica had
a sense of the next figure: its squares should be spatially distributed into two roads,
but neither their numerosity nor the position of the dark square followed the logic
of the first figures. Counting aloud, she started drawing the squares on the first row
from left to right: “one, two, three”. When she was about to draw the forth square,
she came back to Figure 4 of the sequence and, with a left hand finger counted the
number of white squares. She then continued drawing the squares on the top row in
Figure 5, followed by those on the bottom row, putting the dark square at the end
(See Fig. B).
The progressive grasping of the regularity appears as the result of linking two dif-
ferent kinds of structures: a spatial and a numerical one. From the spatial structures
there emerges a sense of the squares’ position, whereas their numerosity emerges
from a numerical structure. The former deals with the question of “where?”; the
latter with the question of “how many?” As a fine-grained video analysis shows, the
spatial-numerical link was mediated by a complex embodied intersensorial process
involving visual, motor, and aural elements. The motor action of drawing was ac-
companied by seeing (seeing where the squares have to be drawn so as to ensure
that the squares are aligned, that they have an approximate same size, etc.) and also
by language. By counting the squares as they are drawn, language helped Erica and
the other students to keep track of the question of “how many?” As in Grade 2 the
students’ motor actions underpinning pen use are still being refined, drawing small
squares takes considerable time and effort. Perceptual efforts are focused on the task
of drawing. Language, then, is used to maintain awareness of the numerical status
of the square that is being drawn.
As the group moved to the next task, i.e. the drawing of Figure 6, Carl and
Cindy—who, like Erica, counted aloud, but managed to coordinate the spatial and
numerical structures so as to make it coherent with the intended logic of the se-
quence in Fig. A—intervened and anticipated the key aspects of the figure. Carl
said:
1. Carl: We do 6 plus 6 equals 12, plus 1
2. Erica: Yes. . . No. . .
3. Cindy: Yes!
4. Carl: Yes, that is what we did in the other ones. Look!
Fig. C To the left, Carl starts counting the squares on the bottom row; then he counts the white
squares on the top. To the right, the moment in which, after finishing counting the top white squares,
he goes back to the dark square and says “plus 1”
5. Cindy: Yes, you add one.

6. Carl: (Talking to Erica) Look. 4 plus 4 equals 8. In there [Figure 4], there are 8.
(And pointing to the successive squares of Figure 4 in Erica’s sheet he continued)
1, 2, 3, 4, 5, 6, 7, 8, plus 1, which is equal to 9. (See Fig. C)
Carl’s addition “6 plus 6” in Line 1 conveys the link between the spatial and
numerical structures. Indeed, the expression “6 + 6” is not a mere addition: it is a
synthetic or global expression that conveys the ideas of where the squares are, and
how many squares there are. In addition to the numeric information, the first 6 refers
to the bottom row; similarly, the second 6 refers to the top row, and indicates that to
the six squares an additional square has to be added—the one corresponding to the
dark square. As Erica’s perception of the figures seemed to be more sequential (she
relies on counting squares one after the other), she was unsure about what her team-
mates meant. Hence Carl illustrated the idea, resorting to one of the figures given to
the students—Figure 4. He started counting the squares on the bottom row (“1, 2, 3,
4.”). Then he switched to the top row and continued counting the white squares, and
ended the counting with the dark square. Erica and Cindy followed Carl’s gestures
attentively.
Through his intervention, Carl made available to his team-mates a specific com-
monality in the terms of the sequence. He talked about Figure 4, but implicit in his
intention was the idea that such a commonality applied to the other figures as well.
Abstraction
The grasping of the commonality is the formation of what, in Aristotelian terminol-
ogy, is called a genus, i.e. that in virtue of which various things are recognized as
belonging together (see e.g. Aristotle’s Categories, 2a13–2a18). The genus results
from an abstraction that requires the students to make distinctions between what is
similar and what is different in the given terms of the sequence. Now, can the stu-
dents use this abstracted commonality to assert whether a given term belongs to the
sequence? We wanted to explore this question and presented the students with the
following task:
308 L. Radford
Monique wants to build Figure 8 of the sequence. She builds the figure shown
below. Are you in agreement with Monique? Please explain!
Resorting to the objectified commonality or genus, the students argued that this
term was not Figure 8. Cindy pointed to the top row and counted the white squares.
She said: “It isn’t because 1, 2, 3, 4, 5, 6, 7”, meaning that there were not 8 white
squares. Carl agreed: “Yes, because, look, here it [pointing to the top row] has 7,
here it [the bottom row] has 8, so it’s not good.”
However, I want to argue that the identification of the genus and its use to extend
the sequence to neighboring terms (e.g. Figures 5 and 6) as well as to discriminate
between terms that belong and those that do not belong to the sequence cannot
be considered the result of an algebraic process. However complex such a process
might be, there is nothing necessarily algebraic in it.
Indeed, noticing what is really common and characteristic of the terms of a se-
quence or set of objects is a central aspect of concept formation. As we shall see
in the next section, it plays an important role in the emergence of the students’
first algebraic ideas, but is not itself the result of an algebraic process. In fact, find-
ing a characterizing attribute of the terms of a sequence or a set of objects is not
specific to humans. Indeed, Sue Savage-Rumbaugh and her team as well as other
researchers in the field of animal cognition have established that chimpanzees (and
birds too) can distinguish between several sorts of objects—classing them as “ed-
ible” and “inedible” (Savage-Rumbaugh et al. 1980). Even if chimpanzees “might
be lacking the ability to organize their categories into a more complex hierarchi-
cal network of superordinate categories of increasing abstraction” (Gómez 2004,
p. 126), they do form abstract commonalities (i.e. abstract genera) typical of con-
cept formation. But this cognitive capability alone is not a warrant to claim that
chimps and birds are thinking algebraically. In a similar vein, the capacity of our
Grade 2 students to grasp a commonality, even in complex sequences like the one
shown in Fig. A, and to extend it to a few subsequent terms does not mean that
the students are already thinking algebraically. Generality is not specific to algebra.
Generality is a typical general trait of human and animal cognition and can be of
diverse nature—arithmetic, geometric or other.
Hence, the question is: when, if at all, did our Grade 2 students start thinking
algebraically?
The Boundaries of Arithmetic and Algebraic Thinking
Algebraic thinking does not appear in ontogeny by chance, nor does it appear as the
necessary consequence of cognitive maturation. To make algebraic thinking appear,
and to make it accessible to the students, some pedagogical conditions need to be
created. This was what we were trying to accomplish through the design of our
activities and their implementation in the classroom.
The rest of our questions in the 2-day activity around the sequence shown in
Fig. A were directed to having the students experience the need to come up with
a general procedure. We asked them to consider Figures 12 and 25. We also asked
them the following question:
Pierre wants to build a big figure of the sequence. Explain to him what to do.
The questions were promptly answered. Thus, talking about Figure 12, with ease
Cindy said: “12 plus 12, plus 1”. Referring to Figure 25, Erica said:
1. Erica: Cindy! Um. . . Okay, What is 25 plus 25?
2. Cindy: (Thinking) Euh. . .
3. Erica: (Smiling) After that, you add one!
Before going further, it is worth noticing that to ask the students to find the num-
ber of squares in “big” figures like Figures 25 was far from trivial. The arithmetic
knowledge of our Grade 2 students was, at the time, very limited. Although they had
some acquaintance with “big” numbers, they were able to make systematic additions
only up to 25. Hence our question about Figure 12 was at the very limit of their cal-
culating capabilities. Our question about Figure 50 was definitely beyond them. But
instead of being a hindrance, not knowing how to make additions beyond 25 was in
fact good. The design of the activities was based on the limits of students’ arithmetic
knowledge to promote the emergence of algebraic thinking. Indeed, by exploiting
the students’ limits of arithmetic thinking, the design of the activity aimed at favour-
ing the students’ awareness of calculation methods. Here the calculator proved to be
of great importance.
To help the students deal with big figures, the teacher made calculators available
to the students. But, before finding the actual number of squares in Figure 25 or
other big figures using the calculator, she asked them to come up with an idea of
how to find the total. This pedagogical strategy induced in the students’ activity an
important shift from the numeric qua numeric to the devising of a rule or calculation
method. This shift is central to an algebraic mind. It was the 13th century Arabian
mathematician Aboû Beqr Alkarkhî who first expressed this idea. He suggested that
algebra consists of rules to calculate numbers (Woepcke 1853). And it was within
this context that the students tackled the Pierre question concerning a big, unspec-
ified figure mentioned above. Surely enough, the students chose “big” particular
figures. Carl suggested considering Figure 500; Cindy preferred Figure 50:
4. Carl: How about doing 500 plus 500?
5. Erica: No. Do something simpler.
6. Carl: 500 plus 500 equals 1000.
7. Erica: plus 1, 1001.
8. Carl: plus 1, equals 1001.
9. Cindy: Non, 50 plus 50, plus 1 equals 101.
It might be worth asking now whether or not there is something algebraic in
these responses. Let me note first that to answer the question about Figure 12 Erica
310 L. Radford
and her team did not go from Figure 4 to Figure 12, building figure after figure. To
deal with Figures 12 and 25 the students accomplished a generalization. And this
generalization was algebraic in nature.
I will start the justification of my claim by first addressing a potential objection. It
might be argued that even though the students accomplished a generalization, their
generalization was not algebraic, since the students did not use “notations”. This
is true. However, as I have argued in previous articles, the use of notations (i.e., al-
phanumeric symbolism) is neither a necessary nor a sufficient condition for thinking
algebraically (Radford 2006a, 2009a). Algebraic thinking is not about using or not
using notations but about reasoning in certain ways. What characterizes thinking
as algebraic is that it deals with indeterminate quantities conceived of in analytic
ways. In other words, you consider the indeterminate quantities (e.g. unknowns or
variables) as if they were known and carry out calculations with them as you do
with known numbers (see Filloy et al. 2007). It is in this sense that 16th century
mathematicians like Viète and Cardano understood the distinctive trait of algebra
and used to call it an analytic art (see e.g. Viète 1983). And it is in this sense that
Erica and her team were dealing with our questions about Figures 12 and 25.
My problem now is to show how indeterminacy and analyticity are present in
the students’ procedures. To understand the subtle sense in which indeterminacy
and analyticity were present in our Grade 2 students’ procedures, we have to bear
in mind that even if indeterminacy is usually expressed through letters (e.g. “x”,
“n”, etc.), there are also other genuine forms in which to express it. Before our
standard algebraic symbolism was invented, mathematicians used various ways to
think about and deal with indeterminacy. For instance, Babylonian scribes used con-
textual names, depending on the problem (e.g., the side of a rectangle, the weight
of a stone); Medieval and Renaissance mathematicians employed a generic term—
la cosa (the thing). In the case of our Grade 2 students, indeterminacy is present
through instances of the independent variable—i.e., the number of the figure (“1”,
“2”, 3”, “4”, “5”, “6”, “12”, “25”, “50”). In other words, indeterminacy appears em-
bodied in its surrogates. Indeterminacy and analyticity are in fact bound together
in a schema or rule that allows the students to deal with any particular figure of
the sequence, regardless of its size. It is a rule instantiated in particular cases (e.g.
“12 plus 12, plus 1”), where numbers are dealt with not as merely numbers but as
constituents of something more general. The suspension of intermediate and final
results in “12 plus 12, plus 1” is well tuned with the algebraic idea of analyticity.
What matters is not really the numeric result, but the rule. The students’ rule (“12
plus 12, plus 1”; “25 plus 25, plus 1”, etc.) attests to a shift in focus: the students’
focus is no longer specifically numeric. It is about numbers, of course, but in an al-
gebraic manner, however simple this manner is. This is why Erica was not stopped
by not knowing the result of 25 + 25. Whatever it is, you have to add 1 (the dark
square). And when Carl hurried to help Erica with her question about 25 + 25, he
offered the answer “14”, which surprised his team-mates. Carl then added: “14 or
whatever!” Indeed, whatever the answer is, the important thing is to add 1. For the
students’ emerging understanding, what matters is not the result. It is the rule, that
is to say, the formula—the algebraic formula.
This “formula” can better be understood as an embodied predicate (e.g. “12 plus
12, plus 1”) with a tacit variable: indeterminacy as such does not reach the level
of symbolization, not even the level of discourse. There are no words in the stu-
dents’ vocabulary to name it. Indeterminacy remains implicit—something whose
presence is only vaguely adverted through particular instances, like clouds antic-
ipating a storm. Indeterminacy is expressed in an indexical manner: its instances
point to something that is in adventus, that is to say to-come.
Layers of Generality
My claim that Erica and her team-mates have effectively stepped into the realm of
algebra makes sense only if, as I suggested, we consider algebra as a particular form
of thinking that, instead of being characterized by alphanumeric signs, is rather char-
acterized by the specific manner in which it attends to the objects of discourse. This
distinctive manner of being of algebraic thinking can be defined, on epistemologi-
cal grounds, by indeterminacy and analyticity. These two elements are what make
algebraic thinking different from arithmetic and other forms of thinking.
Now, indeterminacy and analyticity can take several forms. And this is so because
algebraic thinking can operate at different layers of generality. Some layers are more
concrete, some more general. In some concrete layers, indeterminacy and analyticity
may appear in an intuited form, as in the previous section. In others, they may appear
in a more explicit manner, as when students use alphanumeric symbolism.
Layers of generality can be distinguished in terms of the signs to which the stu-
dents resort to think algebraically. Indeed, thinking, as I am conceiving of it here,
is not something that is restricted to the mental plane. On the contrary, thinking
also occurs in the social plane. Gestures, language, and perception are material con-
stituents of thinking (Radford 2009b). The material (e.g. gestures) and immaterial
(e.g. imagery, inner speech) components of thinking constitute its “semiotic tex-
ture”. As there are some things that we can and others that we cannot say, think,
and intend through certain signs (think for instance of the impossibility of exactly
translating a poem into a painting), the “semiotic texture” of thinking sets the limits
of what is sayable, thinkable and intendable within it. If I spent some time analyz-
ing the manner in which Erica and her group-mates became aware of the genus of
the sequence and used it to objectify an algebraic rule, it was because such analysis
reveals the “semiotic texture” of their algebraic thinking. This rule, I argued, was
not made up of alphanumeric signs but of embodied signs. Of course, epistemo-
logically speaking, the layer of generality in which these students were operating
was not very profound. Yet Erica and her team-mates were thinking algebraically.
I deem this point important for our understanding of what is achievable in terms of
introducing algebra in the early grades. The natural question is: Is this all that Grade
2 students are capable of? I deal with this question in the next section.
312 L. Radford
Fig. D The teacher with a

box containing numbered
cards representing the number
of some terms of the first
day’s sequence. With her
right hand she holds the
envelope where one of the
cards with an unknown
number has been put
Beyond Intuited Indeterminacy

At the end of the second day, the teacher and I were convinced that asking the stu-
dents to write an explanation of their solutions and ideas was not productive. The
students were still in the early stages of writing and to write an explanation of two
sentences was taking them far too long. Furthermore, in writing their explanations,
the discussions turned often into other matters that were of relative importance in
terms of algebraic thinking. For instance, the students were spending huge amounts
of time deciding how to spell a word or discussing whether or not a plural name
should have an s at the end. We decided then to turn to oral explanations. Starting
from the third day all groups were provided with a digital voice recorder. When
a student was ready to offer an explanation, the student would activate the voice
recorder, say her name and start the recording. Our Grade 2 students ended up prac-
ticing oral algebra— perhaps like early Renaissance students before the invention
of the printing press and the spread of writing as a social phenomenon (Radford
2006b). It was in this context of oral-oriented activity that the students spent the rest
of the week working on similar sequences as the one shown in Fig. A.
On the fifth and final day of our pattern generalization teaching-learning se-
quence, the teacher came back to the sequence shown in Fig. A. To recapitulate, she
invited some groups to share in front of the class what they had learned about that
sequence in light of previous days’ classroom discussions and small group work.
Then, she asked a completely new question to the class. She took a box and, in
front of the students, put in it several cards, each one having a number: 5, 15, 100,
104, etc. Each one of these numbers represented the number of a figure of the pat-
tern shown in Fig. A. The teacher invited a student to choose randomly one of the
cards and put it into an envelope, making sure that neither the student herself nor
the teacher nor anybody else saw the number beforehand. The envelope, the teacher
said, was going to be sent to Tristan, a student from another school. The Grade 2
students were invited to send a recorded message that would be put in the envelope
along with the card. In the message the students would tell Tristan how to quickly
calculate the number of squares in the figure indicated on the card (see Fig. D).
The number of the figure was hence unknown. Would the students be able to
generalize the rule that they had objectified when working with “big” figures and
engage with calculations on this unknown number? In other terms, would our Grade
2 students be able to go beyond intuited indeterminacy and its corresponding ele-

mentary form of algebraic thinking?
As in the previous days, the students worked in small groups of three. Let me
dwell on what happened in Erica’s group. In an episode that lasted 30 seconds,
Erica started making a suggestion:
10. Erica: We can say . . .

11. Cindy: You do. . .
12. Carl: You can do. . .
13. Erica: You can do the number. . . (she makes the pointing gesture shown in
Fig. E, Pic. 1).
14. Cindy: You look at the number and then you. . .
15. Carl: He will have
16. Erica: (Continuing her utterance in line 13) The same number. . .
17. Cindy: And then you. . .
18. Erica: (Continuing her utterance in line 16) as at the bottom (she makes the
pointing gesture shown in Fig. E, Pic. 2), after on the side you put another one
(she makes the pointing gesture shown in Fig. E, Pic. 3).
19. Cindy: and then, and then. . .
20. Carl: And then at the bottom he will have the same number of light squares (he
makes the pointing gesture shown in Fig. E, Pic. 4), at the top the same number
of light squares (he makes the pointing gesture shown in Fig. E, Pic. 5), and a
dark one (he makes the pointing gesture shown in Fig. E, Pic. 6).
As the previous dialogue shows, the fact that the number of the figure was un-
specified did not impede the students in thinking of and talking about the figure in
a mathematical way. Through the linguistic expression “the number”, the students
engaged with indeterminacy in an explicit manner. The definite article “the” quali-
fies the noun “number” making it specific even if it is unknown. From the intuited
form in which it appeared in the students’ previous activity, indeterminacy has now
entered the realm of the students’ universe of discourse. In so doing, the students
have reached a new layer of generality. However, this new layer remains deeply
anchored in the students’ perceptual experience, as shown by the students’ fierce
recourse to gestures and contextual clues through which they somehow make visi-
ble the unspecified figure. Gestures and words allow the students to visualize in an
embodied and almost tangible way the figure. Indeed, while Erica did not gesture
when tackling the question about Figure 25, here she made extensive use of ges-
tures and linguistic deictics (“top”, “bottom”). In line 13, she says: “You can do the
number. . .” and points to an imaginary place where would be the bottom row of the
unspecified figure. The utterance continues in lines 16 and 18, where she says: “The
same number. . . as at the bottom”, pointing now to the imaginary place of the top
row. Then pointing to a spot on the right side, she finishes the sentence saying “after
on the side you put another one.” Drawing on Erica’s idea, Carl immediately offered
a recapitulation that, interestingly enough, was accompanied by linguistic deictics
314 L. Radford
Fig. E Erica’ and Carl’s gestures help them to visualize the unknown figure
and a set of three gestures on the table with a pronounced movement of the arms
and the whole upper part of the body.1
However, in contrast to what the students did when dealing with “big” particular
figures, like Figure 50, here the students did not produce a formula. Indeed, instead
of something similar to Cindy’s formula “50 plus 50, plus 1” (line 9), the students
produced a spatial description of the unspecified figure. As Carl said: “at the bot-
tom he will have the same number of light squares; at the top the same number
of light squares and a dark one.” As a result, there are no explicit operations with
the unknown number. In other words, analyticity—this chief feature of algebraic
thinking— seems to be missing.
When the teacher came to see the group’s work, Carl explained the message
they were working on, using an example—Figure 50. Here is an excerpt of the
discussion:
1 The role of gestures and words in visualization and thinking is not specific to our Grade 2 students.
It has been put into evidence with older students (Radford et al. 2007; Sabena et al. 2005) and also
in other contexts. See, e.g. the pioneer work of Presmeg (1986); see also Arzarello and Robutti
(2004), Edwards (2009), Roth (2001), Nemirovsky and Ferrara (2009).
Fig. F Erica refers to the calculator to make the calculations on the unknown number
21. Carl: You do 50, plus 50, plus 1.

22. Teacher: Excellent! That would be a good example. But what if Tristan finds
another number? . . .
23. Carl: 100 plus 100 plus 1 . . .
Here Carl is dealing with generality through particular examples, in a manner that
Balacheff (1987) calls “generic example”, a way of seeing the general through the
particular, as Mason (1996) puts it. Erica continued:
24. Erica: It’s the number he has, the same number at the bottom, the same number
at the top, plus 1 . . .
25. Teacher: That is excellent, but don’t forget: he doesn’t have to draw [the figure].
He just has to add. . . So, how can we say it, using this good idea?
26. Erica: We can use our calculator to calculate!
27. Teacher: Ok. And what is he going to do with the calculator?
28. Erica: He will put the number. . . (she pretends to be inserting a number into the
calculator; see Fig. F, Pic. 1).
29. Cindy: He will do: the number. . .
30. Erica: plus the same number, plus 1 (as she speaks, she pretends to be inserting
the number again (Pic. 2) and the number 1 (Pic. 3)).
31. Carl: Yeah!
32. Teacher: (Repeating) The number, plus the same number, plus 1! Do you think
that Tristan would be able to find the total like that?
33. Cindy and Carl: Yes!
34. Teacher: Very good. I will go to check on the other groups now.
In Line 25 the teacher makes the subtle distinction between drawing and cal-
culating. The formula can be derived from the students’ general description of the
figure, but is not equal to it. An algebraic formula does not include terms such as
“top” and “bottom”. In Line 26, Erica suggested using the calculator and, along
with Cindy, mentioned the sequence of calculations to be carried out in order to
find the total. Naturally, the use of the calculator is merely virtual. In the students’
calculator, all inputs are specific numbers. Nevertheless, the calculator helped the
students objectify the analytic dimension that was apparently missing in the new
layer of generality. Through the calculator, calculations are now performed on this
unspecified instance of the variable—the unknown number of the figure.
316 L. Radford
At the end of the lesson, several groups were invited to come forward and record
the message to Tristan. When Erica’s group was invited, Carl recorded a message
using a particular figure (Figure 50). While congratulating Carl, the teacher stressed
the fact that the message was based on a concrete example and asked if there was
another way of telling Tristan what to do. Using the voice recorder, Erica recorded
the following message:
Hi Tristan. You put the number at the bottom (she makes a gesture like pointing at the
imaginary bottom row) the same number on top (like pointing at the imaginary top row),
plus 1. Afterwards, you use the calculator and (making gestures as if using the calculator
keyboard) you insert the number plus the same number plus 1, and after you press equal
and it will show you what it is.
The message was divided into two parts. In the first part Erica tells Tristan about
the aspect of the figure. In the second part, Erica indicates the calculations to be
performed. It seems that knowing how the figure looks is a prerequisite to making
the calculations. Indeed, Erica imagines entering the numbers in the calculator in
the order that the students imagined the unspecified figure (from bottom to top, then
the dark square). The meaning of the terms in the formula is hence derived from the
spatial configuration of the figure.
An important aspect of the development of algebraic thinking consists in imbu-
ing the terms of a formula with abstract meanings so that formal calculations can be
performed with the unknown numbers. The situated, spatial sense of the unknown
numbers in the formula seems to constitute the limits of our Grade 2 students’ al-
gebraic thinking. Yet there was one group that overcame this limit. When their turn
came to record the message for Tristan, they produced the following message:
Hello Tristan, um. . . we are going to show you a strategy to figure out the sequence. Um,
you have to find the number. If the number you grab is like 50 or 40 or something, you have
to do like the number times two and after plus 1, and you will see what it equals to.
Here the addition of the unknown number with itself is turned into a multiplica-
tion by two. The spatial meaning of the unknown is overcome.
A General Overview
The next week, the students were asked to respond to a questionnaire. Even if ques-
tionnaires are hardly the best instruments to assess cognitive development, the stu-
dents’ responses give us an additional perspective of their general progress towards
algebraic thinking. In this section, bearing in mind the partial and limited view of
questionnaires, I will comment on some of the results. The students dealt with the
sequence shown in Fig. G.
As in previous tasks, they were asked to continue the sequence up to Figure 6 and
indicate the number of circles in Figures 5 and 6. They were also asked to indicate
the calculations to find out the number of circles in Figure 25.
The identification of the commonality and its use in extending the sequence were
very well accomplished. The success rate was 21/25 or 84%. Let me note that
Fig. G The sequence of the test
among the 4 other answers, one comes from a student who did not respond. Two
others come from students who, reading Figure 1 as having “one on the bottom and
two more on top”, interpreted the commonality as follows: Figure 6 has six circles
on the bottom and two more on top; they drew Figure 6 accordingly and argued that
Figure 6 has 6 + 8 = 14 figures. The fourth answer came from a student who drew
the circles with no order (figures were correct from the numeric structure but not
from the spatial one).
What were the results in the question about Figure 25? Thirteen students showed
a formula to obtain the total of circles in Figure 25 (“25 + 25 + 1” or “25 + 26”).
However, a closer look at the data reveals that five additional students knew that
they had to add two consecutive numbers, and used “25 + 24”. I want to suggest
that the difficulty was partly algebraic and partly arithmetic. The algebraic gener-
alization was accomplished, but it was its quantification that became problematic.
If this explanation is accepted, then the ratio of success increases to about 18/25
or 72%. This ratio is equivalent to if not better than those we have obtained in the
introduction of algebra in Grade 7 and Grade 8. All things considered, our Grade 2
students did very well.
Synthesis and Concluding Remarks

In this chapter I presented an overview of a journey of Grade 2 students into algebra.
The journey was motivated by new curricular trends that recommend starting to in-
troduce elementary algebraic concepts at the beginning of primary school. However,
as Carraher et al. note “little is known about children’s ability to make mathematical
generalizations and to use algebraic notation” (Carraher et al. 2006, p. 111). Despite
the extant experimental evidence and recent theoretical developments in early alge-
bra research (Carraher and Schliemann 2007), there is still a lot to do.
The journey this article was about focused on the generalization of patterns. This
topic has become a very popular way to introduce students into algebra. It now
occupies an important place in many contemporary curriculum programs around the
world (MacGregor and Stacey 1992). Yet, it is haunted by many misunderstandings
due to the lack of clear distinctions between arithmetic and algebraic thinking, and
an unambiguous elucidation of the relationships between them. In all fairness we
should note that this problem is not specific to pattern generalization research. It also
permeates the broader mathematics education field that deals with the relationships
between arithmetic and algebra.
In the first part, I offered an interpretative analysis of extending sequence activ-
ities. I scrutinized the type of thinking that is elicited by those activities in order
318 L. Radford
to reflect on the question of whether or not there is something algebraic in it. The
analysis put into evidence the fact that, to extend a sequence, our Grade 2 students
resorted to coordinating spatial and numerical structures that, although highly com-
plex, do not mobilize algebraic concepts as such. Extending sequences require in-
deed the grasping of a commonality, a process that instead of being algebraic is part
in fact of a more generic process of concept formation, accessible also (although
within certain limits) to other species.
Asking questions, however, about “remote” figures, like Figures 25 and 50—
that is, figures beyond the perceptual field—is a different matter. To answer ques-
tions like those, our Grade 2 students did make a generalization. And I tried to
show that the generalizations the students produced were certainly algebraic. To do
so, I had to elaborate on the idea of algebraic thinking and what makes it distinc-
tive. I started by arguing that it is misleading to associate algebraic thinking with
the use of letters. Unfortunately, very often, curricular documents and current re-
search about the relationships between arithmetic and algebraic thinking get caught
in this trap. Algebraic thinking, I suggested, is not about using letters but about rea-
soning in certain ways. On epistemological grounds (Serfati 1999, 2006; Radford
1997, 2001) and drawing on the seminal work of Kieran (1989), Filloy and Rojano
(1989), Filloy et al. (2007), Bednarz et al. (1996), Vergnaud (see, e.g., Cortes et
al. 1990) and others I suggested that what distinguishes arithmetic from algebraic
thinking is the fact that in the latter indeterminate quantities are treated in an ana-
lytic manner. I then moved on to claim, on semiotic grounds, that there are different
ways in which to think of and express indeterminacy. Indeterminacy can indeed be
expressed through signs other than the alphanumeric ones of conventional modern
algebraic symbolism. This claim is fully compatible with the historical development
of algebra. But, even more importantly, it makes room for the investigation of non-
symbolic forms of algebraic thinking—an endeavour that is of great importance if
we are to honour and understand the potential diversity of young students’ algebraic
thinking.
Bearing these ideas in mind, the analysis that we conducted of the students’ math-
ematical activity showed that to tackle the questions about “remote” and “unspec-
ified” figures, the students dealt first with indeterminacy and analyticity in what
turned out to be an elementary algebraic layer of generality. This layer of general-
ity sets the limits and possibilities of the corresponding form of algebraic thinking.
Thus, the most elementary form of algebraic thinking elicited by our mathematical
tasks was one where indeterminacy remained confined to specific figures. Indeter-
minacy appears here in an intuited form: it is expressed through particular instances
of the variable in the form of a concrete rule or formula (like “50 plus 50, plus
1”). Analyticity and indeterminacy remain attached to the level of particular figures
and arithmetic facts. We have encountered this form of thinking in older students
(Grade 7, 8, 9) and called it factual algebraic thinking (Radford 2000, 2003).2 It is
2 This intuited form of variable has also been investigated by Fujii and Stephens (2008), and re-
ferred to as quasi-variables.
interesting to see that this embodied form of algebraic thinking can be accessible to
most of Grade 2 students.
But, as our results imply, Grade 2 students can also engage in more sophisticated
forms of algebraic thinking. The Tristan problem suggests indeed that students can
deal with indeterminacy and analyticity in a more explicit way. We have also found
this form of thinking in older students. We have termed it contextual algebraic think-
ing to stress the fact that the meaning with which algebraic formulas are endowed is
deeply related to the spatial or other contextual clues of the terms the generalization
is about.3
In the case of our Grade 2 students, the calculator proved to be extremely useful
in the emergence of factual and contextual algebraic thinking. Its usefulness, how-
ever, was not limited to producing the numerical answers that were beyond the lim-
ited arithmetic knowledge of our students. Its usefulness resided in the conceptual
frame that it made available for the students to envision the calculations to be per-
formed and to come up with algebraic formulas. It might not be a coincidence that,
historically speaking, the astounding consolidation of algebra in the late Renais-
sance was followed by attempts to construct the first calculating machines (Radford
2006b).
At any rate, the most important point, I believe, is that algebraic thinking is by
no means something “natural”, something that will appear and develop once the
students have matured enough. Algebraic thinking is a very sophisticated cultural
type of reflection and action, a way of thinking that was refined again and again
through centuries before it reached its actual form. This is why its acquisition in on-
togeny raises very difficult problems (Radford 2008b). In a sense, the journey this
article was about is the story of our attempt at creating the pedagogical conditions
for the students’ first encounter with algebra. And as the journey intimates, the kind
of algebraic thinking that emerged from the classroom activities was framed by the
students’ evolving understandings, the questions that we were asking, their interac-
tion with peers, the teacher’s participation, and the historical intelligence embedded
in language and in the tools that were made available to the students.
It is still too early for us to offer an appraisal of the longitudinal development
of the students’ algebraic thinking. We hope, however, to be able to do it relatively
soon. We have already accompanied these students as they moved to Grade 3 and,
this year, we will continue following them in Grade 4. Such a longitudinal accom-
paniment should allows us to document the consolidation of the students’ algebraic
thinking and its evolution into more sophisticated forms, including the dawn of sym-
bolic algebraic thinking.
3 To tackle generalizing tasks like the ones seen in this chapter, older students tend also to mobilize
forms of thinking that are not algebraic. For instance, some of them tend to produce generaliza-
tions that are strictly arithmetic—e.g., guessing rules and inductive generalizations (for a detailed
discussion about the difference between algebraic and non-algebraic generalizations, see Radford
2008a). Knowing little arithmetic seems to inhibit the appearances of these arithmetic generaliza-
tions that compete with the algebraic ones. Instead of being an obstacle, knowing little arithmetic
facilitates, at least to some extent, the focus on the algebraic.
320 L. Radford
Acknowledgement I would like to thank the reviewers of a previous version of this chapter for
their precious insights and comments.
References
Arzarello, F., & Robutti, O. (2004). Approaching functions through motion experiments. In
R. Nemirovsky, M. Borba & C. DiMattia (Eds.), Educational Studies in Mathematics: Vol.
57(3). PME Special Issue of “Approaching functions through motion experiments”. CD-Rom,
Chap. 1.
Balacheff, N. (1987). Processus de preuve et situations de validation. Educational Studies in Math-
ematics, 18, 147–176.
Becker, J., & Rivera, F. (2006a). Establishing and justifying algebraic generalization at the sixth
grade level. In J. Novotná, H. Moraová, M. Krátkná, & N. Stehlíková (Eds.), Proceedings of
the 30th Conference of the International Group for the Psychology of Mathematics Education
(Vol. 4, pp. 465–472). Prague, Czech Republic.
Becker, J., & Rivera, F. (2006b). Sixth graders’ figural and numerical strategies for generalizing
patterns in algebra. In S. Alatorre, J. L. Cortina, M. Sáiz, & A. Méndez (Eds.), Proceedings
of the 28th Annual Meeting of the North American Chapter of the International Group for
the Psychology of Mathematics Education (Vol. 2, pp. 95–101). Mérida, México: Universidad
Pedagógica Nacional.
and Teaching. Kluwer: Dordrecht.
Blanton, M., & Kaput, J. (2000). Generalizing and progressively formalizing in a third grade math-
ematics classroom: Conversations about even and odd numbers. In M. Fernandez (Ed.), Pro-
ceedings of the Twenty-second Annual Meeting of the North American Chapter of the Interna-
tional Group for the Psychology of Mathematics Education PME-NA (pp. 115–119).
Brizuela, B., & Schliemann, A. (2004). Ten-year-old students solving linear equations. For the
Learning of Mathematics, 24(2), 33–40.
Carraher, D. W., & Schliemann, A. (2007). Early algebra and algebraic reasoning. In F. K. Lester
(Ed.), Second Handbook of Research on Mathematics Teaching and Learning (pp. 669–705).
Greenwich, CT: Information Age Publishing.
Carraher, D. W., Schliemann, A., Brizuela, B., & Earnest, D. (2006). Arithmetic and algebra in
early mathematics education. Journal for Research in Mathematics Education, 37(2), 87–115.
Dougherty, B. (2003). Voyaging from theory to practice in learning: Measure up. In N. A. Pateman,
B. J. Dougherty, & J. T. Zilliox (Eds.), Proceedings of the 27th Conference of the International
Group for the Psychology of Mathematics Education (PME & PMENA) (Vol. 1, pp. 17–23).
Honolulu: University of Hawai’i.
Edwards, L. (2009). Gestures and conceptual integration in mathematical talk. Educational Studies
in Mathematics, 70(2), 127–141. doi:110.1007/s10649-10008-19124-10646.
Filloy, E., Rojano, T., & Puig, L. (2007). Educational Algebra: A Theoretical and Empirical Ap-
proach. New York: Springer.
Fujii, T., & Stephens, M. (2008). Using number sentences to introduce the idea of variable. In
C. Greenes & R. Rubenstein (Eds.), Algebra and Algebraic Thinking in School Mathematics:
Seventieth Yearbook (pp. 127–140). Reston, VA: National Council of Teachers of Mathematics.
Gómez, J. C. (2004). Apes, Monkeys, Children, and the Growth of Mind. Cambridge: Harvard
University Press.
Kieran, C. (Ed.) (1989). The Early Learning of Algebra: A Structural Perspective. Virginia:
Lawrence Erlbaum Associates and National Council of Teachers of Mathematics.
MacGregor, M., & Stacey, K. (1992). A comparison of pattern-based and equation-solving ap-
proaches to algebra. In B. Southwell, K. Owens, & B. Penny (Eds.), Proceedings of the 15th
Annual Conference, Mathematics Education Research Group of Australasia (MERGA) (pp.

362–370). University of Western Sydney, July 4–8.
(Eds.), Approaches to Algebra (pp. 65–86). Dordrecht: Kluwer.
Moss, J., & Beatty, R. (2006). Knowledge building and knowledge forum: Grade 4 students col-
laborate to solve linear generalizing problems. In J. Novotná, H. Moraová, M. Krátká, & N.
Stehlíková (Eds.), Proceedings of the 30th Conference of the International Group for the Psy-
chology of Mathematics Education (pp. 193–199). Prague, Czech Republic.
Nemirovsky, R., & Ferrara, F. (2009). Mathematical imagination and embodied cognition. Educa-
tional Studies in Mathematics, 70(2), 159–174. doi:110.1007/s10649-10008-19150-10644.
Ontario Ministry of Education (1997). The Ontario Curriculum, Grades 1–8, Mathematics.
Toronto: Queen’s Printer for Ontario (Revised version, 2005).
Presmeg, N. (1986). Visualization in high school mathematics. For the Learning of Mathematics,
6(3), 42–46.
Radford, L. (1997). L’invention d’une idée mathématique: la deuxième inconnue en algèbre.
Repères (Revue des instituts de Recherche sur l’enseignement des Mathématiques), 28, 81–96.
Radford, L. (2000). Signs and meanings in students’ emergent algebraic thinking: A semiotic anal-
Radford, L. (2001). The historical origins of algebraic thinking. In R. Sutherland, T. Rojano, A.
Bell, & R. Lins (Eds.), Perspectives in School Algebra (pp. 13–63). Dordrecht: Kluwer.
Radford, L. (2003). Gestures, speech and the sprouting of signs. Mathematical Thinking and Learn-
ing, 5(1), 37–70.
Radford, L. (2006a). Algebraic thinking and the generalization of patterns: A semiotic perspective.
In S. Alatorre, J. L. Cortina, M. Sáiz, & A. Méndez (Eds.), Proceedings of the 28th Confer-
ence of the International Group for the Psychology of Mathematics Education, North American
Chapter (Vol. 1, pp. 2–21). Mérida, Mexico.
Radford, L. (2006b). The cultural-epistomological conditions of the emergence of algebraic sym-
bolism. In F. Furinghetti, S. Kaijser, & C. Tzanakis (Eds.), Proceedings of the 2004 Conference
of the International Study Group on the Relations between the History and Pedagogy of Math-
ematics & ESU 4—Revised Edition (pp. 509–524). Uppsala, Sweden.
Radford, L. (2008a). Iconicity and contraction: A semiotic investigation of forms of algebraic gen-
eralizations of patterns in different contexts. ZDM—The International Journal on Mathematics
Education, 40(1), 83–96.
Radford, L. (2008b). The ethics of being and knowing: Towards a cultural theory of learning. In L.
Radford, G. Schubring, & F. Seeger (Eds.), Semiotics in Mathematics Education: Epistemology,
History, Classroom, and Culture (pp. 215–234). Rotterdam: Sense Publishers.
Radford, L. (2009a). Signs, gestures, meanings: Algebraic thinking from a cultural semiotic per-
spective. Plenary Lecture presented at the Sixth Conference of European Research in Mathe-
matics education (CERME 6). Université Claude Bernard, Lyon, France, January 28–February
1, 2009.
Radford, L. (2009b). Why do gestures matter? Sensuous cognition and the palpability of mathe-
matical meanings. Educational Studies in Mathematics, 70(2), 111–126.
Radford, L., Bardini, C., & Sabena, C. (2007). Perceiving the general: The multisemiotic dimen-
sion of students’ algebraic activity. Journal for Research in Mathematics Education, 38, 507–
530.
Roth, M.-W. (2001). Gestures: Their role in teaching and learning. Review of Educational Re-
search, 71(3), 365–392.
Sabena, C., Radford, L., & Bardini, C. (2005). Synchronizing gestures, words and actions in pat-
tern generalizations. In H. L. Chick & J. L. Vincent (Eds.), Proceedings of the 29th Conference
of the International Group for the Psychology of Mathematics Education (pp. 129–136). Uni-
versity of Melbourne, Australia.
Savage-Rumbaugh, E. S., Rumbaugh, D. M., Smith, S. T., & Lawson, J. (1980). Reference: The
linguistic essential. Science, 210(4472), 922–925.
Serfati, M. (1999). La dialectique de l’indéterminé, de Viète à Frege et Russell. In M. Serfati (Ed.),
La recherche de la vérité (pp. 145–174). Paris: ACL—Les éditions du kangourou.
322 L. Radford
Serfati, M. (2006). La constitution de l’ecriture symbolique mathematique. Symbolique et inven-

tion. SMF—Gazette, 108, 101–118.
Viète, F. (1983). Nine Studies in Algebra, Geometry and Trigonometry from the Opus Restitutae
Mathematicae Analyseos, Seu Algebrâ Novâ (Translated by T. Richard Witmer). New York:
Dover.
Warren, E. (2006). Teacher actions that assist young students write generalizations in words and
in symbols. In J. Novotná, H. Moraová, M. Krátká, & N. Stehlíková (Eds.), Proceedings of
(Vol. 5, pp. 377–384). Prague, Czech Republic.
Warren, E., & Cooper, T. (2008). Generalising the pattern rule for visual growth patterns: Actions
that support 8 year olds’ thinking. Educational Studies in Mathematics, 67, 171–185.
Woepcke, F. (1853). Extrait du Fakhrî. Paris: Imprimérie Impériale.
Formation of Pattern Generalization Involving
Linear Figural Patterns Among Middle School
Students: Results of a Three-Year Study
Abstract This chapter provides an empirical account of the formation of pattern

generalization among a group of middle school students who participated in a three-
year longitudinal study. Using pre-and post-interviews and videos of intervening
teaching experiments, we document shifts in students’ ability to pattern generalize
from figural to numeric and then back to figural, including how and why they oc-
curred and consequences. The following six findings are discussed in some detail:
development of constructive and deconstructive generalizations at the middle school
level; operations needed in developing a pattern generalization; factors affecting
students’ ability to develop constructive generalizations; emergence of classroom
mathematical practices on pattern generalization; middle school students’ justifi-
cation of constructive standard generalizations, and; their justification of construc-
tive nonstandard generalizations and deconstructive generalizations. The longitudi-
nal study also highlights the conceptual significance of multiplicative thinking in
pattern generalization and the important role of sociocultural mediation in fostering
growth in generalization practices.
Research on patterning and generalization at least in the last decade has empirically
demonstrated the remarkable, albeit fundamental, view that individuals tend to see
and process the same pattern P differently. Consequently, this means they are likely
to produce different generalizations for P . For example, when we asked forty-two
This is an updated and thoroughly revised version of an earlier article that appeared in ZDM
40(1), 65–82. DOI 10.1007/s11858-007-0062-z. The ZDM article reported on findings drawn
from the first two years of the study. This chapter reports on various aspects of the three-year
study. This work was supported by Grant #DRL 044845 from the National Science Foundation
(NSF) awarded to F. D. Rivera. The opinions expressed are not necessarily those of NSF and,
thus, no endorsement should be inferred. This chapter is dedicated to Douglas Owens,
mathematics educator, in honor of his retirement.
F.D. Rivera () · J.R. Becker
Department of Mathematics, San José State University, San José, USA

324 F.D. Rivera and J.R. Becker
Square Toothpicks Pattern. Consider the sequence of toothpick squares below.
A. How many toothpicks will pattern 5 have? Draw and explain.

B. How many toothpicks will pattern 15 have? Explain.
C. Find a direct formula for the total number of toothpicks T in any pattern
number n. Explain how you obtained your answer.
D. If you obtained your formula numerically, what might it mean if you think
about it in terms of the above pattern?
E. If the pattern above is extended over several more cases, a certain pattern
uses 76 toothpicks all in all. Which pattern number is this? Explain how you
obtained your answer.
F. Diana’s direct formula is as follows: T = 4 · n − (n − 1). Is her formula
correct? Why or why not? If her formula is correct, how might she be thinking
about it? Who has the more correct formula, Diana’s formula or the formula
you obtained in part C above? Explain.
Fig. 1 Adjacent squares pattern task
undergraduate elementary majors to establish a general formula for the total number
of matchsticks at any stage in the Adjacent Squares Pattern shown in Fig. 1, Chuck
obtained his generalization “4 + (n − 1)3” in the following manner:
How many matchsticks are needed to form four squares? So ahm I’m looking for a pattern.
For every square you add three more. So let’s see. So that would be 4 plus 3 for two squares.
Plus 3 more would be for three squares. So it’s 10 matchsticks. So you have 4. So there
would be 13. So 13 plus 3 more is 16. . . . So, for three squares, it would have to be two
3s. So there’d be two 3s. Three 3s is for four squares, and four 3s for five squares. For n
squares, it would just be ahm n minus one 3s. (Rivera and Becker 2003, p. 69)
When we gave the same pattern in Fig. 1 to a group of middle school students
three times over a two-year period, first when they were in sixth grade (after a teach-
ing experiment) and then twice in seventh grade (before and after a teaching exper-
iment), all of their generalizations consistently took the form T = (n × 3) + 1. For
example, in a clinical interview prior to the Year 2 teaching experiment, Dung, in
seventh grade, initially set up a two-column table of values, listed down the pairs
(1, 4), (2, 7), and (3, 11) and noticed that “the pattern is plus 3 [referring to the de-
pendent terms].” He then concluded by saying, “the formula, it’s pattern number x 3
plus 1 equals matchsticks,” with the coefficient referring to the common difference
and the y-intercept as an adjustment value that he saw as necessary in order to match
the dependent terms. When he was then asked to justify his formula, he provided the
following faulty reasoning below in which he projected his formula onto the figures
in a rather inconsistent manner (see Fig. 2 for an illustrative version).
For 1 [square], you times it by 3, it’s 1, 2, 3 [referring to three sides of the square] plus 1
[referring to the left vertical side of the square]. For pattern 2, you count the outside sticks
Formation of Pattern Generalization Involving Linear Figural Patterns 325
Fig. 2 Dung’s justification of

his direct formula for the
Fig. 1 pattern
Fig. 3 Tile patio pattern
and you plus 1 in the middle. For pattern 3, there’s one set of 3 [referring to the last three
sticks of the third adjacent square], two sets of 3 [referring to the next two adjacent squares]
plus 1 [referring to the left vertical side of the first square].
Also, by the end of the Year 2 study, none of Dung’s classmates were able to
come up with a general form similar to Chuck’s. Further, when they were asked to
explain an imaginary student’s formula, T = 4n − (n − 1), for the pattern in Fig. 1,
they found this and other similar tasks difficult.
However, we found it interesting that when the students in Year 3 of the study
were purposefully reoriented to a multiplicative thinking approach to patterning ac-
tivity involving figural stages (i.e., pictorial patterns with known stages such as the
one shown in Fig. 1), they finally settled on figural-based generalizations. For ex-
ample, when they obtained a generalization for the Tile Patio Pattern in Fig. 3 during
a teaching experiment, they developed at least three equivalent direct formulas that
reflected the use of multiplicative reasoning (i.e., in relation to what they perceived
Che: “W = 4n + 4. So there’s like 4 squares and then you add 4 to each one
[corner]. . . . And for patio #2, 2 × 4 is 8, so 1, 2, 3, . . . , 8, then you add 4.
[There are] 4 groups of n.”
Dina: “4(n + 1) coz patio #1 you have two groups of 4, patio #2, 3 groups of 4,
etc.”
Dave: “T = 2(n + 2) + 2n. The top part, 2 + 1 = 3. Then I multiplied by 2, the

bottom, so that’s 6. And the 2n, so here’s 1 [row 2 column 1 square] and 1 here
[row 2 column 3 square], etc.”
Fig. 4 Visuoalphanumeric generalizations of 8th graders on the Fig. 3 pattern
to be the repeated parts of the pattern). In Fig. 4, the students’ constructed direct
formulas are examples of figural-based generalizations in which the alphanumeric
symbols in the formulas conveyed relationships that they have drawn figurally from
the pattern. Such representations are, in fact, effects of particular (i.e., mathematical)
ways of seeing and acquired knowledge and experience (Metzger 2006/1936).
In this chapter, we address issues relevant to the following two related questions:
What is the content and structure of algebraic generalization that middle school
students (i.e., Grades 6 to 8, ages 11 to 14) develop in the case of linear figural pat-
terns? Further, to what extent are they capable of establishing and justifying their
algebraic generalizations? By algebraic generalization of a figural pattern, we refer
to, in Radford’s (2008) words, the “[students’] capability of grasping a common-
ality noticed on some particulars (in a sequence), extending or generalizing this

commonality to all subsequent terms, and being able to use the commonality to pro-
vide a direct expression of any term of the sequence” (p. 115). The two research
questions address various aspects of what we label as pattern generalization, which
involves constructing and justifying an algebraic generalization within the means
available to a learner. Our notion of pattern generalization extends Radford’s (2008)
view to include justification. Also, we note that the above definition of algebraic
generalization shares the basic conceptual intent surrounding all processes relevant
to the task of generalization, which involves constructing an invariant and stable
structure, property, attribute, or relation from particular known cases (or samples or
domains) and extending, applying, and projecting it to the unknown cases or larger
classes of cases (Dreyfus 1991). But we further refine Dreyfus’s (1991) sense above
by acknowledging the complex of factors (cognitive, cultural, extra-cultural such as
linguistic and classroom practices, etc.) that influence the construction of a “gener-
ality” that, according to Dörfler (2008), is a way of practice of using and interpreting
“signs, like graphs or letters, are not general by themselves” (p. 1).
In addressing the first research question, we initially survey relevant research in
the area of middle school algebraic thinking. We then consider how findings in our
three-year longitudinal research at the middle grades further confirm and/or extend
the current knowledge base in this area. Our response to the second research ques-
tion is grounded on how our students dealt with factors that influenced the manner
in which they obtained their pattern generalizations. Our decision to investigate lin-
ear figural patterns has been drawn from our survey of various school mathematics
curricula across states that show value and interest in this mathematical topic and its
connections to other concepts as well.
Anticipating What Is to Come: Initial Reflections on Our

Three-Year Data from the Clinical Interviews
Table 1 provides a summary of the results of the clinical interviews before and
after every teaching experiment we conducted over the course of three years with
our middle school students beginning at sixth grade. We briefly note the following
observations:
• About 63% of the students in the Year 1 study employed figural-based strategies
in obtaining a generalization for patterns that were mostly linear in content before
a teaching experiment involving pattern generalization. But we also point out a
dramatic shift to a numerical strategy (100%) after the teaching experiment in the
same year.
• In the Year 2 clinical interviews with eight students, seven students maintained
a numerical strategy in obtaining a generalization before and after a teaching
experiment.
• In the Year 3 clinical interviews, a shift to figural-based strategies (about 69%)
occurred.
Table 1 Summary of pattern generalization

Year 1 Results Before Teaching Experiment After Teaching Experiment
(n = 29) (n = 11)
Overall Visual 63% 0%

Overall Numeric 37% 100%
Constructive Standard 0% 100%
Generalizations
Constructive Nonstandard 0% 0%
Generalizations
Deconstructive Generalizations 0% 0%
Year 2 Resultsa Before Teaching Experiment After Teaching Experiment
(n = 8) (n = 8)
Overall Visual 12% 25%

Overall Numerical 88% 75%
Increasing Patterns
Decreasing Patterns
Generalizations
Year 3a Before Teaching Experiment After Teaching Experiment
(n = 18; 5 newb ) (n = 14; 3 newb )
Overall Visual 67%c 71%

Overall Numeric 33%c 29%
Generalizations
Generalizations
a Some tasks had multiple questions

b Did not participate in earlier two-year interviews
c More visual tasks than numerical
Considering three years of collected data, in this article we extrapolate factors

that explain why such shifts in generalization strategies took place among the stu-
dents over the course of three years. Also, we provide a description of the quality,
content, and form of generalizations at each phase. Further, we point out the progress
in students’ ability to deal with various aspects and types of pattern generalization,
which was especially evident in Year 3. Our overall intent in this chapter is to de-
scribe teaching-learning conditions that enable meaningful pattern generalization to
occur at the middle school level.
Cognitive Issues Surrounding Pattern Generalization: What We

Know from Various Theoretical Perspectives and Empirical
Studies
Clarifying the Definition of Pattern Generalization
Several researchers have pointed out that the initial stage in generalization involves
“focusing” or “drawing attention” on a candidate invariant property or relation-
ship (Lobato et al. 2003), “grasping” a commonality or regularity (Radford 2006),
and “noticing” or “becoming aware” of one’s own actions in relation to the phe-
nomenon undergoing generalization (Mason et al. 2005). Lee (1996) poignantly
surfaces the central role of “perceptual agility” in patterning and generalization,
which involves “see[ing] several patterns and [a] willing[ness] to abandon those
that do not prove useful [i.e., those that do not lead to a formula]” (p. 95). Ma-
son et al. (2005) points out as well how specializing on a particular case in a
pattern on the route to a generalization necessitates acts of “paying close atten-
tion” to details, especially those aspects that change and/or stay the same, best
summarized in Mason’s (1996) well-cited felicitous phrase of “seeing the gen-
eral through the particular.” Results of our earlier work with 9th graders (Becker
and Rivera 2005) and undergraduate majors (Rivera and Becker 2003) also con-
firm such a preparatory act whereby perception—as a “way of coming to know”
an object or something property or fact about the object (Dretske 1990)—is nec-
essary and fundamental in generalization. Of course, there are other researchers
who emphasize the fundamental, genetic role of invariant acting in the construc-
tion of an intentional generalization (Dörfler 1991; Garcia-Cruz and Martinón 1997;
Iwasaki and Yamaguchi 1997). In our longitudinal study, which focuses exclusively
on the pattern generalization of figural objects, we affirm the above views about the
nature of generalization.
Our contribution to the above characterizations deals with the mutually deter-
mining relationship between individual and sociocultural activity in the formation
of pattern generalization. That is, while we acknowledge the constructivist nature
of pattern generalization among individual students (every individual sees what s/he
finds meaningful to see that influences how and what s/he constructs), collective
action—that is, shared ways of seeing—makes the above characterizations even
more meaningful than when performed in isolation.
As we have noted in the introduction, pattern generalization refers to both ac-
tions of constructing an algebraic generalization and justifying it on the basis of the
students’ repertoire of available explanatory mechanisms (Rivera 2010a, 2010b).
Constructing and justifying a generalization are two equally important tasks. In con-
structing an algebraic generalization, we expect closure in mathematical activity
via the construction of a direct formula (i.e., a closed formula in function form).
In the case of justification, in light of the cognitive level of middle school students
who have just begun learning domain-specific knowledge and practices in algebra,
we are more or less concerned with their capacity to reason, in the sense follow-
ing Hershkovitz (1998), “to understand, to explain, and to convince” (p. 29). Knuth
(2002), for instance, talks about the importance of having students perform a figural
demonstration that explains, that is, using the relevant features in a figure in order
to provide insights regarding a particular claim. Lannin’s (2005) work with 25 US
sixth graders had him pointing out how justification seemed to have been relegated
to the “realm of geometric proofs” when, in fact, students’ justifications in the con-
text of pattern generalization could “provide a window for viewing the degree to
which they see the broad nature of their generalizations and their view of what they
deem as a socially accepted justification” (p. 232).
Types of Algebraic Generalization Involving Figural Patterns
There are two basic algebraic ways of developing a pattern generalization involving
figural patterns. (For an extended list, see Rivera 2010a.) The first way involves
what we classify as constructive generalizations (CG), which refer to those direct
or closed polynomial formulas that learners construct from the known stages in a
figural pattern as a result of cognitively perceiving figures that structurally consist of
non-overlapping constituent gestalts or parts. For example, in the case of the Fig. 1
pattern, some students may perceive the stages as a sequence of growing squares
that are produced by repeatedly adding three sides to form a new square. Dung’s
formula, “pattern number times 3 plus 1 equals matchsticks” in relation to Fig. 2
exemplifies a CG that exhibits the standard linear form y = mx + b, hence, it is a
CSG. Chuck’s direct expression for the Fig. 1 pattern, “4 + (n − 1)3,” on the other
hand, is an example of a constructive nonstandard generalization (CNG) since the
terms in his expression still need further simplification.
The second way of developing a pattern generalization involves what we clas-
sify as deconstructive generalizations (DG), which refer to those direct or closed
polynomial formulas that learners construct from the known stages as a result of
cognitively perceiving figures that structurally consist of overlapping constituent
gestalts or parts. Consequently, the corresponding general formulas involve a com-
bined addition-subtraction process of separately counting each sub-configuration
and taking away parts (sides or vertices) that overlap. For example, some students
may initially infer the appropriate number of squares at each stage in the Fig. 1 pat-
tern (i.e., stage 1 has one square, stage 2 has two squares that are adjacent to each
other, stage 3 has three adjacent squares, . . . ) and then multiply that number by 4
(since there are four sides to a square) and subtract the appropriate number of over-
lapping sides (i.e., stage 2 has two groups of 4 sides with an overlapping “interior”
side, stage 3 has three groups of 4 sides with two overlapping “interior” sides, . . . ).
In a DG, further actions of deconstructing and decomposing are necessary in order
to reveal the overlapping part(s).
In pattern generalization involving figural stages, we note that because there are
many different ways of expressing a generalization for the same pattern, we fore-
ground Duval’s (2006) view about the cognitively complex requirements of semi-
otic representations—that is, a primary resolve is to assist learners to recognize
the viability and equivalence of several generalizations that are drawn from several
“semiotic representations that are produced within different representation systems”
(p. 108). For example, Dung obtained his general formula for the Fig. 1 pattern by
initially manipulating the corresponding numerical stages that he later justified fig-
urally (Fig. 2), while Chuck established his formula for the same Fig. 1 pattern from
the available figural stages. Both learners operated under two different representa-
tional systems and, thus, produced two different, but equivalent, direct expressions
for the same pattern.
Methodology
This section is divided into five sections. The first section provides information
about the middle school participants involved in the three-year study. The next two
sections provide details of the teaching experiments on pattern generalization. The
fourth section provides samples of the tasks used in the clinical interviews. The fifth
section deals with matters involving data collection and analysis and relevant study
protocols.
Classroom Contexts from Years 1 to 3 of the Study
In Fall 2005 and Fall 2006 (i.e., Years 1 and 2 of the study), the first author collabo-
rated with two middle school mathematics teachers in developing and implementing
two related design-driven teaching experiments on pattern generalization. From Fall
2007 to Spring 2008 (i.e., Year 3 of the study), the first author taught the class the
whole academic year. The second author conducted the pre- and post-clinical inter-
views with the participating students in all three years of the study. Learnings from
the pre-interviews were incorporated in the evolving teaching experiments with the
participants, and the post-interviews were meant to assess students’ abilities to es-
tablish and justify their generalizations, including the extent of influence of class-
room practices in their developing capacity to generalize. In the Year 1 study, the
sixth-grade class consisted of twenty-nine students (12 males, 17 females; mean age
of 11; most of Southeast Asian origins). In the Year 2 study, three students moved
to different schools and were replaced with six new students. In the Year 3 study,
only fifteen students from the earlier two-year project were allowed to complete
the project. They were then mixed with a new cohort of nineteen 7th and 8th grade
students (22 females, 12 males) that together comprised an Algebra 1 class.
Nature and Content of Classroom Teaching Experiments in Years 1

and 2
A basic instructional objective of the Years 1 and 2 classroom teaching experiments

on pattern generalization involves providing students with every opportunity to en-
gage in problem-solving situations that would enable them to meaningfully acquire
the formal mathematical requirements of algebraic generalization. The instructional
theory that was initially used in Years 1 and 2 was Realistic Mathematics Education
(RME). In RME, learners use models of their informal mathematical processes to as-
sist them in developing models for more formal processes. Formalizing is, thus, seen
as “growing out of their mathematical activity” and mathematizing, more generally,
involves “expanding [their] common sense” with the same reality as “experiencing”
in everyday life (Gravemeijer and Doorman 1999, p. 127).
In the Year 1 teaching experiment on pattern generalization, two algebra units
in the Mathematics-in-Context (MiC) curriculum were used. Also, taking note of
the algebra requirements of the California state standards for sixth graders, sec-
tions were selected from the units Expressions and Formulas (MiC Team 2006b)
and Building Formulas (MiC Team 2006a) that became the basis of a three-phase
classroom teaching experiment on pattern generalization. In the first two phases,
activities drawn from the two algebra units were used to foster the development of
algebraic generalization through a series of horizontal and vertical mathematization
tasks. Horizontal mathematization involves transforming real and experientially real
problems into mathematical ones by using strategies such as schematizing, discover-
ing relations and patterns, and symbolizing, while vertical mathematization involves
reorganizing mathematical ideas using different analytic tools such as generalizing
or refining of an existing model (Treffers 1987). In both units, the students initially
explored horizontal activities that allowed them to build an informal mathematical
model. They then engaged in vertical activities.
In the Expressions and Formulas unit, each section had the students starting out
with a problem situation that involved using an arrow language notation to initially
organize the situation and later to express relationships between two relevant quan-
tities. An example is shown in Fig. 5. The arrow notation was meant to articulate
the different numerical actions and operations that were needed to carry out a string
of calculations in an activity. Also, the task situations were either stated in words or
accompanied by tables, and they contained items that necessitated either a straight-
forward or a reverse calculation.
The Patterns section in the Building Formulas unit was the only one that was
used in the teaching experiment because of constraints in the stipulated sixth-grade
algebra requirements of the state’s official mathematics framework. In this sec-
tion, arrow language was employed less in favor of recursive formulas and di-
rect formulas in closed, functional form. The students dealt with problem situa-
tions that consistently contained the following tasks relevant to generalizing: ex-
tending a near generalization problem physically (for example: drawing or demon-
strating with the use of available manipulatives) and/or mentally (reasoning about
Fig. 5 Arrow notation activity (MiC 2006b, p. 21)
it logically); calculating a far generalization task (i.e., finding a total number be-
yond stage 10) using either a figural or a numerical strategy; developing a gen-
eral formula recursively and/or in closed, functional form, and; solving problems
that involve inverse or reverse operations. In all problem situations, tables were
presented and employed as an alternative representation for organizing the given
data.
Fig. 6 Two multiplication tasks
Finally, the students dealt with tasks that asked them to reason and to make judg-
ments about the equivalence of several different formulas for the same problem sit-
uation. In the third phase of the teaching experiment, they worked through several
decontextualized patterning problems whose basic structure was similar to the ones
that have been described in the paragraph above (see, for e.g., Figs. 1 and 3). Also,
they explored problems that enabled them to develop both numerical and figural
generalization.
In the Year 2 teaching experiment on pattern generalization, the same three-phase
process occurred. The seventh-grade class used Building Formulas and portions of
Patterns and Figures (MiC 2006a) in the first two phases with the third phase the
same as in the description provided above.
Nature and Content of Classroom Teaching Experiments in Year 3
The Year 3 study on pattern generalization took place in three phases. The initial
phase of learning pattern generalization focused on helping students develop a bet-
ter understanding of multiplicative thinking, which actually was the unifying thread
that connected all the algebra concepts and processes that were learned throughout
the year. The students initially investigated counting activities that emphasized mul-
tiplicative thinking. For example, the activity in Fig. 6 asked the students to establish
a mathematical expression involving multiplication. In the second phase, they ex-
plored pattern generalization activities that involve developing a structural analysis
of a pattern (i.e., in terms of what stays the same and what changes in the pattern).
In the third phase, they connected multiplicative thinking and structural analysis.
Nature and Content of Clinical Interview Tasks from Years 1 to 3
In the Year 1 clinical interview prior to the teaching experiment on pattern general-
ization, the students were given five tasks that addressed various aspects of pattern
generalization. The task shown in Fig. 7a asked the students to determine a near and
a far generalization item, with far items as arbitrarily referring to figural stages 10
and above in a given pattern and near items as pertaining to stages 9 and below. All
five tasks required students to calculate a near and a far item. Three of the five tasks
had figural patterns that show at least four consecutive initial stages. One task was
presented numerically using a table of values. The fifth task began with an interme-
diate figural stage in some pattern and the students were asked to reconstruct a set of
stages prior to the figural stage and then to use that knowledge to extend the pattern
and deal with far items. In the Year 1 clinical interview after the teaching experiment
on pattern generalization, analogous tasks were presented with some changes in the
questions. For example, the task shown in Fig. 7b uses the same task structure in
Fig. 7a but there is an increased emphasis in the following aspects of direct-formula
construction: justification; the use of numerical or figural strategies; assessing for
equivalence.
In the Year 2 clinical interviews before and after the teaching experiment on
pattern generalization, tasks similar to Fig. 7b were presented to the students with
the inclusion of two decreasing patterns.
In the Year 3 clinical interviews before and after the teaching experiment on
pattern generalization, the students were asked to justify a given direct formula of a
given figural pattern (increasing and decreasing; see, for e.g., Fig. 8a) and to obtain
several equivalent pattern generalizations for the same pattern (see, for e.g. Fig. 8b).
A semi-free construction task was also added (see Fig. 9) in response to Dörfler’s
(2008) “plea for ‘free’ generalization tasks” (p. 153). Dörfler notes that patterns with
well-defined stages impress on learners the view that “there is an expected direction
of generalizing,” which would then “intimate one and only one way [of continuing]
a figural sequence” and, consequently, harbor “a strong regulating or even restrictive
impact” on their thinking (p. 153). He recommends a different approach by asking
students to think about (figural) patterns, as follows:
How otherwise can one ask for, say, the number of matchsticks . . . in an “arbitrary” item
of the sequence? The situation would presumably be much more open if one asked simply
“How can you continue?” or “What can you change and vary in the given figures?” . . . I
rather want to hint to possible further directions for research . . . a plea for “free” general-
ization tasks not restricted by pre-given purposes. (Dörfler 2008, p. 153)
Data Collection and Analysis and Relevant Study Protocols
Each project year, we collected the following data: students’ written work on various
homework, classroom, and performance assessments involving pattern generaliza-
tion; videos and transcripts of clinical interviews before and after every teaching
How many tiles does Picture 10 have? How do you know?

How many tiles does Picture 100 have? How do you know?
Find a formula to calculate the number of tiles in Picture “n.” How did you
obtain your formula?
Fig. 7a A sample Year 1 clinical interview task prior to a teaching experiment
experiment, and; videos and transcripts of relevant classroom episodes taken during
a teaching experiment.
With respect to the analysis of data drawn from the clinical interviews, we en-
gaged in several repeated processes of individual and shared reading within and
across cases. We have carefully described this important step in several published
research papers (Becker and Rivera 2005, 2006, 2007, 2008; Rivera and Becker
2008). Basically, individual cases were analyzed, developed, and later synthesized
in order to construct individual cognitive maps with the aim of schematically cap-
turing their generalizing schemes from problem to problem. Next, those individual
cases were compared, analyzed, and categorized using grounded theory that enabled
us to develop some empirical claims about aspects of their pattern generalizing ac-
Tiles are arranged to form pictures like the ones below.

See pattern in Fig. 7a
A. Find a direct formula that enables you to calculate the number of square tiles
in Picture “n.” How did you obtain your formula? If the solution has been ob-
tained numerically, respond to the following question: Is there a way to explain
your formula from the figures?
B. How many square tiles will there be in Picture 75? Explain.
C. Can you think of another way of finding a direct formula?
D. Two 6th graders came up with the following two formulas:
Kevin’s direct formula is: T = (n × 2) + (n × 2) + 1, where n means Picture
number and T means total number of squares. Is his formula correct? Why or
why not?
Melanie’s direct formula is: T = (n × 2) + 1 + (n × 2) + 1 − 1, where n and T
mean the same thing as in (D) above. Is her formula correct? Why or why not?
Which formula is correct: Kevin’s formula, Melanie’s formula, or your formula?
Explain.
Fig. 7b Analogous Year 1 task given after a teaching experiment on pattern generalization
tivity. The results, findings, and observations we have developed were consequences
of several iterated processes of reading and analyzing the within- and across-case
studies in order to ensure greater validity.
The first author was also responsible for the analysis of data drawn from the
classroom episodes. Relevant transcripts of key classroom episodes were obtained in
order to provide additional support about the claim of shifts in pattern generalization
practices of the participating students over the course of the longitudinal study.
Undergraduate student assistants videotaped all the classroom and collaborative
group sessions involving pattern generalization. Each teaching experiment lasted
three consecutive weeks on average. All clinical interviews were also videotaped.
During a clinical interview, each student was requested to think aloud and to use the
available and relevant manipulatives (pattern blocks, calculators, centimeter graph-
ing paper, etc.) to help them deal with the tasks. In cases when a student incorrectly
performed a calculation, the interviewer (second author) sought clarification to bet-
ter assess the nature of the error. Also, in cases when a student had a difficult time
articulating a verbal response, the interviewer sought clarification until both of them
felt satisfied with the response.
A. Consider the sequence of three figures shown below.
Three 8th grade students have been asked to find the total number of stars (S) at
any given stage number (n). Explain how each student might be thinking of his or
her formula.
Marcia’s formula: S = n · 3 + 1
Pete’s formula: S = (n + 1) · 3 − 2
Jayme’s formula: S = (n · 4 + 1) − n
B. Consider the sequence of three figures below.
Two seventh grade students came up with two different direct formulas for the
total number of circles (C) for any step number (n). How might each student be
thinking about his formula?
Jake: C = −n + 8
Bharath: C = 4 + 4 − n
Fig. 8a Two sample Year 3 tasks given before and after a teaching experiment on pattern general-
ization (increasing and decreasing figural patterns)
Findings and Discussion Part 1: Accounting for Constructive and

Deconstructive Generalizations
Findings in Our Study An analysis of Table 1 shows the predominant use of

CSGs from Years 1 to 3. In fact, CNGs were not evident until Year 3. The pro-
cess used to establish a CSG was predominantly numerical in Years 1 and 2 with
a mean of 87.5% but a shift to figural took place in Year 3 at about the same per-
centage. Also, the generalizations that were developed in the case of all increasing
and decreasing linear patterns were CSGs. Further, while the students were success-
ful (100%) in establishing CSGs in the case of increasing linear patterns, they had
considerable difficulty in the case of decreasing linear patterns (a success rate of
38% before a teaching experiment and 75% after). In the case of DGs, while none
of the students could construct them by the end of the Year 2 study, they, however,
Obtain two different ways (or formulas) that will enable you to find the total num-
ber of gray squares (S) at any stage number (n). Then explain why you think each
way (or formula) makes sense to you.
1. Formula 1: _________________________
Explanation:
2. Formula 2: _________________________
Explanation:
Fig. 8b A sample Year 3 task given before and after a teaching experiment on pattern generaliza-
tion
had considerable success in explaining them (from 50% to 100% before and after a
teaching experiment, respectively).
Discussion Results drawn from our Years 1 and 2 study actually confirm find-
ings from several research studies at the middle school level that also provide suf-
ficient evidence indicating students’ predisposition toward producing more CSGs
than DGs. For example, when Taplin and Robertson (1995) asked 40 Australian 7th
graders to establish a generalization for the pattern sequence in Fig. 1, while none
The figure below shows five gray circles that enclose a white circle. Call it stage 1.
1. First, find a way to continue the above figure so that you end up with several
stages that altogether form a pattern of figures. Draw your figures below.
Your Stage 2:
Your Stage 3:
Your Stage 4:
2. Next, try to find a formula for the pattern of figures you constructed above. If a
formula is not possible, describe your pattern in a general way.
Fig. 9 A semi-free construction task given before and after a teaching experiment on pattern
generalization in the Year 3 study
of them could state an algebraic generalization, their incipient generalizations took

the form of CS verbal statements. Seven students perceived four toothpicks that per-
tained to the original square in stage 1 and the repeated addition of 3 toothpicks each
time from stage to stage. There were eight students who offered the CN verbal gen-
eralization, 3(n − 1) + 4, although none offered an articulation that was as clear as
Chuck’s in the introduction above. Only one student began to think about the pattern
in a deconstructive way; however, that student was not able to figure out the number
of toothpicks that needed to be taken away despite seeing the pattern as consisting of
overlapping squares. When the same problem was given to a cohort of four hundred
thirty 12- to 15-year old Australian students, findings from English and Warren’s
(1998) study also showed that, among the less than 40% of students who success-
fully obtained a generalization, they expressed their generalities on this and other
patterning tasks in constructive terms similar to what Taplin and Robertson (1995)
found. For example, a student developed the general expression 2x + (x + 1), where
2x refers to the top and bottom row sticks and (x + 1) to the column sticks in the
Fig. 1 pattern, after seeing two invariant properties within and across stages.
While descriptions of CGs for figural patterns abound, the more important
question involves the formation of CSG and CNG, in particular, how does con-
structive objectification come about? First, Radford (2003) notes that there are
different semiotic means of objectification in relation to pattern stages, that is,
possibly different ways in visibly surfacing attributes and properties of, or re-
lationships among, stages with the use of signs and relevant processes or oper-
ations. Second, Radford (2003, 2006) advances the view that there are at least
three layers of algebraic generalization—factual, contextual, and symbolic—based
on his three-year longitudinal work with middle and junior high school students.
Third, purposeful instruction through well-designed classroom teaching experi-
ments could scaffold the development of closed forms of constructive generaliza-

tions in middle school children (Lannin et al. 2006; Martino and Maher 1999;
Steele and Johanning 2004). In the following paragraphs, we dwell on cognitive-
related issues at the entry stage of generality, that is, factual, since both contextual
and symbolic layers are marked indications of further essentializing and increasing
formality on the basis of the stated factual expressions.
At the pre-symbolic stage of factual generalizing involving increasing linear pat-
terns, students oftentimes start with a recursive relation that is both additive and
arithmetical in nature. As a matter of fact, studies done in different settings (for
e.g., countries) and in different contexts (prior to formal instruction in algebra,
during and/or after a teaching experiment, etc.) with middle school students have
asserted the use of recursion as the entry (and, in some cases, the final) stage in
factual generalizing (Becker and Rivera 2006; Bishop 2000; Lannin et al. 2006;
Orton et al. 1999; Radford 2003; Sasman et al. 1999; Swafford and Langrall 2000).
For example, in the case of increasing figural sequences, it is usually easy to first
perceive the dependent terms as constantly being increased by a common difference.
As soon as this takes place, students’ thinking is then characterized in two ways, as
follows:
• First, they see two consecutive stages as being different and, using the method
of “differencing” (Orton and Orton 1999, p. 107), the same number of objects is
constantly being added from one stage to the next, leading to a recursive, arith-
metical generalization (of the type un = un−1 + c, where c is the common differ-
ence). Then, some students further develop emergent factual generalizations from
the arithmetical generalization. Two possible factual generalizations involving the
Fig. 1 pattern are as follows: 4 + 3 + 3 + 3 + · · · ; 1 + 3 + 3 + 3 + · · · .
• Second, a structural similarity is observed among and, thus, connects two or more
stages in a relational way. Especially in the case of increasing linear patterns that
figurally demonstrate growth, constructing a succeeding stage from a preceding
one oftentimes involves a straightforward process of simply adding a constant
number of objects on particular locations of the preceding stage. That is, the basic
structure of the unit figure is perceived to stay the same despite the fact that equal
amounts of objects are conjoined in various parts of the figure in a particular,
predictable manner. Such method of construction does not necessitate making a
figural change (in Duval’s 1998 sense) on the part of the learner.
Radford (2003) further notes how in the factual stage of generalizing, invari-
ant acting from one stage to the next operates at the concrete level that eventually
leads to the abstraction of a numerical or operational scheme for the figural pattern.
Hence, generalizations that have been mediated by such actions tend to be conse-
quentially constructive and almost always standard (whether rhetorical, syncopated,
or symbolic in form).
Even with pattern generalization tasks that require middle school students to first
specialize (in Mason’s 1996 sense) on the route to establishing a generality as a
consequence of not being provided with the usual consecutive sequence of figural
stages, many of them were predisposed to establishing constructive generalizations.
For example, Swafford and Langrall (2000) asked ten middle- to high-math achiev-
ing 6th grade students to solve the Fig. 3 pattern prior to a formal course in algebra.
In their case, the task began with a drawn 10 by 10 square grid in which the four bor-
ders of the grid are shaded. The students were asked to figure out the total number of
squares on the border, and the task was repeated in a 5 by 5 grid. The students were
then asked to describe how to determine the total number of squares in the border of
an N by N grid. Results on this task show that, while none of the students offered a
recursive rule, the general verbal descriptions ranged in form from the constructive
to the deconstructive. When translated in symbolic form, two of the verbal general-
izations were CNGs and obeyed the following forms: (1) n + n + (n − 2) + (n − 2);
(2) n + (n − 1) + (n − 1) + (n − 2). Only one student in their study offered a verbal
DG that followed the form 4n − 4. When the above task and other similar ones were
given to eight 7th grade students in Steele and Johanning’s (2004) study in the con-
text of a problem-solving enriched teaching experiment, only three students came
up with DGs.
Findings and Discussion Part 2: Understanding the Operations

Needed in Developing a Pattern Generalization
Findings in Our Study This section addresses issues our middle school partici-
pants had relative to developing DGs and CGs involving decreasing linear patterns.
Due to space constraints, we illustrate in this section students’ difficulties with de-
creasing linear patterns. Decreasing linear patterns could be expressed as CSGs in
the form y = mx + b, where m < 0. In Year 2, the students’ primary cognitive dif-
ficulty with decreasing patterns prior to a teaching experiment (with a success rate
of 38%) was how to handle negative differencing and, especially, how to perform
operations involving negative and positive integers. While we found that they were
attempting to transfer the existing generalization process they developed in the case
of increasing linear patterns, they could not, however, make sense of the negative
integers and the relevant operations that were used with such types of numbers.
For example, in a clinical interview prior to a teaching experiment, Tamara easily
obtained CSGs on two increasing linear patterns. Also, she was able to justify given
several CNGs relative to another figural pattern. When she was then presented with
the Losing Squares Pattern in Fig. 10 as a third task to analyze, she immediately saw
that every stage after the first involves “minusing 2” squares. She used multiplication
to count the total number of squares at each stage. In obtaining a direct formula,
however, she was perturbed by the negative value of the common difference and
said,
I was trying to think of, just like the last time, I was trying to get a formula. . . . I was
thinking of trying to do with the stage number but I don’t get it.
The presence of the negative difference, including the necessity of multiplying

two differently signed numbers, partially and significantly hindered her from ap-
plying the method she learned in the case of increasing patterns. Tamara’s situation
Take a look at the three different stages in the design below.
1. How many squares are there in stage 1? stage 2? stage 3?

2. How many squares are there in stage 10? How do you know for sure?
3. How many squares are there in stage 15? How do you know for sure?
4. Find a direct formula for the total number of squares in stage n, where n is a
positive integer.
If you obtained your formula numerically, what might it mean in the context of
the above pattern?
5. How many squares are there in stage 20? What might your answer mean in the
context of the given pattern?
6. For what stage number will there be no more squares left? How do you know
for sure?
Fig. 10 Losing squares pattern
exemplifies the thinking of those students interviewed who were also unsuccessful
and, thus, unable to overcome such difficulties before (about 62%) and even after
(25%) the teaching experiment in Year 2. Further, her thinking in relation to decreas-
ing patterns after the teaching experiment captures the actions of those students who
were also successful by the end of the Year 2 study (about 75%).
Discussion A relevant issue we considered in relation to linear pattern generaliza-

tion involves the operations that are employed in formulating CGs and DGs. Devel-
oping CGs in the case of increasing linear patterns requires students to have solid
grounding in addition and multiplication of whole numbers. Developing DGs and
CGs in the case of decreasing linear patterns necessitates knowledge relevant to
manipulating addition, subtraction, and multiplication of integers (cf: English and
Warren 1998; Stacey and MacGregor 2001).
Gelman and colleagues (Gelman 1993; Gelman and Williams 1998; Hartnett and
Gelman 1998) have advanced and empirically justified a rational constructivist ac-
count of cognitive development among young children that presupposes the exis-
tence of innate or core skeletal mental structures (such as arithmetical structures)
that enable them to easily develop and process new information as long as it is con-
sistent with their core structures. Hartnett and Gelman (1998) write:
As long as inputs are consistent with what is known, then novice’s active participation in
their Learning can facilitate knowledge acquisition. But when available mental structures
are not consistent with the inputs meant to foster new Learning, such self-initiated interpre-
tative tendencies can get in the way (p. 342).
Among middle school students who develop CSGs and CNGs in the case of
increasing linear patterns, perhaps it is the case that their generalizations, which in-
volve using the operations of addition and multiplication of whole numbers, map
easily onto their current understanding of what numbers are and how such entities
are used, represented, and manipulated. Thus, constructive generalizing will proceed
naturally and smoothly. Moreover, middle school students are likely to associate in-
creasing growth patterns with counting objects over several non-overlapping con-
stituent gestalts and then use the addition and multiplication of counting numbers
as useful operators in obtaining a final count. Hence, their core arithmetical struc-
tures assist in this developing capacity towards making constructive generalizations.
This being the case, it is less likely that students will apprehend increasing patterns
as being embedded in a figural process that involves the operation of subtraction
via, say, the utilization of a figural change process of seeing sub-configurations and
removing overlapping parts as in all cases of DGs.
In the case of decreasing linear patterns, students like Tamara have to first
broaden their knowledge of multiplication to include two factors having opposite
signs in order to establish, say, the formula S = −2 × n + 34. However, we note
that, as with the other students, while Tamara was able to explain the terms in her
CSG consistently across increasing linear patterns, she was unable to justify the
formulas she established for decreasing linear patterns.
Findings and Discussion Part 3: Factors Affecting Students’

Ability to Develop CGs
Even when middle school students are found sufficiently capable of producing more
CSGs than DGs and CNGs, we discuss three additional factors that influence their
ability to establish the former.
Findings in Our Study In our Year 1 study, 69% of our sixth grade students’ ini-
tial verbal generalizations (correct and incorrect) in relation to the T Circle Pattern
in Fig. 11 could not be conveniently translated in closed form. Table 2 provides a
list of these verbal responses. The remaining 31% provided verbal generalizations
that are considered to be algebraically useful. The responses are listed in Table 3.
For example, when Dina was asked to obtain a generalization for the total number
of dots in the Fig. 10 pattern, her circle chip-based stages in Fig. 12 revealed the
Fig. 11 T circle pattern
Table 2 Summary of non-algebraically useful verbal responses in relation to the Fig. 11 pattern
(20 out of 29 students)
Frequency Verbal Generalizations
14 Figure 1 has 1 circle, figure 2 has 3 circles, figure 3 has 5 circles, . . . , figure 10 has
19 circles. So figure 100 has 190 circles since 10 × 10 = 100 and 10 × 19 = 190
1 Figure 1 has 1 circle, figure 2 has 3. So you’re adding two each time. So figure 100
has 100 × 2 = 200 but the numbers are always 1 less than the actual multiple of 2.
So figure 100 has 199 circles
1 Since the pattern is always adding 2, it’s the same thing as multiplying by 2. So
figure 100 should have 100 × 2 = 200 circles
1 Always keep adding 2
1 There is a pattern in the units digit. If figure 10 has 19 circles, then figure 20 has 29
circles, figure 30 has 39 circles, etc.
1 There is a pattern in the units digit. If figure 5 has 9 circles and figure 10 has 19
circles, then figure 15 should have 29 circles, figure 20 has 39 circles, etc.
1 Since figure 1 has 1 circle, then figure 5 has 5 circles. Since figure 2 has 2 row circles
and 1 column circle, then figure 6 has 6 row circles and 1 column circle. Since figure
3 has 3 row circles and 2 column circles, then figure 7 has 7 row circles and 2 column
circles. Since figure 4 has 4 row circles and 3 column circles, then figure 8 has 8 row
circles and 3 column circles, etc.
Table 3 Summary of algebraically useful verbal responses in relation to the Fig. 11 pattern (9 out
of 29 students)
Frequency Verbal Generalizations
7 Since figure 2 has 1 circle in the column and 2 circles in the row, and figure 3 has
2 circles in the column and 3 circles in the row, then figure 10 has 9 circles in the
column and 10 circles in the row
2 The pattern keeps adding two by adding a circle on the right side of the row and
another circle at the top of the column
extent of her perception of the stages, that is, the stages just kept going up by twos
and nothing else. Those who used a figural multiplicative strategy, on the other
hand, initially employed analogical reasoning. Employing multiple instead of unit
counting, their general statements reflect the invariant structure they thought was
evident from stage to stage.
Fig. 12 Dina’s interpretation of the Fig. 10 T circles pattern using colored chips
Fig. 13 The triangular

toothpick pattern
Discussion Language and the use of variables and analogies are all important fac-
tors in direct-formula construction involving all CGs and DGs. Concerning lan-
guage, Stacey and MacGregor (2001) point out the importance and necessity of the
“verbal description phase” in the “process of recognizing a function and expressing
it algebraically” (p. 150). Also, based on results drawn from Year 7 to Year 10 (ages
12 to 15) Australian students and their reflections on a national recommendation for
a pattern-based approach to algebra, MacGregor and Stacey (MacGregor and Stacey
1992; Stacey and MacGregor, 2001) surface students’ difficulties in “Transition[ing]
from a verbal expression to an algebra rule” since “students with poor English skills”
are oftentimes unable to “construct a coherent verbal description” and many of their
“verbal description[s] cannot be [conveniently and logically] translated directly to
algebra” (MacGregor and Stacey 1992, pp. 369–370).
Concerning variables, Radford (2006) points out the problematic status of vari-
able use in students’ expressions of generality. In his layers of algebraic general-
ization, the presence and use of variables in their proper form and meaning sig-
nal the accomplishment of the final stage of symbolic generalization. He notes that
while some students may display knowledge of using algebraic language to express
a CG, the variables used in such contexts have to reach their objective state of being
desubjectified and disembodied placeholders. Radford’s (2001) characterization of
algebraic language at the layer of symbolic generalizing is best exemplified in the
thinking of two small groups of 8th graders on the Triangular Toothpick Pattern in
Fig. 13 who obtained the generalities (n + n) + 1 and (n + 1) + n and perceived them
as being different on the basis of having been derived from two different actions.
Radford (2001) astutely points out that the use of variables to convey a generality
has to evolve. In particular, when students employ a variable in relation to the inde-
pendent term of the general expression, they need to eventually see that the variable
has to shift meaning from being a “dynamic general descriptor of the figures in [a]
pattern” to being “a generic number in a mathematical formula” (Radford 2000,

p. 255). Thus, their general algebraic language in expressions of generality involves
semantically transposing the independent variable from its ordinal character (index-
ical, positional, deictically-based) to the cardinal (as a “number capable of being
arithmetically operated” (ibid.)).
Concerning analogies, since all linear patterns could take the CSG formula
y = mx + b, perceiving and using analogies can quickly facilitate the generaliz-
ing process. While middle school students are likely to offer a constructive recur-
sive expression, some have been documented to be capable of developing construc-
tive analogical expressions in varying formats even prior to a formal study of alge-
bra and algebraic notation (Becker and Rivera 2006; Bishop 2000; Lannin 2005;
Stacey 1989; Swafford and Langrall 2000). Performing analogy involves “per-
ceiv[ing] and operat[ing] on the basis of corresponding structural similarity in ob-
jects whose surface features are not necessarily similar” (Richland et al. 2004,
pp. 37–38).
In our Year 1 study, we identified a possible source of difficulty among the sixth
grade students in relation to constructing algebraically useful analogies for partic-
ular figural-based patterns. We distinguished between students who perceived and
generalized additively from those who employed a multiplicative approach. Those
students who used a figural additive strategy, on the one hand, were not thinking in
analogical terms at all, and they frequently employed counting objects one by one
from stage to stage. Further, when some of them were provided with manipulatives
to copy figural stages that had been drawn on paper, their manipulative-constructed
stages did not preserve the structure of individual stages like Dina in relation to
Fig. 12 above; they, in fact, used the available manipulatives only as counters.
Findings and Discussion Part 4: A Three-Year Account

of Classroom Mathematical Practices that Encouraged
the Formation of Generalization Among Our Middle School
Students
Findings in Our Study In this section, we describe how our middle school par-
ticipants established six classroom mathematical practices on pattern generalization
over the course of three teaching experiments that occurred over three consecutive
years. We note that very few studies at the middle school level have focused on
the manner in which students develop pattern generalizations over some extended
timeframe. Thus, in this section, we aim to highlight how certain legitimate mathe-
matical practices could be viewed not as conceptual, received objects that learners
simply acquire rather unproblematically but as part of their individual and sociocul-
tural developmental transformations drawn from and embodied in their activity with
other learners.
Year 1 Classroom Practices: From Figurally- to

Numerically-Driven CSGs
In the Year 1 study, four pattern generalization practices were constructed and be-
came taken-as-shared in collaborative activity. Two of the practices had their origins
in the first MiC unit they used in class (i.e., Expressions and Formulas). First, the
students initially employed arrow strings as a method for organizing a sequence
of arithmetical operations (see, for e.g., Fig. 5). They also explored the notion of
equivalence through arrow strings that could either be shortened or lengthened de-
pending on the nature of the numbers being manipulated. Second, the use of the
arrow strings evolved as the students were asked to deal with more complicated
problem situations that were still arithmetical in context. In several more sessions,
they developed a connection between constructing an arrow string and a formula in
such a way that they used arrow strings as a means of describing invariant opera-
tional schemes in the context of generalizing situations. In transitioning from the
arrow strings to formulas, the students developed an understanding that a formula,
like the arrow strings, consists of a starting number or input, a rule in the form of a
sequence of operations, and an output value or expression.
Two additional practices emerged when the students began to generalize figural-
based patterns that have been initially drawn from the Patterns section in the MiC
unit Building Formulas. The third classroom practice that became taken-as-shared
involves generalizing figurally and is exemplified in the classroom episode below
in which the students were engaged in developing a formula for the total number
of grey and white tiles for new path number n whose figural stages are shown in
Figs. 14a and 14b. Initially, the students explored specific instances when n = 3 to
5, 9, 15, 30, and 100. In particular, they were not merely asked to obtain the output
values but also to describe the patterns without actually drawing them explicitly.
The class then generated a recursive rule for each tile type. In the episode below, the
discussion that took place between the first author and the class shifted from the use
of recursive rules to the construction of general expressions in relation to the new
path patterns.
Fig. 14a Urvashi’s tile

patterns (MiC Team 2006a,
p. 2)
Fig. 14b Urvashi’s design for new path 3 (MiC Team, 2006a, p. 3)
FDR: Suppose I want you to describe new path 1,025. That’s a big number. I
want you to figure out the total number of white and grey tiles for new
path 1,025. Emily, how do we do this?
Emily: The whites will be 2,054?
Ford: That’s the grey.
Emily: It is?
Ford: Yeah, the white’s the middle.
Emily: 1,029.
FDR: Why 1,029?
Emily: Because it’s in the middle and in the corners it has four.
FDR: Alright. What about the grey ones? Mark.
Mark: The grey ones are 2,052.
FDR: Why 2,052?
Mark: Because you added the top and the bottom and then you add the two
middle.
FDR: Okay, this will be a challenge for some of you. Can you find a formula
for me? Suppose, I say, I’m going to use a variable, new path number n.
n could mean 1, 2, 3, 4, all the way to 1025. All the way to a billion.
Dung: n plus 4 equals white.
FDR: Why n + 4 equals white?
Dung: Coz n is the number of whites in the middle plus 4 whites on the sides.
FDR: Does that make sense? [Students nodded in agreement.] What about the
grey ones? The grey ones are a bit more difficult. What’s a formula for
the number of grey ones?
Che: n times 2 and then you plus 2.
FDR: It’s n × 2 + 2. What about if I express it as n plus?
Deb: n plus n plus 2.
FDR: n + n + 2. Are they the same?
Jack: Yes.
FDR: Why?
Nora: You have two grey ones.
FDR: Yes, you have the two gray ones plus the two on both sides. So now if I
know these formulas here, can I figure out new path number 50,000?
Students: Yeah.
FDR: So how do we do this, using the formula here. Number of whites. n plus 4
for whites. What do we do?
Tamara: It’s 50,004.
FDR: What about the grey ones?
Mark: 100,002.
One indication of the students’ individual appropriation of learnings from the
above social event involves their work on succeeding figural patterns. The formulas
they established were all CSGs that they oftentimes justified in figural terms. From
the above discussion, the students acquired an understanding of using figural gen-
eralizing in explicitly articulating structural similarities among the available pattern
Fig. 15 Two layer circles

pattern
stages and, hence, figurally identifying properties or relationships that remained sta-
ble and invariant over a sequence of known stages. Further, they learned to express
those properties or relationships in algebraic form, including the need to justify the
reasonableness and validity of the direct formulas. We classify such formulas as
figural-based representations.
The fourth classroom practice came about when the students tackled the Two
Layer Circles Pattern (Fig. 15). All the students initially perceived a recursive rela-
tion with the constant addition of one circle per layer. Two groups of students offered
the figural-based formula C = (n + 1) + (n + 2), where n represents figure number
and C stands for the total number of circles, which they established analogically.
That is, since Fig. 1 had two and three circle rows, Fig. 2 had three and four circle
rows, and so on, then figure n had to have (n + 1) and (n + 2) circle rows. The first
author then suggested organizing the two sets of numerical values in the form of a
table without making any recommendation that might have encouraged a numerical
strategy. The basic purpose in introducing the table in several classroom instances
was primarily to foster students’ growth in their representational skills, that is, pat-
terns could also be expressed in tabular form. In the classroom episode below, Anna
shared her group’s thinking with the class which eventually was taken as shared
and became the fourth classroom practice, that of generalizing numerically using
differencing, which was reflective of an appropriation of a standard institutional nu-
merical strategy.
Anna: We made up a formula. Like we got the figures until figure 5, and we tried
it with other ones. We got n × 2 + 3, where n is the figure number and
timesed it by 2. So 5 × 2 equals 10, plus 3, that’s 13. So for figure 25, it’s
53.
FDR: I like that formula. So tell me more. So your formula is?
Anna: n × 2 + 3.
FDR: So how did you figure this out?
Anna: First we were like making the numbers to 25. We kept adding 2 and for
figure 25, it was 53.
FDR: Wait. So you kept adding all the way to 25?
Anna: Yeah. . . . Then we used our chart. Then finally we figured out that if we
timesed by 2 the figures and plus 3, that would give us the answer.
FDR: Does that make sense? [Students nodded in agreement.] So what Anna
was suggesting was that if you look at the chart here, Anna was suggest-
ing that you multiply the figure number by 2, say, what’s 1 × 2?
Tamara: 2.
FDR: 2. And then how did you [referring to Anna’s group] figure out the 3 here?
Anna: Because we also timesed it with figure number 13.
FDR: What did you have for figure 13?

Anna: That was 29. And then 13 × 2 equals 26 plus 3.
FDR: Alright, does that work? So what they were actually doing is this. They
noticed that if you look at the table, it’s always adding by 2. You see this?
[Students nodded.] They were suggesting that if you multiply this number
here [referring to the common difference 2 by figure number, say figure
number 1, what’s 1 × 2?
Students: 2.
FDR: Now what do you need to get to 5? What more do you need to get to 5?
[Some students said “3” while others said “4.”] Is it 4 or 3?
Students: 3.
FDR: It’s 3 more. So what is 1 × 2?
Students: 2.
FDR: Plus 3?
Students: 5. [The class tested the formula when n = 2, 3, and 25.]
Year 2 Practice: Continued Use of Numerically-Driven CSGs and a

Refinement in the Case of Decreasing Linear Patterns
In the Year 2 study, the students were once again involved in a teaching experiment
that focused on linear patterning. While the first author observed that the students,
in seventh grade, seemed to have remembered how to generalize patterns figurally
(weak) and numerically (strong), results of our clinical interviews with a subgroup
of ten students prior to the teaching experiment confirmed this observation.
In the classroom episode below, the students were asked to obtain an algebraic
generalization for increasing and decreasing linear patterns in both figural and nu-
merical forms. Emma and her group (with Dave below as a member) have been con-
sistently applying the shared practice of generalizing numerically. However, Emma
introduced her process of “zeroing out” in the case of decreasing linear patterns that
resulted in a further refinement of the numerical generalizing process.
FDR: Alright. So I have my x and my y. [FDR sets up a table of values consist-
ing of the following pairs: (1, 17), (2, 14), (3, 11), (4, 8), (5, 5), (6, 2).]
So what’s the answer to this one?
Dave: y = −3x + 20. [FDR writes the formula on the board.]
FDR: This is always the problem, here [pointing to the constant 20]. Before we
figure that out, how did you figure out the −3?
Dave: The difference between the ys, between the numbers.
FDR: So what’s happening here [referring to the dependent terms]. Is this in-
creasing by 3 or decreasing by 3?
Students: Decreasing by 3.
FDR: So if it’s decreasing by 3, what’s our notation?
Students: Negative.
FDR: Alright, so negative 3. So this one is clear [referring to the slope]. Look
at this. This one I get [the slope]. If you keep doing that [i.e., differenc-
ing], it’s always true. That’s why you have this. The difficult part is this
[referring to the constant 20].
Emma raised her hand and argued as follows:
Emma: If you did a negative times a positive, it’s gonna be a negative. So what I’d
do is zero it out.
FDR: So what do you mean by zero out?
Emma: So like if it’s −3 times 1, that’s −3 [referring to the product of the common
difference (−3) and the first independent term (1)]. . . . So I’d zero out by
adding 3.
FDR: So you try to zero out by adding 3. So, what does that mean?
Emma: Coz a −3 plus 3 equals 0.
FDR: So what’s the purpose of zeroing out?
Emma: So it’s easier to add to 17. Coz if it’s 0, all you have to do is add 17.
FDR: So you’re suggesting if you’re adding 3 here, if this is −3 plus 3, that goes
0. So what do you do with the plus 3 here?
Emma: Just remember it and write it down.
FDR: Suppose I remember it, adding 3. So how does that help me?
Emma: Then ahm it’s easier to add to 17. So just add 17 [to 3 to get 20].
The class then tried Emma’s method in a different example. The first author asked
the class to first generate a table of values, and they came up with the following
(x, y) pairs: (1, 10), (2, 8), (3, 6), (4, 4), (5, 2). Using Emma’s method, one student
offered the general formula y = −2x + 12, where the constant 12 was obtained after
initially adding the common difference and its opposite to get 0 (i.e., −2 + 2 = 0)
and then adding 2 to the first dependent term to yield the constant value of 12 (i.e.,
2 + 10 = 12). The class then verified that the formula worked in any instance of the
sequence. Finally, when the first author asked if there was a limitation to Emma’s
strategy, Emma quickly pointed out that “it only works for 1” (i.e., when the case of
n = 1 is known) and that her method would fail when the initial independent term
was any other number besides 1. Hence, the fifth mathematical practice that became
taken-as-shared was generalizing numerically using Emma’s zeroing out strategy,
which was a further refinement of an institutional practice involving decreasing lin-
ear patterns.
Year 3 Practices: A Third Shift Back to Figural-based

Generalization and the Consequent Occurrence of CSGs, CNGs,
and DGs
Prior to the Year 3 teaching experiment on pattern generalization, the students ex-
plored activities involving multiplicative thinking. In the following episode, they
obtained a mathematical expression for the two sets of figures shown in Fig. 6.
FDR: So what mathematical expression corresponds to what you see here [re-
ferring to the top set]?
Francis: 6 circles.
FDR: Yes, there are 6 circles but I want a mathematical expression that shows
how you got 6.
David: Add them one by one.
FDR: Yes, you can certainly add them one by one. But are there other ways of
getting 6?
Eric: 2 times 3. [FDR writes the answer on the board.]
FDR: So what do you mean by 2 × 3, Eric?
Eric: It means 2 threes.
FDR: Uhum, 2 threes or we say 2 groups of threes. Does that make sense?
[Students nod in agreement.] Okay, so now what expression works with
the second item here [referring to the bottom set]?
Salina: Three groups of 6.
FDR: Uhum, and how do we write that using multiplication?
Students: 3 times 6.
FDR: Times meaning what?
Salina: Groups of.
For homework, the students were given similar figural tasks that required them to
construct different mathematical expressions with a focus on articulating multiplica-
tive relationships.
The following day, they worked in pairs to obtain a pattern generalization for the
Fig. 3 pattern. Considering the fact that there were old and new participants in the
Year 3 study, the first author and the classroom teacher saw to it that every table
that had two pairs of students had at least one experienced student who could guide
the remaining table members in setting up a direct formula. During the classroom
discussion, the students offered three different constructive generalizations (Fig. 4)
that the class then assessed for equivalence. In obtaining a pattern generalization, the
students first addressed structural issues of what stayed the same and what changed
from stage to stage. Then they found a multiplicative expression for those aspects or
parts that changed from stage to stage and then added a number corresponding to the
remaining parts that stayed the same. This process became the sixth mathematical
practice that was taken as shared in class.
A further refinement in this figural-based strategy occurred when the students
began to inspect a particular stage number in a pattern in terms of multiples of
the stage number and then either added or subtracted a number corresponding to
the remaining parts. For example, Che in Fig. 3 circled four groups of stage 1,
four groups of stage 2, four groups of stage 3, . . . , four groups of stage n, which
justifies her use of the expression 4n. Then she saw added 4 corresponding to the
four corners, which led her to conclude that her formula made sense.

Capability in Justifying CSGs
Findings in Our Study Results of the Year 1 teaching experiment indicate dif-
fering levels of competence in providing justifications. In particular, based on a
follow-up clinical interview with nine students after the teaching experiment on
pattern generalization, they justified in several different ways on five linear patterns,
as follows:
1. They employed extension generation (7%), which involves using more examples
to verify the correctness of a formula.
2. Some used a generic case (7%), which involves describing a perceived structural
similarity in an imagined general instance.
3. Some employed formula projection (22%), a figural-based explanation that in-
volves demonstrating the validity of a direct formula as it is seen on the given
figural stages.
4. Some used formula appearance match (71%), a numerical-based explanation
that involves merely fitting the formula onto the corresponding generated table of
values that have been initially drawn from the figural stages (Rivera and Becker
2009a, 2009b).
We also note that, in our study, because the students in Year 1 initially developed
the emergent practice of figural-based generalizing, they were in fact constructing
and validating their direct formulas at the same time. For example, Dung estab-
lished and justified his direct expression n + 4 for the total number of white tiles in
Fig. 14b as soon as he saw “the number of white [square tiles] in the middle plus
[the] 4 white [tiles] on the sides.” Also, Che, Deb, and Nora established and justified
their direct expressions, n × 2 + 2, when they perceived “two grey [squares] plus
the two squares on both sides [in a given figural stage].” All four students came up
with their justifications above after empirically verifying them on several extensions
and then either employing formula projection or imagining a generic case that high-
lights the invariant properties common to all stages. The formula appearance match
was used only later after the class developed the emergent practice of generalizing
numerically.
When the students in our Year 1 study fully appropriated the above numerical
strategies in establishing CSGs, as exemplified in the thinking of Anna and Emma
above in relation to the Fig. 14b pattern, we observed a shift that took place from
a figural to a numerical mode of generalizing among them. In fact, in both the pre-
and post-clinical interviews in the Year 2 study, very few (about 19%) initiated a
figural-based approach with most of them developing numerical-based CSGs (about
82%). Consequently, the shift affected their capacity to justify algebraic generaliza-
tions correctly on the basis of faulty responses that used either formula projection or
formula appearance match. For example, Dung, in two clinical interviews when he
was in sixth grade, primarily established and justified his generalizations figurally
and oftentimes with the use of a generic example. However, in two clinical inter-
views when he was in seventh grade, Dung primarily established his generalizations
W-Dot Sequence Problem. Consider the following sequence of W-patterns be-

low.
A. How many dots are there in pattern 6? Explain.

B. How many dots are there in pattern 37? Explain.
C. Find a direct formula for the total number of dots D in pattern n. Explain how
you obtained your answer. If you obtained your formula numerically, explain it
in terms of the pattern of dots above.
D. Zaccheus’s direct formula is as follows: D = 4(n + 1) − 3. Is his formula
correct? Why or why not? If his formula is correct, how might he be thinking
about it? Which formula is correct: your formula or his formula? Explain.
E. A certain W-pattern has 73 dots altogether. Which pattern number is it? Ex-
plain.
Fig. 16a W-dot pattern task
Fig. 16b Anna’s figural justification of the W-dot pattern in Fig. 16a
numerically and justified inconsistently using formula projection. An example of a

faulty argument that uses formula appearance match is exemplified in the thinking
of Anna who first developed the generalization D = n × 4 + 1 numerically for the
figural pattern in Fig. 16a. When she was then asked to justify her formula, she
circled 1 group of 4 circles, 2 groups of 4 circles, and three groups of 4 circles in
patterns 1, 2, and 3, respectively, beginning on the left and then referred to the last
circle as the y-intercept (Fig. 16b). As a matter of fact, in the post-interview in Year
2, only three of the eight students saw the sequence in Fig. 16a in the same way
Dung perceived it (Fig. 16c).
Discussion The phenomenological shift from the figural to numerical modes in es-
tablishing generalizations involving figural linear patterns among our middle school
Fig. 16c Dung’s figural justification of the pattern in Fig. 16a
students in the first two years of the study is not uncommon in empirical ac-
counts of cognitive development. Induction studies in developmental psychology
have demonstrated shifts in students’ abilities to categorize (from perceptual to con-
ceptual; from object- or attribute-oriented to relation-oriented, etc.). Also, Davydov
(1990) has noted similar occurrences of change on the basis of his work on gener-
alization with Soviet students, including his critique of mathematics instruction that
seems to favor one process over the other.
Based on the empirical data we collected in Years 1 and 2, the shift from the
figural to the numerical could be explained initially in terms of the predictive and
methodical nature of the established numerical strategy (as exemplified by Anna’s
group thinking relative to the Fig. 15 pattern). That is, the students found them
to be compact and easy to use particularly in far generalization tasks that asked
them to determine an output value for a large input value. What was difficult with
figural strategies, which could be dispensed with the established numerical strategy,
was the cognitive perceptual distancing that was necessary in order to: (1) figurally
apprehend and capture invariance in an algebraically useful manner; (2) selectively
attend to aspects of sameness and differences among figural stages and; (3) create
a figural schema or a mental image of a consistent generic case and then transform
the schema or image into symbolic form. In terms of Radford’s (2006) definition
of algebraic generalization of a pattern—grasping of a commonality, applying the
commonality to all the terms in the pattern, and providing a direct expression for the
pattern—the almost, albeit not fully, automatic process of numerical generalizing
requires only a surface grasp of a commonality (i.e., a common difference in the
case of a linear pattern) that would then be used to set up a direct expression. In
particular, when the students surfaced a commonality among stages in a numerical
generalizing process involving linear patterns, most of them did not even establish
it figurally since the corresponding numerical representation was sufficient for their
purpose.
In articulating our argument of a figural-to-numerical shift in mode of general-
izing in the first two years of our study, we have already noted how most of them
could correctly establish CSGs numerically but had difficulty justifying them. We
also discussed how some of them employed formula projection in an inconsistent
(faulty) manner. Another significant source of difficulty in justifying CSGs was the
students’ misconstrual of the multiplicative term in the general form y = mx + b for
linear patterns. Toward the end of the Year 1 teaching experiment, they would often-
times express their algebraic generalization in the form O = n × d + a, where the
variable O refers to the total number of objects being dealt with (like matchsticks,
circles, squares, etc.), n the pattern number, d the common difference, and a the
adjusted value. For example, the general form for the pattern sequence in Fig. 1 is
T = n × 3 + 1. The students would then justify their formula by locating n groups of
3 matchsticks respecting invariance along the way. In the Year 2 study, they learned
more about the commutative property, which then encouraged them to write all their
generalizations in the equivalent form O = dn + a. This became a source of confu-
sion among some of them because they interpreted the expressions n × d and d × n
as referring to the same grouping of objects. For example, in the clinical interviews
that we conducted immediately after the Year 2 teaching experiment, some of those
who wrote the form D = 4n + 1 for the sequence in Fig. 16a justified its validity by
looking for 4 groups of, say, 2 circles in pattern 2 when, in fact, they should have
been looking for 2 groups of 4 circles. Thus, the algebraic representation proved
to be especially confusing among those who established their generalizations nu-
merically because of their misinterpretations involving some of the mathematical
concepts and properties relevant to integers (such as the commutative law for mul-
tiplication).
The final shift in Year 3, from numerical to figural mode of generalizing, as a mat-
ter of fact, settled the above issues. Because the students understood the relationship
between multiplicative thinking and grouping relations, they reinterpreted their pat-
tern structural analysis in terms of how grouping could be accomplished so that it
is stable and consistent across stages. For example, in Fig. 3, Che noted the four
corner squares that stayed the same. She also saw stability in grouping the middle
parts on all four sides from stage to stage. Hence, her direct formula, W = 4n + 4,
captured her figural interpretation of the structure that she saw in Fig. 3.

Capability in Constructing and Justifying CNGs and DGs
Findings in Our Study Considering the results drawn from our longitudinal work
(and, in fact, relevant patterning studies discussed in this paper), we can conclude
with sufficient sample that the task of establishing and justifying CNGs and DGs
could be both easy and difficult for middle school students. For most students, com-
petence in pattern generalization that leads to a CNG and DG could be considered
as an effect of acquired knowledge and experience. We have found that individual
and classroom-generated practices on pattern generalization with minimal scaffold-
ing from the teacher, while helpful in many simple cases of linear patterns, appear
limited in many respects. In our three-year study, the first two years in which the stu-
dents were numerically driven to producing CSGs constrained them from obtaining
more complex and equivalent generalizations for the same pattern. The numerical
method of table differencing assisted in simplifying the process of constructing di-
rect formulas, however, it had a negative effect on the students’ ability to justify. We
note as well the limited form in which such formulas took shape at least in the case
of linear figural patterns, that is, they were oftentimes CSGs.
In Year 3 of the study, when the students acquired knowledge of multiplicative
thinking and found ways to link such thinking in patterning activity, the resulting
generalizations they produced and justified included CNGs and DGs. Results of
the Year 3 clinical interviews with fourteen students after the teaching experiment
on patterning and generalization show continued use of CSGs (100%), then DGs
(86%), and finally CNGs (36%). Dina and Dave in Fig. 4 constructed and justified
two equivalent CNGs for the Fig. 3 pattern. Figure 17 shows the generalizations of
five students on the T Stars Pattern that ranged in complexity from CSGs to CNGs
to DGs.
We discuss briefly the nonlinear pattern generalization of Diana, 7th grader from
Cohort 2, whose pattern of growing segment-triangles is shown in Fig. 19 on a free
construction task in Fig. 18 that was given after a teaching experiment on pattern
generalization. We should point out that Diana ignored the differences in lengths
of the diagonal and horizontal line segments, a fact that applies to a significant
number of students who saw segment length as unimportant on this task. Despite
that fact, Diana clearly identified an underlying structure in her growing pattern.
Hence, assuming all segments are equal, her stage 1 triangle consists of two diag-
onal segment-sides, a horizontal base that has two segments, and with no interior
segments. She then doubled each segment in stage 1 so that in stage 2, each diago-
nal side has two segments, the horizontal base has four segments, and two interior
horizontal segments. She then circled in two colors to distinguish the groupings she
was counting, one the interior horizontal segments, and the other, the outer segments
on the perimeter of the triangle. In her written description of what to her stayed the
same and what changed, she wrote:
Number 1 [the original triangle] will stay in all of them. The x(x − 1) is for the lines in the
middle of the triangle. The +4x is for the triangle borders. It’s really short for 2x + 2(x).
But it was pretty much the same.
To illustrate, Diana counted the interior horizontal segments of her growing triangle
pattern as: 1 group of 0 segment in stage 1; 2 groups of 1 segment in stage 2; 3
groups of 2 segments in stage 3; 4 groups of 3 segments in stage 4; 5 groups of
4 segments in stage 5 leading to the expression n(n − 1). Then she counted the
segments on the perimeter of the growing triangle in two parts. Part A pertains to
the two diagonal sides of the growing triangle: 2 groups of 1 segment in stage 1; 2
groups of 2 segments in stage 2; 2 groups of 3 segments in stage 3 leading to 2n.
Part B pertains to the base of the growing triangle: 2 groups of 1 segment in stage 1;
2 groups of 2 segments in stage 2; 2 groups of 3 segments in stage 3; 2 groups of 4
segments in stage 4 leading to 2n. Clearly, central to her pattern generalization was
her understanding of multiplicative thinking that enabled her to count in ways that
corresponded to how she was circling the parts of her figures. Finally, she simplified
her pattern of L = n(n − 1) + 4n to L = n2 + 3n.
Fig. 17 Year 3 students’ work on the T stars pattern in Fig. 8a

From the following three figures below, pick at least two figures to create a
pattern sequence of five stages. Use the attached grid paper to draw your four
additional stages.
1. What stays the same and what changes in your pattern?

2. Obtain a generalization for your pattern either by describing it in words or by
constructing a formula. How do you know that your generalization works?
Fig. 18 Semi-free construction task
Discussion Central to the students’ success in the Year 3 study was the sociocul-
tural mediation that took place in the context of activities that encouraged them to
explicitly engage in multiplicative thinking. When they began to see the significance
of multiplicative thinking on matters that involve grouping and invariance in rela-
tion to patterning activity, their pattern generalization further progressed in ways
that could not be simply done in the case of the numerically driven method of table
differencing. We should note, however, that the students interviewed by the end of
the Year 2 study were all successful in justifying given DGs. But their success was
task-sensitive with some of them providing correct justification in one task and then
an incorrect justification in some other task.
For example, results of the two clinical interviews in our Year 2 study separated
by a teaching experiment show that almost all the students had more difficulty deal-
ing with the Fig. 16a pattern than the Fig. 1 pattern. Results of the clinical interviews
prior to the teaching experiment show only one student correctly justifying a DG in
the case of the Fig. 16a pattern and six students in the case of the Fig. 1 pattern.
Further, all students interviewed after the teaching experiment were able to justify
the DG for the Fig. 1 pattern, but only six students in the case of the Fig. 16a pat-
tern. Thus, it seems that some overlaps in a deconstructive generalization task are
easier to see than others. For example, the students above found it easier to see over-
laps among the shared adjacent sides of the squares (Fig. 1) than the shared interior
vertices in a W-dot formation (Fig. 16a).
In a reported study by Steele and Johanning (2004), their middle school partic-
ipants found DGs difficult at least in the context of their teaching experiment. The
authors asked eight U.S. 7th graders to generalize five linear and three quadratic
problem situations that pertained to growth, change, size, and shape. Their results
show that, in the case of tasks that contained figural stages, only three students were
Fig. 19 Diana’s pattern generalization in relation to Fig. 18 task
able to establish and justify DGs (or “well-connected subtracting-out schemas”).

The notion of multiplicative thinking was not used in their teaching experiment.
In the Year 1 clinical interviews after the teaching experiment on pattern gen-
eralization, none of them were found to be capable in establishing and justifying
a DG. Further, in clinical interviews in Year 2 after the teaching experiment on pat-
tern generalization, none of them were capable of constructing DGs. However, there
was a marked gain in their ability to interpret and justify a stated DG (with a success
rate of 50% to 100% in pre- and post-clinical interviews, respectively). All the stu-
dents interviewed saw the overlapping sides in the adjacent squares pattern in Fig. 1
and six could see the overlapping interior vertices in the case of the W-dot pattern
in Fig. 16a. We further note that despite their success in justifying, seeing an over-
lap was not immediate for most of the students; it became evident only after they
had initially employed formula appearance match followed by formula projection.
Of course, some students employed formula projection incorrectly. For example,
Jana justified the subtractive term 3 in Zaccheus’s DG (item D in Fig. 16a) in the
following manner:
FDR: So if you look at this [referring to the formula (item D, Fig. 16a) in which
Jana substituted the value of 2 for n], this one’s four times two plus one,
right? And then minus 3. So how might he be looking at 4 times 2 plus 1 and
then minus 3?
Jana: Uhum, the 2 is for the pattern number.
FDR: Uhum. Because when Zaccheus was thinking about it, he said multiply 4 by
n + 1 and then take away 3. So how might he be thinking about it?
Jana: Like it’s gonna be 3 [referring to 2 + 1] and then it’s gonna be 12 [referring
to 4 × 3]. But I counted there’s only 9, so he has to subtract 3.
FDR: So how might he be doing that? Suppose I do this? [FDR builds pattern
2 with circle chips in which the three overlapping “interior” vertices are
colored differently.]
Jana: Hmm, like he has this group of 4 [Jana sees only two sides in W in pattern
2 with the top middle interior dot connecting the two sides. Hence, one side
has 4 dots.].
FDR: Is there a way to see these 4 groups of 3 here [referring to pattern 2]?
Jana: Like he imagines there’s 3 and he has to subtract 3.
FDR: So can you try it for other patterns? [Jana builds pattern 4.]
Jana: He has 1 group of 4. So there’s 3 groups of 4 and he imagines 3 more [to
form 4 groups of 4] and then he subtracts them [the three circles added].
FDR: So he imagines there’s three more. But why do you think he would add and
then take away?
Jana: Because there’s supposed to be 4 groups of 4 and then you don’t have enough
of these ones [circles] so he adds 3. You add these ones.
Conclusion
This paper began with two broad questions that have guided the longitudinal study
summarized in this work: What is the nature of the content and structure of gener-
alization involving figural patterns among middle school students? To what extent
are they capable of establishing and/or justifying more complicated generalizations?

Various patterning studies that have been conducted at the middle grades level pro-
vide strong evidence that students’ generalizations shift from the recursive to the
closed, constructive form. In this article, we discussed in some detail at least three
epistemological forms of generalization involving figural linear patterns, namely:
CSG, CNG, and DG. The general forms are further classified according to percep-
tual complexity. CSGs are the easiest for most middle school students to establish
and, thus, most prevalent. CNGs and DGs are relatively difficult and less preva-
lent. This classification scheme of generalizations emerged from detailed analyses
of students’ pattern generalization over three years. Also, it elucidates the content
and structure of such generalizations.
We have also discussed how students’ approaches to establishing generalizations
are intertwined with their justification schemes. Further, results drawn from our lon-
gitudinal work show shifts in pattern generalization schemes among middle school
students at least in the case of figural patterns. We note two consequences. First,
we highlight changes in their representational skills and fluency, that is, from be-
ing verbal (situated) to symbolic (formal) and to figural (formal). Second, the phe-
nomenological shifts affect the manner in which they justify their generalizations.
We have documented at least four types of justifications, namely: extension gen-
eration; generic example use; formula projection, and; formula appearance match.
The entry level of justification oftentimes involves generating extensions (i.e., cal-
culating and/or producing more stages after the initial ones). Students who then
generalize numerically without having a strong figural foundation are most likely to
employ formula appearance match and use formula projection inconsistently. Stu-
dents who understand multiplicative thinking in relation to figural patterning activity
oftentimes employ formula projection but the success and validity of such formulas
are relative to the associated structural analyses.
References
Becker, J. R., & Rivera, F. (2005). Generalization strategies of beginning high school algebra stu-
dents. In H. Chick & J. L. Vincent (Eds.), Proceedings of the 29th Conference of the Interna-
tional Group for the Psychology of Mathematics Education (Vol. 4, pp. 121–128). Melbourne,
Australia: University of Melbourne.
Becker, J. R., & Rivera, F. (2006). Establishing and justifying algebraic generalization at the sixth
grade level. In J. Novotná, H. Moraová, M. Krátká, & N. Stehlíková (Eds.), Proceedings of
(Vol. 4, pp. 465–472). Prague: Charles University.
Becker, J. R., & Rivera, F. (2007). Factors affecting seventh graders’ cognitive perceptions of pat-
terns involving constructive and deconstructive generalization. In J. Woo, K. Park, H. Sew,
& D. Seo (Eds.), Proceedings of the 31st Conference of the International Group for the Psy-
chology of Mathematics Education (Vol. 4, pp. 129–136). Seoul, Korea: The Korea Society of
Educational Studies in Mathematics.
Becker, J. R., & Rivera, F. (2008). Nature and content of generalization of 7th- and 8th-graders on
a task that involves free construction of patterns. In O. Figueras, J. L. Cortina, S. Alatorre, T.
Rojano, & A. Sepulveda (Eds.), Proceedings of the Joint Meeting of PME 32 and PMENA XXX
(Vol. 4, pp. 201–208). Morelia, Mexico: Cinvestav-UMSNH.
Bishop, J. (2000). Linear geometric number patterns: Middle school students’ strategies. Mathe-
matics Education Research Journal, 12(2), 107–126.
Davydov, V. (1990). Types of Generalization in Instruction: Logical and Psychological Problems
in the Structuring of School Curricula. Reston, VA: National Council of Teachers of Mathe-
matics (J. Teller, Trans.).
Dörfler, W. (1991). Forms and means of generalization in mathematics. In A. Bishop, S. Mellin-
Olsen, & J. van Dormolen (Eds.), Mathematical Knowledge: Its Growth Through Teaching
(pp. 63–85). Dordrecht, Netherlands: Kluwer Academic Publishers.
Dörfler, W. (2008). En route from patterns to algebra: Comments and reflections. ZDM, 40(1),
143–160.
Dretske, F. (1990). Seeing, believing, and knowing. In D. Osherson, S. M. Kosslyn, & J. Holler-
back (Eds.), Figural Cognition and Action: An Invitation to Cognitive Science (pp. 129–148).
Cambridge, Massachusetts: MIT Press.
Dreyfus, T. (1991). Advanced mathematical thinking process. In D. Tall (Ed.), Advanced Mathe-
matical Thinking (pp. 25–41). Dordrecht, Netherlands: Kluwer.
Duval, R. (1998). Geometry from a cognitive point of view. In C. Mammana & V. Villani (Eds.),
Perspectives in the Teaching of Geometry for the 21st Century (pp. 29–83). Boston: Kluwer.
Duval, R. (2006). A cognitive analysis of problems of comprehension in a learning of mathematics.
Educational Studies in Mathematics, 61(1&2), 103–131.
English, L., & Warren, E. (1998). Introducing the variable through pattern exploration. Mathemat-
ics Teacher, 91, 166–170.
Garcia-Cruz, J. A., & Martinón, A. (1997). Actions and invariant schemata in linear generalizing
problems. In Proceedings of the 23rd Conference of the International Group for the Psychology
of Mathematics Education (Vol. 4, pp. 161–168). Haifa: Israel Institute of Technology.
Gelman, R. (1993). A rational-constructivist account of early learning about numbers and objects.
In D. Medin (Ed.), Learning and Motivation (Vol. 30) New York: Academic Press.
Gelman, R., & Williams, E. (1998). Enabling constraints for cognitive development and learning:
Domain specificity and epigenesis. In W. Damon, D. Kuhn, & R. Siegler (Eds.), Handbook of
Child Psychology: Vol. 2. Cognition, Perception, and Language (5th ed., pp. 575–630). New
York: Wiley.
Gravemeijer, K. P. E., & Doorman, L. M. (1999). Context problems in realistic mathematics educa-
tion: A calculus course as an example. Educational Studies in Mathematics, 39(1–3), 111–129.
Hartnett, P., & Gelman, R. (1998). Early understandings of numbers: Paths or barriers to the con-
struction of new understandings? Learning and Instruction: The Journal of the European As-
sociation for Research in Learning and Instruction, 8(4), 341–374.
Hershkovitz, R. (1998). About reasoning in geometry. In C. Mammana & V. Villani (Eds.), Per-
spectives on the Teaching of Geometry for the 21st Century (pp. 29–37). Boston: Kluwer.
Iwasaki, H., & Yamaguchi, T. (1997). The cognitive and symbolic analysis of the generalization
process: The comparison of algebraic signs with geometric figures. In E. Pehkonnen (Ed.), Pro-
ceedings of the 21st Annual Conference of the Psychology of Mathematics Education (Vol. 3,
pp. 105–113). Lahti, Finland.
Knuth, E. (2002). Proof as a tool for learning mathematics. Mathematics Teacher, 95(7), 486–490.
Lannin, J., Barker, D., & Townsend, B. (2006). Recursive and explicit rules: How can we build
student algebraic understanding. Journal of Mathematical Behavior, 25, 299–317.
Lee, L. (1996). An initiation into algebra culture through generalization activities. In C. Bednarz,
C. Kieran, & L. Lee (Eds.), Approaches to Algebra: Perspectives for Research and Teaching
(pp. 87–106). Dordrecht, Netherlands: Kluwer Academic Publishers.
Lobato, J., Ellis, A., & Muñoz, R. (2003). How “focusing phenomena” in the instructional environ-
ment support individual students’ generalizations. Mathematical Thinking and Learning, 5(1),
1–36.
MacGregor, M., & Stacey, K. (1992). A comparison of pattern-based and equation-solving ap-
proaches to algebra. In B. Southwell, K. Owens, & B. Perry (Eds.), Proceedings of the Fifteenth
Annual Conference of the Mathematics Education Research Group of Austalasia (pp. 362–371).
Brisbane, Australia: MERGA.
Martino, A. M., & Maher, C. A. (1999). Teacher questioning to promote justification and general-
ization in mathematics: What research has taught us. Journal of Mathematical Behavior, 18(1),
53–78.
(Eds.), Approaches to Algebra: Perspectives for Research and Teaching (pp. 65–86). Dordrecht,
Netherlands: Kluwer Academic Publishers.
Mason, J., Graham, A., & Johnston-Wilder, S. (2005). Developing Thinking in Algebra. London:
The Open University.
Mathematics in Context Team (2006a). Building Formulas: Algebra. Chicago, IL: Encyclopaedia
Brittanica, Inc.
Mathematics in Context Team (2006b). Expressions and Formulas: Mathematics in Context.
Chicago, IL: Encyclopaedia Brittanica, Inc.
Metzger, W. (2006/1936). Laws of Seeing. Cambridge, MA: MIT Press.
Orton, A., & Orton, J. (1999). Pattern and the approach to algebra. In A. Orton (Ed.), Patterns in
Orton, J., Orton, A., & Roper, T. (1999). Pictorial and practical contexts and the perception of
pattern. In A. Orton (Ed.), Patterns in the Teaching and Learning of Mathematics (pp. 121–
136). London: Cassell.
ysis. Educational Studies in Mathematics, 42, 237–268.
Radford, L. (2001). Factual, contextual, and symbolic generalizations in algebra. In M. van den
Hueuvel-Panhuizen (Ed.), Proceedings of the 2th Conference of the International Group for the
Psychology of Mathematics Education (Vol. 4, pp. 81–88). Netherlands: Freudenthal Institute.
Radford, L. (2006). Algebraic thinking and the generalization of patterns: A semiotic perspec-
tive. In S. Alatorre, J. Cortina, M. Sáiz, & A. Méndez (Eds.), Proceedings of the 28th Annual
Meeting of the North American Chapter of the International Group for the Psychology of Math-
ematics Education (Vol. 1, pp. 2–21). México: UPN.
Radford, L. (2008). Iconicity and contraction: A semiotic investigation of forms of algebraic gen-
eralizations of patterns in different context. ZDM, 40(1), 83–96.
Richland, L., Holyoak, K., & Stigler, J. (2004). Analogy use in eight-grade mathematics class-
rooms. Cognition and Instruction, 22(1), 37–60.
Rivera, F. (2010a). Visual templates in pattern generalization activity. Educational Studies in Math-
ematics, 73(3), 297–328.
Rivera, F. (2010b). Toward a Visually-Oriented School Mathematics Curriculum: Research, The-
ory, Practice, and Issues. Mathematics Education Library, Vol. 49. New York, NY: Springer.
Rivera, F., & Becker, J. R. (2003). The effects of figural and numerical stages on the induction
processes of preservice elementary teachers. In N. Pateman, B. Dougherty, & J. Zilliox (Eds.),
Proceedings of the Joint Meeting PME and PMENA (Vol. 4, pp. 63–70). Honolulu, HI: Univer-
sity of Hawaii.
Rivera, F., & Becker, J. (2008). Sociocultural intimations on the development of generalization
among middle school learners: Results from a three-year study. In O. Figueras, J. L. Cortina, S.
Alatorre, T. Rojano, & A. Sepulveda (Eds.), Proceedings of the Joint Meeting of PME 32 and
PMENA XXX (Vol. 4, pp. 193–200). Morelia, Mexico: Cinvestav-UMSNH.
Rivera, F., & Becker, J. (2009a). Developing generalized reasoning in patterning activity. Mathe-
matics Teaching in the Middle School.
Rivera, F., & Becker, J. (2009b). Visual templates in pattern generalization. In M. Tzekaki, M.
Kaldrimidou, & H. Sakonidis (Eds.), Proceedings of the 33rd Conference of the International
Group for the Psychology of Mathematics Education (Vol. 4, pp. 473–480). Thessaloniki,
Greece: PME.
Sasman, M., Olivier, A., & Linchevski, L. (1999). Factors influencing students’ generalization
thinking processes. In O. Zaslavsky (Ed.), Proceedings of the 23rd Conference of the Interna-
tional Group for the Psychology of Mathematics Education (Vol. 4, pp. 161–168). Haifa: Israel
Institute of Technology.
Stacey, K. (1989). Finding and using patterns in linear generalizing problems. Educational Studies
Stacey, K., MacGregor M. (2001). Curriculum reform and approaches to algebra. In R. Sutherland,
T. Rojano, A. Bell, & R. Lins (Eds.), Perspectives on School Algebra (pp. 141–154). Dordrecht,
Netherlands: Kluwer Academic Publishers.
Steele, D., & Johanning, D. (2004). A schematic-theoretic view of problem solving and develop-
ment of algebraic thinking. Educational Studies in Mathematics, 57, 65–90.
Swafford, J., & Langrall, C. (2000). Grade 6 students’ preinstructional use of equations to describe
and represent problem situations. Journal for Research in Mathematics Education, 31(1), 89–
112.
Taplin, M. L., & Robertson, M. E. (1995). Spatial patterning: A pilot study of pattern formation
and generalization. In Proceedings of the 19th Conference of the International Group for the
Psychology of Mathematics Education (Vol. 3, pp. 273–280), Recife, Brazil.
Treffers, A. (1987). Three Dimensions: A Model of Goal and Theory Description in Mathematics
Education. Reidel: Dordrecht.
Commentary on Part II
Bharath Sriraman and Kyeong-Hwa Lee
Introductory Remarks
In this commentary to the nine chapters in the cognitive section of early algebraiza-
tion, we synthesize and critically discuss common themes found in them such as
components of non-formal algebraic thinking, the purported dichotomy between
arithmetic and algebra; meaningful arithmetic, and generalizing ability, among oth-
ers using the frameworks of William Brownell, Ernst Haeckl and Jean-Baptiste
Lamarck.
The nine chapters that comprise Part II of this book consist of 4 revised ZDM
articles and 5 new chapters, which together explore the cognitive aspects of early
algebraization. As spelled out in the preface of the first volume of the Advances in
mathematics education series, “the purpose of a commentary is not only to elucidate
ideas present in an original text, but. . . [t]o take them forward in ways not conceived
of originally” (Sriraman and Kaiser 2010, p. vi). Therefore, our aim in this commen-
tary is to synthesize common themes found in the nine chapters of this section and to
discuss the significance of the ideas and claims made in the chapters through differ-
ent theoretical lenses. We add that a commentary cannot occur in the void meaning
that it needs to be anchored in what is already existent in the literature. So for the
sake of better inter-textuality in the existent literature (Sriraman 2010) and preserv-
ing the integrity of what is already known in the field, we refer the reader to two
related books released by Springer in the MEL1 series, namely vol. 43 focused on
Educational Algebra (Filloy et al. 2008) and vol. 22 which explored Perspectives
1 Mathematics education Library Series, Springer.

B. Sriraman ()
Department of Mathematical Sciences, University of Montana, Missoula, USA
K.-H. Lee
Department of Mathematics Education, Seoul National University, Seoul, Korea

368 B. Sriraman and K.-H. Lee
on School Algebra (Sutherland et al. 2001) that contain complementary perspec-

tives to those found in this commentary and book. The critically reflective question
we asked ourselves is whether these nine chapters really represented advances in
the domain of early algebraization or whether they were simply a regurgitation or
recycling of old ideas in new clothing?
Early Algebraization Versus Meaningful Arithmetic
There is no rigid definition of early algebraization per se found in the literature,

which is acknowledged by the editors of the book (Cai and Knuth). Instead the term
is used to refer to algebraic thinking initiated from the early grades (elementary
school) via the use of student’s informal knowledge, different modes of representa-
tions, pattern activities that lead to generalization, and building on natural linguistic
and cognitive mechanisms by reflection, verbalization, articulation and sense mak-
ing (Greer 2008; Kaput 1999). The question that arises in our mind when reading
these chapters that address cognitive aspects of early algebraization is whether they
are really addressing the notion of “meaningful arithmetic” as proposed by William
Brownell (1895–1977)? Amongst mathematics educators there is consensus that the
traditional separation of arithmetic from algebra hinders the development of alge-
braic thinking in the later grades. Hence it makes sense to push for developing al-
gebraic ideas in the earlier grades through various activities such as pattern finding
and relationships in the curriculum without the formalism of notation or symbols,
which collectively is termed early algebraization. We do not dispute this, however
are these ideas really advancing our field, or are they a recycling of previous work?
This is the focus of our discussion in this section.
Brownell (1947) was the chief spokesperson for the “meaningful” arithmetic.
Meaningful arithmetic refers to instruction which is deliberately planned to teach
arithmetical meanings and to make arithmetic more sensible to children through its
mathematical relationships. Brownell categorized the meanings of arithmetic into
the following groups.
1. A group consisting of a large list of basic concepts. For example: meanings of
whole numbers, of fractions, ratios and proportions etc.
2. A second group consisting of arithmetical meanings which includes understand-
ing of fundamental operations. Children must know when to add, subtract, mul-
tiply, and divide. They must also know what happens to the numbers used when
a given operation is performed.
3. A third group of meanings consisting of principles that are more abstract. For ex-
ample: relationships and generalizations of arithmetic, like knowing that 0 serves
as an additive identity, the product of two abstract factors remains the same re-
gardless of which factor is use as a multiplier, etc.
4. A fourth group of meanings that relates to the understanding of the decimal num-
ber system, and its uses in rationalizing computational procedures and algorithms
(Brownell 1947).
Commentary on Part II 369
Meaningful arithmetic is “deliberately planned to teach arithmetical meanings

and to make arithmetic sensible to children through its mathematical relationships”.
Brownell emphasized that learning arithmetic through computations required con-
tinuous practice. He suggested that teaching meaningful arithmetic would reduce
practice time, encourage problem solving, and develop independence in students
and remarked that there was lack of research in 1947, on teaching and learning arith-
metic meaningfully. He also doubted that quantitative research could address ques-
tions in the area of teaching and Learning arithmetic. One could say that Brownell
was clairvoyant for his time, emphasizing qualitative research in the age of behav-
iorism and he was noted for his use of a variety of techniques for gathering data,
including extended interviews with individual children and teachers, as well as his
careful, extensive and penetrating analyses of those data (Kilpatrick 1992).
Sixty three years after Brownell’s recommendations, we see ample research ev-
idence in the nine chapters of this part that the algebra underpinning the arithmetic
operations can be made accessible to students without the need for sophisticated
notation that impedes understanding. Two chapters report on the results of longi-
tudinal research projects. For instance, the chapter by Britt and Irwin on Algebraic
thinking with and without algebraic representation: a pathway to learning reports
on substantial gains in introducing algebraic thinking within arithmetic in the New
Zealand Numeracy Project with 4–7 year olds, with the gains remaining in a fol-
low up large scale study with the same students at the age of 11–12 on a 21-item
test consisting of various algebraic properties. Similarly Cai, Moyer, Wang and Nie
report the findings of the LieCal2 Project on the algebraic development of middle
school students (grades 6–8) exposed to reform based curricula (CMP) versus a tra-
ditional curricula (non CMP). They found that CMP students performed better on
generalization tasks than their peers in the traditional (non CMP) track. The CMP
curriculum used a functional approach and emphasized conceptual understanding
whereas the non CMP curriculum took a structural approach and emphasized proce-
dural understanding. Interestingly enough in this rather massive study, both groups
performed equally well on equation solving!
Generalized Arithmetic, Generalizing, Generalization
Amongst mathematicians, algebra is typically understood as generalized arithmetic

and some would qualify this characterization by saying it is meaningful general-
ized arithmetic. In other words there is a significant shift (abstraction) from talk-
ing about basic arithmetic operations concretely, to talking about operations arbi-
trarily. There is an even greater shift (abstraction) when one views numbers as a
special case of polynomial evaluations. However mathematics educators would ob-
ject to such characterizations as being top-down, i.e., viewing particulars through
general lenses as opposed to discovering the general via particulars (Mason 1992;
2 Longitudinal Investigation of the Effect of Curriculum on Algebra Learning.

Sriraman 2004) and argue that algebraic thinking which fosters the elements of ab-
straction and generalization inductively from the early grades through contextual
and rich mathematical activities is more valuable than the deductive approach.
The theme of generalization and generalizing ability recurs in the chapter by
Cooper and Warren (Years 2 to 6 students’ ability to generalize: models, repre-
sentations and theory for teaching/learning), and the chapter by Rivera and Becker
(Formation of pattern generalization involving linear figural patterns among middle
school students: Results of a three-year study). The former chapter by Cooper and
Warren is the only one in this entire part to cite the work of Dienes (1961) and push
his ideas of multiple embodiments fostering abstraction and generalization, by ex-
ploring the interrelations between generalization and verbal/visual comprehension
of context, and arguing for the value of communicating commonalties seen across
different representations. The latter chapter delineates the different nuances of gen-
eralization such as constructive versus deconstructive generalizations, and the role
of sociocultural mediation in fostering/facilitating verbalization of generalizations
in teaching experiments. The different patterning activities found in these chapters
serve well to validate the claims made by these authors.
Many of the chapters in the cognitive part try to move beyond or work their way
around the debate that algebra is generalized arithmetic. However Radford’s chapter
on Grade 2 students’ non-symbolic algebraic thinking tackles this issue head on by
trying to draw a clear distinction between what is arithmetic and what is algebraic.
This is accomplished by focusing on non-symbolic means of expression, going back
to the ideas of other ways (e.g., linguistic, gestural, figural) of expressing algebraic
notions as opposed to expressing them symbolically. This chapter brings to the fore-
ground the need to have a historical perspective on the development of mathemati-
cal ideas (algebraic or otherwise). Diophantus’ Arithmetica is a misnomer and deals
with solving algebraic equations with integer co-efficients. Does that mean alge-
bra as we define it today preceded arithmetic or that they developed concurrently,
just as the notion of Integrals preceded the rigorous development of convergence of
sequences and limits but the approximation of areas and volumes contained the in-
tuitive notions that were later formalized? We leave these questions for the reader to
ponder over. But we foray briefly into the role of history of mathematics in mathe-
matics education before we conclude our commentary. In doing so, we highlight the
need to change the dominant discourses present in arithmetic-algebra dichotomy,
and adopt a historical-cultural perspective in addition to a strictly cognitive perspec-
tive. The commentary offered by Balacheff (2001) on Perspectives in school algebra
offers another perspective on the didactical dilemma/distinction between symbolic
arithmetic and algebra.
From Haeckel to Lamarck to Early Algebraization

The emphasis on the important role of the history of mathematics in mathemat-
ics education research is one that has been sporadically addressed by the com-
munity. Furinghetti and Radford (2002) traced the evolution of Haeckel’s (1874)
law of recapitulation from the point of view that parallelism is inherent in how
mathematical ideas evolve and the cognitive growth of an individual (Piaget and
Garcia 1989). In other words the difficulties or reactions of those who encounter
a mathematical problem can invariably be traced to the historical difficulties dur-
ing the development of the underlying mathematical concepts. The final theoret-
ical product (namely the mathematical theorem or object), the result of the his-
torical interplay between phylogenetic and ontogenetic developments of mathe-
matics, where phylogeny is recapitulated by ontogeny, has an important role in
pedagogical considerations (Bagni et al. 2004; Furinghetti and Radford 2002;
Sriraman and Törner 2008). Psychological constructs as well as the study and
formation of intellectual mechanisms are not as tenable as the clearly dated and
archived transformations of mathematics in its historical development. Further, the
apparent free use of Haeckel’s recapitulation theory as the link between the psycho-
logical and historical domains is in need of re-examination. It is a well known fact
that Haeckel’s law in its original form was rejected by the community of biologists
and has been transformed numerous times by some, over the last 100 years to bet-
ter explain the relationship between phylogeny and ontogeny in different species.
However in mathematics education we are referring to psychological recapitula-
tion or the use of recapitulation metaphors to explain the evolution of mathematical
ideas. A neo-Lamarckian perspective needs to be introduced into the recapitulation
discussion for the following reasons. Recapitulation cannot be applied or transposed
directly to the study of didactical problems because it does not take into account the
influence of experience (or more broadly culture). Just as Jean-Baptiste Lamarck
proposed in vain to his peers in 1803, that hereditary characteristics may be influ-
enced by culture, we need to take into account how culture influences the mutation
of historical ideas. Gould (1979) wrote that
Cultural evolution has progressed at rates that Darwinian processes cannot begin to ap-
proach. . . [t]his crux in the Earth’s history has been reached because Lamarckian processes
have finally been unleashed upon it. Human cultural evolution, in strong opposition to our
biological history, is Lamarckian in character. What we learn in one generation, we trans-
mit directly by teaching and writing. Acquired characters are inherited in technology and
culture. Lamarckian evolution is rapid and accumulative. . .
The teaching and learning of mathematics bears strong evidence to this Lamar-
ckian nature. Indeed, what took Fermat, Leibniz and Newton a 100 years is taught
and often digested by students in one year of university Calculus. Any higher level
mathematics textbook is a cultural artifact which testifies to rapid accumulation and
transmission of hundreds (if not thousands of years) of knowledge development.
So, evolutionary epistemologists have now begun to accept the fact that for humans,
cultural evolution in a manner of speaking is neo-Lamarckian (Callebaut 1987;
Gould 1979).
The neo-Lamarckian perspective becomes evident in the research reported by
Izsak in his chapter on Representational competence and algebraic modeling. In
this chapter, the reader confronts students with a significant and complex “substrata
of knowledge” and criteria developed from the culture of learning they were previ-
ously exposed to that resulted in them inventing their own private inscriptions and
criteria for evaluating representations. Developing notational competence in stu-

dents, i.e., the ability to adapt their inscriptions to fit the external representations
they are confronted with, and progress to the necessity for uniformity in notation
for the purposes of communication is an inference we draw from this chapter. The
history of algebra as seen in the development of the theory of equations shows that
the notation (or private inscriptions) developed by Galois on general methods for
the solvability of equations by radicals, took his peers a significant amount of time
to decipher, nearly 15 years after his death (until 1846), when Louiville published
it in his journal commenting on Galois’ solution, “. . . as correct as it is deep of this
lovely problem: Given an irreducible equation of prime degree, decide whether or
not it is soluble by radicals”.
In the chapter by Ellis on Algebra in the Middle School, the informal or cultural
notion of comparing quantities is used to scaffold the building of quantitative rela-
tionships and the sophisticated idea of covariation in younger students. In spite of
creating relevant, contextual and quantitatively rich situations, Ellis reports that the
“students unique interactions with an interpretations of real world situations remind
us that these contexts are not a panacea. Introducing a quantitatively rich situation
does not guarantee that students build quantitative relationships. . . [s]tudents may
focus on any number of features in a problem situation. . . [t]herefore teachers play
an important role in shaping a classroom discussion, . . .”
Knuth et al. in their reprint of the 2005 ZDM article suggest that students “pre-
algebraic” experiences are crucial in laying the foundation for the study of more
formal algebra. These authors view the middle school grades as the link from early
algebraic reasoning to more complex and abstract reasoning. Five years later, the
studies from New Zealand, Australia, Canada and the U.S. reported in this part,
indicate that algebraic thinking can be cultivated from the very early grades on if
teachers are cognizant of non-symbolic modes of reasoning. It is a testament to our
development as a field that the seemingly divergent ideas of Brownell, Haeckel and
Lamarck converge in our understanding of early algebraization, which can no longer
be viewed as a neologism but a clearly defined term!
References
Bagni, G. T., Furinghetti, F., & Spagnolo, F. (2004). History and epistemology in mathematics
education. In L. Cannizzaro, A. Fiori, & O. Robutti (Eds.), Italian Research in Mathematics
Education 2000–2003 (pp. 170–192). Milano: Ghisetti e Corvi.
Balacheff, N. (2001). Symbolic arithmetic vs algebra: The core of a didactical dilemma. In R.
Sutherland, T. Rojano, A. Bell, & R. Lins (Eds.), Perspectives on School Algebra (pp. 249–
260). Dordrecht, The Netherlands: Kluwer Academic.
Brownell, W. A. (1947). The place and meaning in the teaching of arithmetic. The Elementary
School Journal, 4, 256–265.
Callebaut, W. (1987). Why it makes sense to extend the genotyoe/phenotype distinction to culture.
Problemi epistemology della biologia, 2, 65. La Nuova Critica, Nuova Serie I-II.
Dienes, Z. P. (1961). On abstraction and generalization. Harvard Educational Review, 31, 281–
301.
Filloy, E., Rojano, T., & Puig, L. (2008). Educational Algebra. A Theoretical and Empirical Ap-
proach. Berlin: Springer.
Furinghetti, F., & Radford, L. (2002). Historical conceptual developments and the teaching of
mathematics: From phylogenesis and ontogenesis theory to classroom practice. In L. English
(Ed.), Handbook of International Research in Mathematics Education (pp. 631–654). Hillsdale:
Erlbaum.
Gould, S. J. (1979). Another look at Lamarck. New Scientist, 4, 38–40.
Greer, B. (2008). Guest editorial: Reaction to the final report of the national mathematics advisory
panel. The Montana Mathematics Enthusiast, 5(2&3), 365–370.
Haeckel, E. (1874). Anthropogenie oder Entwickelungsgeschichte des Menschen. Leipzig: Engel-
mann.
Kaput, J. J. (1999). Teaching and learning a new algebra. In E. Fennema & T. A. Romberg (Eds.),
Mathematics Classrooms that Promote Understanding (pp. 133–155). Mahwah, NJ: Lawrence
Erlbaum Associates.
Kilpatrick, J. (1992). A history of research in mathematics education. In D. A. Grouws (Ed.),
Handbook of Research on Mathematics Teaching and Learning (pp. 3–38). New York: NCTM:
Macmillan Publishing Company.
Mason, J. (1992). The Discipline of Noticing. Sunrise Research Laboratory, Melbourne: RMIT.
Piaget, J., & Garcia, R. (1989). Psychogenesis and the History of Science. New York: Columbia
University Press.
Sriraman, B. (2004). Reflective abstraction, uniframes and the formulation of generalizations. The
Journal of Mathematical Behavior, 23(2), 205–222.
Sriraman, B. (2010). Commentary on theorizing in mathematics education research: Differences
in modes and quality. Nordic Studies in Mathematics Education, 15(1), 53–58.
Sriraman, B., & Kaiser, G. (2010). Series preface to advances in mathematics education. In B.
Sriraman & L. English (Eds.), Theories of Mathematics Education: Seeking New Frontiers
Berlin: Springer.
Sriraman, B., & Törner, G. (2008). Political union/mathematical education disunion: Building
bridges in European didactic traditions. In L. English, M. Bussi, G. A. Jones, R. Lesh, B. Srira-
man, & D. Tirosh (Eds.), The Handbook of International Research in Mathematics Education
(2nd ed.) (pp. 656–690). New York: Routledge, Taylor & Francis.
Sutherland, R., Rojano, T., Bell, A., & Lins, R. (Eds.) (2001). Perspectives on School Algebra.
Dordrecht, The Netherlands: Kluwer Academic Publishers.
Part III
Instructional Perspective
Preface to Part III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

Eric Knuth
Jinfa Cai
Newark, USA
Prospective Middle-School Mathematics Teachers’ Knowledge of

Equations and Inequalities . . . . . . . . . . . . . . . . . . . . . . . . 379
Department of Mathematics, Illinois State University, Normal, USA
The Algebraic Nature of Fractions: Developing Relational Thinking in

Elementary School . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Susan B. Empson
College of Education, University of Texas at Austin, Austin, USA
Linda Levi
Cognitively Guided Instruction (CGI) Professional Development
Initiatives, Teachers Development Group, Madison, USA
Thomas P. Carpenter
Department of Curriculum and Instruction, University of
Wisconsin-Madison, Madison, USA
Professional Development to Support Students’ Algebraic Reasoning:

An Example from the Problem-Solving Cycle Model . . . . . . . . . 429
Karen Koellner
School of Education, Hunter College, New York, NY, USA
Jennifer Jacobs
School of Education, University of Colorado at Boulder, Boulder, USA
376 Instructional Perspective
Hilda Borko
School of Education, Stanford University, Stanford, USA
Sarah Roberts
Department of Curriculum and Instruction, Iowa State University,
Ames, USA
Craig Schneider
Using Habermas’ Theory of Rationality to Gain Insight into Students’

Understanding of Algebraic Language . . . . . . . . . . . . . . . . . 453
Department of Mathematics, University of Genova, Genova, Italy
Theoretical Issues and Educational Strategies for Encouraging Teachers

to Promote a Linguistic and Metacognitive Approach to Early
Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Mathematics Department, University of Modena and Reggio Emilia,
Modena, Italy
A Procedural Focus and a Relationship Focus to Algebra: How U.S.

Teachers and Japanese Teachers Treat Systems of Equations . . . . . 511
Margaret Smith
Department of Mathematics, Iona College, New Rochelle, USA
Teaching Algebraic Equations with Variation in Chinese Classroom . . . 529

Jing Li
School of Mathematics and Statistics, Southwest University, Chongqing,
40071, China
School of Mathematics and Information, Langfang Teachers’ College,
Langfang, 065000, China
Aihui Peng
Institute of Higher Education, Southwest University, Chongqing,
400715, China
Naiqing Song
School of Mathematics and Statistics, Southwest University, Chongqing,
40071, China
Commentary on Part III . . . . . . . . . . . . . . . . . . . . . . . . . . . 557

John Mason
Open University, Milton Keynes, UK
University of Oxford, Oxford, UK
Preface to Part III
Although curricula can provide elementary and middle school students with oppor-
tunities to develop their algebraic thinking, teachers are arguably the most important
influence on what students actually learn. Thus, the success of efforts to develop
students’ algebraic thinking prior to formal algebra courses rests largely with the
ability of teachers to foster the development of such thinking. The chapters in this
part present a diverse set of perspectives regarding early algebra instruction and
ways of supporting teachers’ efforts to foster the development of students’ algebraic
thinking. The chapters range in focus from teachers’ algebra content knowledge to
teachers’ use of classroom opportunities to foster algebraic thinking to international
comparisons of instructional approaches with regard to algebra.
Ellerton and Clements present data from a study that examined teacher educa-
tion students’ knowledge about equations and inequalities. Although their results
are disheartening, especially given that the teachers were seeking endorsement to
become middle school mathematics specialists, the authors do offer suggestions
gleaned from their work regarding the teaching and learning of algebra. Also, focus-
ing on teacher education albeit with an experienced teacher rather than pre-service
teachers, Koellner and her colleagues document the influence of a particular profes-
sional development model (Problem-Solving Cycle model) on one teacher’s algebra
instruction. Two years worth of data provide the basis for their description and il-
lustration of changes in the teacher’s instructional practice, and also serve to link
the changes to emphases in the professional development program. In the chap-
ter by Cusi and colleagues, they focus on the importance of enculturation of early
algebra teachers, emphasizing the critical role that teacher reflection on the pro-
cesses of teaching and learning play in their enculturation. The authors also discuss
E. Knuth ()
J. Cai

378 E. Knuth and J. Cai
the importance of viewing early algebra from both linguistic and socio-constructive
perspectives, and how such perspectives can reshape the nature of teachers’ encul-
turation. Finally, Smith’s chapter presents a comparison of instructional approaches
taken by teachers as they teach a lesson about simultaneous equations. In particular,
she details the qualitative differences in the instruction between U.S. and Japanese
teachers, contrasting the procedural-driven approach of U.S. teachers versus the
relationship-driven approach of Japanese teachers.
The final three chapters in this part focus on ways in which teachers can build
upon students’ thinking in ways that can promote the development of their alge-
braic thinking. For example, Empson, Levi, and Carpenter suggest that students’
work with fractions can also serve as an opportunity for teachers to enhance their
students’ understanding of algebraic structure. They argue that many of the strate-
gies used by students for fraction problems are often based on the same mathemat-
ical relationships that are essential in formal algebra. In the chapter by Jing and
colleagues, they describe a teaching approach commonly used in China for helping
students learn to represent and operate of algebraic equations. Their illustrations
also underscore the important role that whole class discussions play in facilitating
students’ learning. Finally, Moselli and Boero’s chapter present a model for un-
derstanding students’ difficulties in algebra that is based on Habermas’ construct
of rational behavior. The authors apply their model to students’ uses of algebraic
language in the context of solving mathematical modeling problems, demonstrating
how the model provides insight into students’ algebraic thinking as well as guidance
for teachers to understand their students’ algebraic thinking.
The seven chapters that comprise this part highlight the important role that teach-
ers play in developing students’ algebraic thinking—the need for teachers to possess
an adequate understanding of algebra themselves and the need for teachers to be able
to recognize and capitalize on classroom opportunities to foster students’ algebraic
thinking. The authors have provided not only insight into both the challenges and
opportunities with which teachers are presented, but also valuable suggestions for
supporting teachers as well as for continued research.
Prospective Middle-School Mathematics
Teachers’ Knowledge of Equations
and Inequalities
Abstract This chapter describes an investigation into the algebra content knowl-
edge, in relation to elementary equations and inequalities, of 328 US teacher-
education students who were seeking endorsement to become specialist middle-
school mathematics teachers. Most of these prospective teachers had done well in
high school mathematics and were taking their last algebra course before becoming
fully qualified teachers of mathematics. After reviewing the scant literature on the
teaching and learning of quadratic equations, and of linear inequalities, we sum-
marize a pencil-and-paper instrument, developed specifically for the study, which
included linear and non-linear equations and inequalities. The students were also
asked to comment, in writing, on a “quadratic equation scenario” that featured four
common errors in relation to quadratic equations. Data analysis revealed that hardly
any of the 328 students knew as much about elementary equations or inequalities
as might reasonably have been expected. Brief details of a successful intervention
program aimed at improving the pre-service teachers’ knowledge, skills and con-
cepts relating to quadratic equations and inequalities are given, and implications of
the findings for mathematics teacher education and, more generally, for the teaching
and learning of algebra, are discussed.
The Context
It has been well established by research that, in many nations, college students who
are preparing to become specialist teachers of mathematics at the middle-school
level often have unsatisfactory knowledge with respect to the algebra content that
N.F. Ellerton () · M.A. Clements

Department of Mathematics, Illinois State University, Normal, USA
M.A. Clements

380 N.F. Ellerton and M.A. Clements
they might be expected to teach (see, e.g., de Castro 2004). There is some evidence
that in the United States of America the problem is particularly acute. Schmidt
et al. (2007), for example, reported a large study which showed that US prospec-
tive middle-school teachers’ knowledge about “functions”—an important theme in
middle-school and high school mathematics—was low when compared with the
knowledge of corresponding cohorts in Taiwan, South Korea, Bulgaria, and Ger-
many. Indeed, Schmidt et al. claimed, the US performance “lagged almost three-
fourths of a standard deviation below the international mean” (p. 1).
Over the past five years we (Ellerton and Clements) have studied the algebra
content knowledge and attitudes toward algebra of mathematics education students
enrolled in an “Algebra for Teachers” second-year course (hereafter denoted AT2) at
a large US university. We have gathered and analyzed data on what mathematics the
prospective middle-school teachers had studied at school and at college, and have
analyzed, quantitatively and qualitatively, their responses to middle-school algebra
tasks.
Kieran (2007) emphasized that unless students “come to realize that algebra is an
arena of sense-making and that they can arrive at rules that will permit them to obtain
the same results as their teacher or classmates, they will never be able to control
their algebraic work” (p. 732). In this chapter we show how carefully developed
tasks were used to identify prospective middle-school teachers of mathematics who
were struggling with elementary equations and inequalities. We also outline what
we did to improve the situation.
Mathematical Considerations Relating to the Teaching and

Learning of Equations and Inequalities
This chapter describes an investigation into 328 prospective teachers’ thinking about
equations and inequalities, especially (but not only) quadratic equations and linear
and quadratic inequalities. In order to frame our subsequent discussion it will be
useful to begin with an examination of what it means to “solve” an equation or an in-
equality. Henry Pollak, a distinguished applied mathematician with a demonstrated
interest in mathematics education, stated, during an interview with Alexander Karp
(2007), that during the late 1950s, as a consultant to one of the School Mathematics
Study Group’s (SMSG’s) curriculum development teams, he found himself con-
fronted with the fundamental question of what it meant to solve an equation or an
inequality. Pollak stated that SMSG team members were surprised when they re-
alized that there did not seem to be any universally accepted agreement on such a
basic matter.
According to Pollak (see Karp 2007), the team that worked on SMSG’s ninth-
grade Algebra program agreed to adopt the definition that “to solve” an “open sen-
tence” involving a variable (usually an equation, or an inequality) is to find all ac-
ceptable values of the variable that will make the sentence true. The word “accept-
able” was important, because if x were constrained to being a real number then some
Prospective Mathematics Teachers’ Knowledge of Equations and Inequalities 381
apparently simple equations (like, for example, x 2 + 4 = 0) would have no solution.

However, the open sentence x 2 + 4 > 0 would have infinitely many real-number
solutions. One might say that if the replacement set for x is the set of real numbers
(hereafter denoted R), then the truth set for x 2 + 4 = 0 would be ∅, the null set.
On the other hand, with the same constraints operating, the truth set for x 2 + 4 > 0
would be R.
The SMSG team viewed equations and inequalities as part of the same conceptual
framework. As team members, Allen et al. (1965), wrote:
We write open [original emphasis] sentences which involve variables and for which the no-
tion of a truth set becomes important. It is essential that the student consider both equations
and inequalities as sentences, as objects of algebra with equal right to our attention. (p. 44)
The SMSG approach brought to the forefront what it meant to “solve” an equation
or inequality. It emphasized the need to state the replacement set for the variable. It
did not directly use the language of functions, but what the team members decided
could easily have been expressed in function terminology if the team had deemed
that necessary. We believe that the SMSG approach still provides a cohesive frame-
work within which educational issues associated with the teaching and learning of
equations and inequalities can be meaningfully discussed.
Before proceeding further, a brief note on the word “inequality” will be in order.
In the 1960s—the high point of the New Mathematics era—some authors (e.g.,
Clements et al. 1967, p. 88) preferred the term “inequation” to “inequality.” The
argument in favor of the former was that one might refer to “3 + 2 = 5” as an
equality, and to 3x + 5 = 0 as an “equation”; and, if this were done, it would seem
reasonable to regard a statement such as “5 > 3 + 1” as an inequality, and “5 > 3x +
1” as an inequation. Although we find the logic supporting the preference for the
term “inequation” in relation to an open sentence like 5 > 3x + 1 to be compelling,
in this paper we bow to common curriculum usage in the United States (and in most
other parts of the world), and call a sentence like 5 > 3x + 1 an inequality.
The New Math(s) period gave way to a back-to-the-basics era, and the truth set
approach developed by SMSG was forgotten (or, at best, put into the background).
Since then, there has been a large amount of research on conceptions and mis-
conceptions associated with what, and how, students think when they attempt to
solve linear equations. There has been relatively little research, however, on student
conceptions and misconceptions in relation to quadratic equations, or to linear and
quadratic inequalities (Kieran 2007), and we focus on these aspects in this chapter.
Some writers (e.g., Tsamir et al. 1998) have argued that it is theoretically de-
sirable to insist that equations and inequalities should be explicitly taught, from the
outset, in the context of the development of the function concept. Boero and Bazzini
(2004) have maintained, for example, that teaching inequalities without taking due
account of the concept of function “implies a ‘trivialisation’ of the subject, resulting
in a sequence of routine procedures which are not easy for students to understand,
interpret and control” (p. 140). They added that “as a consequence of this approach,
students are unable to manage inequalities which do not fit the learned schemas”
(p. 140). Such an argument tends to be bolstered by references to the need for a
stronger theoretical framing of algebra education relating to equations and inequal-

ities. Often it is suggested that any such framing should emphasize the concept of
function. The power and availability of modern technology is also likely to feature
in arguments put forward by proponents of a function approach.
Like Pimm (1995), we do not accept the view that any decent theory relating to
the role of equations and inequalities in school mathematics curricula must be linked
from the outset to the function concept. For example, we see nothing wrong with
fifth- or sixth-graders considering which natural numbers could replace in the
open sentences 2 × + 3 = 15 and 3 + 2 × < 15 so that true statements would
be obtained. Such tasks would be both age- and curriculum-appropriate, and other
age- and curriculum-appropriate open sentences could be devised for various groups
of students at different grade levels, or of different ages. Furthermore, if the SMSG
set-theoretical position for equations and inequalities is accepted, then those teach-
ing elementary algebra courses to prospective middle-school teachers will need not
only to be aware of, but also to understand thoroughly, that framework for equations
and inequalities. From that perspective, the sign-chart method for inequalities sum-
marized by Dobbs and Peterson (1991) could be valuable, provided it is carefully
and meaningfully tied to the truth value approach.
Research by Vaiyavutjamai (2004) and by Blanco and Garrote (2007) has sug-
gested that a question like “Which real numbers could replace x in the open sentence
3x − 4x 2 = 0, so that a true statement would be obtained?” is best approached, at
least initially, by the SMSG approach. With such an approach, 3x − 4x 2 would be
factored and the null factor law applied to find the truth set. From our perspective,
and our experience, that method can lead to meaningful consideration of quadratic
inequalities like 3x − 4x 2 > 0.
In Karp’s (2007) interview with Henry Pollak, Pollak made the following
thought-provoking comment on relationships between the concepts of open sen-
tence, equation and inequality:
There are different kinds of sentences and you want to know the sets of numbers that makes
these sentences true. And that is how equations and inequalities are all going to be done
together. Each of these is an open [original emphasis] sentence because you have left open
what x is. Each of these open sentences has a truth set, a set of x’s that makes the sentence
true. Maybe it’s empty, that’s possible, or it may be all x’s, or it may be some particular
collection, some set of x’s. But you leave open, what it is you’re trying to find. Now, what
does it mean to solve an equation? Well, what you do is you find that there are a number
of operations which do not change the truth set. You can add anything you want, you can
multiply, but not by zero, for equations at least. So, solving an equation means carrying out
operations which don’t change the truth set. For how long? Until the truth set is obvious.
That is, when you finally get to where it says x = 3 or x > 5, then you know that’s the truth
set. And since you have never changed the truth set as you went through, that’s where it was
to begin with. (p. 73)
Adopting an open-sentence approach with equations and inequalities need not in-
terfere with the development of function concepts in learners, for such an approach
can easily be linked to finding zeros of functions, or to reasoning about inequali-
ties from graphs of functions (or relations). We believe that students should learn
what it means to solve equations, and inequalities, before being given a formal in-
troduction to the concept of function. Thus, for example, a ninth-grade student can
learn to solve the quadratic equation 3x − 4x 2 = 0 with a degree of understanding

by factoring, using the null factor law, and thinking in terms of truth values for the
open sentence. By contrast, most ninth-grade students would find it difficult to know
what was going on in relation to the equation 3x − 4x 2 = 0 by merely examining
that part of the graph of the function f : R → R, f (x) = 3x − 4x 2 shown on a
graphing calculator. If a function approach is adopted too early then confusion can
abound: for example, although the equation 3x − 4x 2 = 0 is equivalent to (i.e., has
the same truth set as) 4x 2 − 3x = 0, the functions f : R → R, f (x) = 3x − 4x 2 and
g : R → R, g(x) = 4x 2 − 3x are certainly not the same.
In the study described in this chapter, teacher-education students who would soon
become accredited middle-school mathematics teachers, were asked to state, for
each of a given set of equations and inequalities, “all the real number(s) which could
replace x to make the statement true.” Although students seemed to recognize that
they were being asked to “solve” the equations and inequalities, analysis revealed
that some did not comprehend the expression “real number(s),” and most did not
know how to find the solutions.
Student Misconceptions in Regard to Quadratic Equations
Although this chapter focuses on prospective middle-school teachers’ conceptions

and misconceptions with respect to equations and inequalities, it will be helpful here
to summarize the findings of Vaiyavutjamai and Clements (2006a) with respect to
how 231 ninth-grade students in Thailand responded to quadratic equations before
and after a sequence of 10 lessons specifically dealing with quadratic equations.
That investigation is one of the few studies in the literature dealing with the ef-
fects of instruction on student skills and conceptions relating to quadratic equations.
Vaiyavutjamai and Clements (2006a), employing both quantitative and qualitative
analyses, found that after the sequence of lessons most students still did not give
both solutions to an equation like x 2 = 16 (when x could take any real-number
value). Most students also struggled, conceptually, with equations expressed in the
form (x − a)(x − b) = 0, where a, b represent any constant real numbers. Thus,
for example, although at the post-instruction stage, most interviewees obtained 3
and 5 as solutions to the equation (x − 3)(x − 5) = 0, when they were asked to
check their solutions they substituted x = 3 into (x − 3) and x = 5 into (x − 5)
and concluded that since 0 × 0 = 0 their solutions were correct. Students who did
this thought that the two x’s in (x − 3)(x − 5) = 0 represented different variables
and needed to take different values. These students obtained correct solutions, and
would have had their written solution scripts assessed as correct. Yet, as Kieran
(2007) noted, their responses “suggested the presence of serious gaps in the theo-
retical linking underpinning students’ work when solving such equations” (p. 732).
This same serious misconception was found among teacher-education students in
the United States who had chosen to seek endorsement as specialist middle-school
mathematics teachers (and who had all done well in high school algebra) (Clements
and Ellerton 2006, 2009; Vaiyavutjamai et al. 2005).
A different misconception was identified with respect to equations like

x 2 − x = 12. In Vaiyavutjamai and Clements (2006a), a majority of interviewees
correctly rearranged x 2 − x = 12 to x 2 − x − 12 = 0, and then to (x − 4)(x + 3) = 0.
They then equated both x − 4 and x + 3 to zero and got the correct solutions. But,
they thought that the x in the x 2 term in the original equation represented a dif-
ferent variable from the other x in the same equation. When asked to check their
solutions, some interviewees said they did not know how to do that. When asked to
try, some “checked” into the (x − 4)(x + 3) = 0 form of the equation, replacing x
by 4 in (x − 4) and x by −3 in (x + 3). After noting that 0 × 0 = 0 was true, they
concluded that their solutions were correct. Others substituted in x 2 − x = 12, but
let x equal one “solution” with x 2 , and x equal the other “solution” for the “−x.”
These students arrived at “16 + 3 = 12” and wondered why this was false.
Vaiyavutjamai and Clements (2006a) found evidence that after lessons on
quadratic equations many high school students thought that each of the equations
(x −3)(x −5) = 0, x 2 −x = 12 and 2x 2 = 10x had two variables, not one. Clements
and Ellerton (2006) and Vaiyavutjamai et al. (2005) conjectured that the same would
be true of most U.S. teacher-education students seeking endorsement as specialist
middle-school mathematics teachers. The misconception could have arisen from
statements, often made by algebra teachers, and often written in algebra textbooks,
that expressions like 2x 2 and 10x are “unlike terms.” It might also have arisen from
students having misinterpreted their teachers’ (correct) statements that quadratic
equations can “have two different solutions.” In students’ minds, such a statement
could mean that if two x’s appeared in an equation then each should take a dif-
ferent value. That could explain why, even at the post-instruction stage, relatively
few students gave both solutions to x 2 = 9. In the words of an interviewee, “in that
equation, x appears only once, and therefore there is only one solution.”
Student Misconceptions with Regard to Linear Inequalities
Although both Blanco and Garrote (2007) and Vaiyavutjamai and Clements (2006b)
identified numerous fossilized misconceptions that guided students’ thinking with
respect to linear inequalities, both pairs of researchers identified two particularly
common misconceptions: the most common arose from a tendency to give the an-
swer to the corresponding equation; and the second arose from an expectation that
there is only one real-number value of the variable in an inequality that makes that
inequality true. Vaiyavutjamai and Clements (2006b) investigated the thinking of
231 high school students with respect to linear algebraic inequalities before and
after a sequence of 13 algebra lessons specifically dealing with linear equations
and inequalities. At both the pre-instruction and post-instruction stages, Vaiyavutja-
mai asked the 231 students to solve a sequence of carefully chosen linear inequal-
ities, and then conducted pre- and post-instruction interviews with 18 students (6
high-performers, 6 middle-performers, and 6 low-performers). She reported that
before the lessons most students had little idea of what they were doing when
they attempted to solve linear inequalities, and that the same was still true after
the lessons—even though the teachers thought that their students had developed a
sound understanding. This lack of understanding was confirmed when, on a pencil-
and-paper retention test administered 6 to 12 weeks after the post-instruction test,
most of the 231 students still did not do well on inequality tasks.
The Tendency to “Give the Solution to the Corresponding Equation” Vaiyavut-

jamai and Clements’s (2006b) analysis of errors, at pre- and post-instruction stages,
and also at a retention stage, on all of the inequality tasks revealed that the most
common, and most stubborn, misconception was that the solution to an inequal-
ity could be obtained by merely solving the corresponding equation. Those whose
thinking was dominated by this misconception thought that they should treat the in-
equality as an equation, except that the inequality symbol in the original statement
should be used instead of an equals sign.
The Tendency to Think Only One Value of the Variable Will Make an Inequal-
ity True Vaiyavutjamai and Clements (2006b) provided space for students to show
their working when responding to tasks and, in addition, they provided “answer
boxes” in which correct answers were supposed to be written. Students attempting
to solve inequalities often showed correct working and arrived at correct answers,
but in the “answer boxes” they wrote single numerals as answers. Thus, for example,
for 3x ≤ 6, even if a student had written something like “3x/3 ≤ 6/3,” and “x ≤ 2,”
the numeral “2” would be all that would appear in the answer box. Interviews re-
vealed that most of the interviewees who did this usually thought that the inequality
should have just one numerical solution. Interviews also revealed that many students
who gave correct written answers, and whose written scripts suggested that they had
understood what they had written, harbored serious misconceptions with respect to
the inequalities that they had just solved. They thought that a final answer should
not “still have an x in it.” For them, a statement like “x ≥ 1” could not be the final
answer for 3 − 4x ≤ 6x − 7. For years they had learned to “solve for x,” and they
had come to believe that the answer had to be a single numeral.
Thus, for example, for the inequality 3 − 4x ≤ 6x − 7, 15%, 20%, and 24% of
the 231 high school students in Vaiyavutjamai and Clements’s (2006b) study gave
“1” as their answer at the pre-teaching, post-teaching, and retention stages, respec-
tively; another 38%, 24%, and 30% gave single-numeral answers other than “1” at
the respective stages. Altogether, 53%, 44% and 54% of the students, respectively,
responded to a request to solve 3 − 4x ≤ 6x − 7 by giving a single numeral as the an-
swer. Furthermore, interviews revealed that many of those who arrived at x ≥ 1, and
left that as their written answer, still thought that the answer was “1,” in the sense
that they believed that 1 was the only real-number value of x which would make
3 − 4x ≤ 6x − 7 true. The “single number answer” tendency, with several variations
(like giving the “additive inverse” as the answer), guided much of the students’
thinking about linear inequalities, with 68%, 39%, and 56%, respectively, of the
231 students giving single numerical answers for 1 − x < 0 at the pre-instruction,
post-instruction and retention stages. In other words, except for a period immedi-
ately after the lessons on inequalities, most of the 231 participating students thought
that “1” was the only value that made the inequality 1 − x < 0 true.
Many episodes in Vaiyavutjamai and Clements’s (2006b) interview data indi-
cated that interviewees did not know what an answer to a request to solve an in-
equality should look like. When “solving” inequalities they attempted to mimic the
methods shown by their teachers in model examples worked on the board, or on a
projector, or in examples used in the textbook. Inequality symbols, number lines,
and operations were used, liberally, but how everything could be linked together,
and what the actual answers were (after the sequence of thought-to-be equivalent
inequalities had been dealt with) seemed to be beyond most of the interviewees.
Both Blanco and Garrote (2007) and Vaiyavutjamai and Clements (2006b) con-
cluded that a major part of the difficulty that students experienced with inequalities
arose from the semantic complexity of inequality tasks. To put the matter bluntly,
other than manipulate symbols dutifully to get “right answers,” most of the students
did not really know why they did what they did. Students in the Vaiyavutjamai and
Clements (2006b) study tended to think that the numeral on the right-hand side of
the final line of their setting out was the solution to the inequality. High-stream stu-
dents were more likely to think that the solution to an inequality comprised a set of
numbers, but in most cases they were not clear what that meant in relation to the
initial inequality. Despite an instruction to find all real numbers that made an open
sentence true, many thought in terms of natural number or integer solutions only.
Interview data revealed that often interviewees did not know that they were making
mistakes, and they did not know how to check their solutions.
In Vaiyavutjamai and Clements’s (2006b) study, mean differences between high-,
medium-, and low-stream classes, at corresponding stages, on individual questions
were large. Most medium- or low-stream Grade 9 students could not correctly solve
simple linear inequalities before or after lessons on inequalities. This was even more
pronounced at the retention stage, several months after the lessons on inequalities.
The Pre-Service Teachers Involved, and Tasks Used, in the

Present Study
This chapter reports analyses of data from 328 prospective middle-school teach-
ers who, at various times during the period 2006–2009, were taking, in classes of
between 25 and 30 students, an “Algebra for Teachers 2” (“AT2”) algebra content
course at a large North American university. AT2 was the last algebra course these
students would take before they would graduate and become formally qualified to
teach Algebra 1 to Grade 7 or Grade 8 students. The analysis will be especially
concerned with the pre-AT2 knowledge of these students, particularly in relation
to linear and quadratic equations and inequalities. In another paper (Clements and
Ellerton 2009) we have described, in detail, a so-called “5-R intervention program”
which enabled these prospective middle-school teachers ultimately to identify and
eliminate serious fossilized misconceptions, and then to apply (and retain) the ap-
propriate conceptions that they had learned. Towards the end of this chapter we shall
present a summary of the 5-R procedure as well as results of its application with the
328 students.
The 328 students involved in our study had done well at school in their math-
ematical studies, and that was one of the reasons why they had decided to seek
endorsement as specialist middle-school mathematics teachers. In written responses
to a questionnaire administered at the beginning of a semester, almost all of the stu-
dents indicated that they really liked the subject and were confident that they would
not have too much difficulty completing the necessary combination of mathematics,
mathematics teacher-education, and education studies that would enable them to be-
come fully qualified, competent teachers of middle-school mathematics (Clements
and Ellerton 2009).
This section of the chapter focuses on the instruments developed to assess the
students’ algebraic thinking (a) immediately before a sequence of lessons aimed
at helping the students to understand quadratic equations and linear and quadratic
inequalities; and (b) 6 to 12 weeks after the planned intervention. In fact, the inter-
ventions proved to be successful with all the different AT2 classes. Almost all of the
students were able to bring their knowledge of elementary equations and inequali-
ties to a level that would enable them to teach that topic competently—at least from
a content perspective—to middle-school mathematics students.
“Clever” Tasks
For several decades, now, there has been a strong emphasis on developing apparently
simple tasks that have the potential to reveal how students think about algebra. In
this chapter we have coined the term “clever tasks” to refer to tasks which, although
apparently simple to persons for whom the tasks were designed, are such that they
have the power to identify fossilized misconceptions related to cognitive aspects of
the tasks. Clever tasks can be questions on pencil-and-paper tests, or questions asked
in interviews, or even questions asked in normal classroom discourses. They have
often been used in algebra education research, and clearly can play an important
role in algebra classes. The main aim of this section is to outline how we developed
a set of clever tasks of the pencil-and-paper variety.
Data from tasks used in studies involving middle-school and high school students
in the 1970s and early 1980s in the United Kingdom by Küchemann (1981), Hart
(1981), and Booth (1984) have been much cited in the literature. Perhaps the most
famous clever task in algebra education, though, was developed in the United States
of America. It has come to be known as the “students and professors” problem:
“At this university there are six times as many students as professors.” If S stands for the
number of students and P for the number of professors, write down an equation which
shows how S is related to P . (Clement 1982, p. 19)
Clement (1982), and other researchers, used this task with students at many levels.
Clement reported that even college engineering students sometimes gave the “re-
versal” response P = 6S (rather than S = 6P ). MacGregor (1991) decontextualized
the “students and professors” task, and reported data from 235 Grade 9 students in
12 schools in Melbourne, Australia, on the following task, which might be called
the “y = 8x” task.
“The number y is eight times the number x.”
Write that information in mathematical symbols. (p. 85)
Almost all teachers of algebra find it difficult to believe that more than 50% of high
school students who had been taking algebra classes for at least two years answered
the “y = 8x” task incorrectly (see MacGregor 1991 for Australian data; Lim and
Clements 2000 for Bruneian data; Vaiyavutjamai 2004 for Thai data). The most
common error, found in each of Australia, Brunei Darussalam and Thailand, was
“x = 8y,” but other common errors included y 8 = x, y 8 > x, 8y > x, y 8 x, and 8yx.
Fujii’s (2003) research in Japan and the United States was based on students’
responses to “clever” tasks like the following:
Mary has the following problem to solve:
“Find the value(s) of x for the expression x + x + x = 12.”
She answered in the following manner (a) 2, 5, 5; (b) 10, 1, 1; (c) 4, 4, 4.
Which of her answers is/are correct? . . . State the reason for your selection.
(Fujii 2003, p. 51)
Fujii (2003) reported that high proportions of elementary, middle-school, and high
school students in both Japan and the United States answered questions like this
incorrectly.
There are many other tasks in the algebra education research literature that might
be regarded as “clever.” Creating such tasks usually requires a combination of teach-
ing experience and associated pedagogical content knowledge. A person attempting
to create a clever task has to reflect on precisely what he or she wants to find out
about learners, and on how to create an apparently simple task that is likely to re-
sult in non-trivial proportions of students making errors that will unambiguously
indicate serious misconceptions.
The tasks cited as “clever” in the above discussion were associated with linear
relationships. As Vaiyavutjamai and Clements (2006a, 2006b) and Kieran (2007)
have noted, there has been an abundance of research on solving linear equations,
but hardly any on solving quadratic equations or inequalities. It will be useful
here to summarize some tasks developed in relation to quadratic equations by Lim
(2000), in Brunei Darussalam, and in relation to linear and quadratic inequalities,
by Vaiyavutjamai (2004) in Thailand.
Pencil-and-paper tasks developed by Lim (Lim 2000; Lim and Clements 2000),
and used by Vaiyavutjamai (2004), asked high school algebra students to find real-
number solutions to quadratic equations expressed in the form x 2 = K(K > 0),
or in the form (x − a)(x − b) = 0 (where a, b can represent any real numbers).
Vaiyavutjamai (2004) used tasks of this kind, and also linear inequality tasks, in her
doctoral study involving 231 ninth-grade students in Thailand.
Tsamir and Bazzini (2002) reported that 16- and 17-year old Israeli and Italian
students’ solutions to inequalities tended to be confused as a result of misapplica-
tions of a “balance” model. Tsamir et al. (1998) found that drawing graphs could
assist understanding, and from that perspective it is interesting that the students in
Vaiyavutjamai’s (2004) sample had not been taught to draw graphs when solving
linear inequalities (for that was not on the national Thai syllabus at that level).
Developing the Pencil-and-Paper Instruments
We developed an instrument that would enable us quickly and accurately to identify

the meanings that prospective middle-school teachers give to the processes they
use when solving linear and quadratic equations and inequalities. In keeping with
conclusions reached in the relevant literature (e.g., Kieran 2004, 2007; Tsamir et al.
1998; Tsamir and Bazzini 2002; Vaiyavutjamai 2004; Vaiyavutjamai and Clements
2006b), we also deemed it to be important that any tasks that we developed should
be such that students’ responses to them would help to reveal their thinking about
connections between solving equations and solving associated inequalities.
The Eight Equation/Algebraic Inequality Pairs
Our main instrument comprised eight pencil-and-paper equations and eight pencil-
and-paper inequalities, each inequality corresponding to one of the equations. The
eight equations and corresponding inequalities are shown in Table 1. Seven of the
eight pairs were expected to be clever tasks, in the sense that student responses
to them would be likely to reveal fundamental misconceptions. The tasks in the
first equation-inequality pair were included as “warm-up” exercises. At the time of
its administration to the prospective teachers, the pencil-and-paper instrument was
introduced with the written statement: “For each equation or inequality, state all
real number(s) which could replace x to make the statement true.” Ample room for
working was available immediately after the statement of an equation or inequality,
and for each question an “answer box” was also provided in which students were
expected to write their final answer. Note that an equation and its corresponding
inequality were not placed next to each other on the actual test instrument. For the
purposes of subsequent analysis, however, they were regarded as a pair. A simi-
lar overall format had been used, successfully, by Lim (2000) and Vaiyavutjamai
(2004).
The prospective teachers were not permitted to use any form of electronic cal-
culator (graphing or otherwise) when answering these questions. That restriction
was made for two reasons. First, from experience we knew that although all of
the teacher-education students had been introduced to graphing calculators in high
school, most of them did not know how to use them competently and confidently.
Table 1 The eight equations

and eight corresponding The 8 Equation Tasks The 8 Inequality Tasks
inequalities (pre-teaching (Pre-Teaching) (Pre-Teaching)
versions)
9(x − 1) = 0 4(x − 1) > 0
1 1
=3 >4
x x
x =9
2 2
x >4
9 4
x= x>
x x
x2 + 6 = 0 x2 + 2 > 0
4(x + 1) = 4(x − 3) 9(x + 1) > 9(x − 2)
(x − 3)(x − 2) = 0 (x − 3)(x − 1) > 0
x + 5 = 8 − (3 − x) x + 3 > 6 − (3 − x)
Second, and more importantly, we wanted students to show their working. In partic-
ular, if they employed mental imagery involving number lines, or Cartesian graphs,
we wanted them to represent that imagery in the section on the test paper designed
for showing working.
For each equation and inequality, students were asked to indicate, with respect to
each solution they offered, how confident they were that that solution was correct.
For each response a student could choose any one of five possible statements: “I’m
certain I’m right,” “I think I’m right,” “I’ve got a 50-50 chance of being right,”
“I think I’m wrong,” and “I’m certain I’m wrong.”
Rationales for including each of the equation/inequality pairs in the instrument
will now be given.
The First Equation/Inequality Pair
Equation 1: 9(x − 1) = 0 Corresponding inequality: 4(x − 1) > 0

Equation 1 and the corresponding inequality were expected to be relatively easy
for the prospective teachers, certainly the easiest of the eight equation/inequality
pairs. The pair was never thought of as defining a “clever task.” That said, we were
interested in finding out the percentage of prospective middle-school students who
gave “1,” or “x < 1,” as the answer to the inequality.
The seven other equation/inequality pairs might all be regarded as “clever tasks.”
They were developed out of the experience and pedagogical content knowledge of
the writers, and a decision to use them in the study was made only after careful
analysis of relevant trial data.
The Second Equation/Inequality Pair
1 1
Equation 2: =3 Corresponding Inequality: >4
x x
The algebra education literature has established that in their middle- and high-school
years, most students are told, many times, that solving equations is just a matter
of maintaining balance (Kieran 2007). As a result, many students come to believe
that it is always satisfactory to “multiply both sides” or to “divide both sides,” by
anything, including expressions involving a variable. Thus, with Equation 2, we
expected that some of the prospective teachers would multiply both sides by x, to
obtain the equivalent equation 1 = 3x, and then, after dividing both sides by 3,
would find the correct solution, x = 1/3. It was expected that that procedure would
be standard for most of the 328 prospective middle-school teachers, and that most
would get the correct solution for Equation 2.
Some students who did not multiply both sides by x were expected to adopt a
“reciprocation” method by which a/b = c/d would be regarded as equivalent to
b/a = d/c. This method is not recommended, for it can generate incorrect solutions
(e.g., consider what would happen with 3/x = 4/x). However, provided 3 is thought
of as 3/1, its application to Equation 1 would give x/1 = 1/3, which would quickly
give the correct solution. A third method involves “cross-multiplying” from 1/x =
3/1 to obtain 3x = 1, etc.
Students who attempted to solve the corresponding inequality by merely “solv-
ing the corresponding equation and maintaining the direction of the given inequal-
ity” would quickly arrive at “1/4 > x” without having realized that since x could
be negative there was no guarantee that the “direction” of the inequality should be
maintained. Students who thought about the meaning of the inequality should rec-
ognize that negative values of x would make the inequality false, and that x = 0
would generate a meaningless statement. Therefore, “1/4 > x” (or the equivalent
“x < 1/4”) is clearly wrong.
It was also expected that some students would offer x = 1/4, or simply 1/4, as
the solution to the corresponding inequality.
The Third Equation/Inequality Pair
Equation 3: x2 = 9 Corresponding Inequality: x2 > 4

It might reasonably have been expected that prospective middle-school teachers
seeking endorsement to teach mathematics, almost all of whom had been success-
ful with Algebra 1, Algebra 2 and Pre-Calculus (and some of whom had also suc-
cessfully formally studied Calculus, as well as some university-level mathematics),
might have no difficulty with Equation 3 (x 2 = 9). However, our previous experi-
ences with prospective middle-school teachers, and our knowledge of Vaiyavutja-
mai’s (2004) and Lim’s (2000) findings, suggested to us that many of the 328 stu-
dents would give “x = 3” as the only solution. Those same students would be likely
to give “x > 2,” or even just “2,” as their solution to the corresponding inequality
(x 2 > 4). Many students who correctly gave “x = ±3” as solutions to Equation 3
might be expected to offer statements like “x > ±2,” or “x > −2 and x > 2” (or
“x > −2 or x > 2”) as solutions to the corresponding inequality.
The Fourth Equation/Inequality Pair
9 4
Equation 4: x= Corresponding Inequality: x >
x x
Equation 4 is equivalent to Equation 3, and it would be expected that almost ev-
eryone who found the two solutions for Equation 3 would also find the same two
solutions for Equation 4. Similarly, those who found only one solution for Equation
3 would find that same solution for Equation 4. So far as the corresponding inequal-
ity is concerned, it was expected that many students who correctly gave “x = ±3”
as solutions to Equation 4 would offer incorrect statements such as “x > ±2,” or
“x > −2 and x > 2” (or “x > −2 or x > 2”) as their solutions for the correspond-
ing inequality. Those who indicated that x = 3 was the only solution to Equation
4 would be likely to assert that x > 2 was the only solution to the corresponding
inequality.
The Fifth Equation/Inequality Pair
Equation 5: x2 + 6 = 0 Corresponding Inequality: x2 + 2 > 0

Written instructions on the test indicated that students should find all real-number
solutions to the equations and inequalities.
√ For Equation
√ √5 it was expected
√ that many
students would offer solutions like i 6, or ±i 6, or −6 or ± −6. We antici-
pated that students who answered in any of those ways would either not have read
the instructions carefully or would not be knowledgeable about the concept of a
real number. The prospective middle-school teachers had only occasionally dealt
with equations with no solutions, and it was expected, therefore, that some would
be confused by Equation 5.
The corresponding inequality provides an elegant test of the extent to which
respondents thought about meaning, for although Equation 5 has no real-number
solutions, the corresponding inequality is true for all real-number values of x. Al-
though it might be expected that most of the students who gave incorrect answers to
Equation 5 would give similar answers for the corresponding inequality, only with
6 replaced by 2, and with a “greater than” sign indicated, it was also conjectured
that there would be some students who, although giving an incorrect answer for the
equation, would recognize that x 2 + 2 > 0 must be true for any real-number, x.
The Sixth Equation/Inequality Pair
Equation 6: 4(x + 1) = 4(x − 3)

Corresponding inequality: 9(x + 1) > 9(x − 2)
Equation 6 is another example of an equation that has no real-number solution.

Students arriving at a statement like “4 = −12” might be expected to experience
cognitive tension, and of interest was whether this would result in their abandoning
the question without any further thinking, or whether they would reflect on what this
might imply for the original equation—in other words, this task invites students to
think holistically. The corresponding inequality is such that any real-number value
of x will make it true. This inequality also invites holistic thinking.
The Seventh Equation/Inequality Pair
Equation 7: (x − 3)(x − 2) = 0
Corresponding inequality: (x − 3)(x − 1) > 0
Equation 7 was included because of the interesting data generated with respect to
such equations in Lim’s (2000) study, in Clements and Ellerton’s (2006) study, and
in Vaiyavutjamai’s (2004) study. It was expected that for the corresponding inequal-
ity many students would answer “x > 3, x > 1,” their thinking being guided by the
misconception that there is a rule for inequalities analogous to the null factor law
for equations.
The Eighth Equation/Inequality Pair
Equation 8: x + 5 = 8 − (3 − x)
Corresponding inequality: x + 3 > 6 − (3 − x)
School students are rarely asked to think about equations like x + 5 = 5 + x, and it
was conjectured that those who had never been encouraged to think meaningfully
when attempting to solve equations would find Equation 8 difficult.
The corresponding inequality provides another test of the extent to which respon-
dents think about meaning. Although Equation 8 is true for all real-number values of
x, the corresponding inequality is not true for any real-number value of x. It might be
expected that most of the students who gave incorrect answers to Equation 8 would
also answer the corresponding inequality incorrectly. It was conjectured, though,
that there would be students who, although giving an incorrect answer for Equation
8, would nevertheless recognize that, since x + 3 is never greater than 3 + x, the
inequality has no real-number solution.
The Quadratic Equation Scenario Lim (2000), Vaiyavutjamai (2004), Clements

and Ellerton (2006), and Vaiyavutjamai et al. (2005) had all noted that, when asked
to solve a quadratic equation in the form (x − a)(x − b) = 0, many students who
gave correct solutions x = a, x = b, thought that the x in (x − a) was equal to a,
and simultaneously the x in (x − b) was equal to b. Our aim was to develop a task
which was such that students who harbored this misconception would be quickly
and unambiguously identified.
After considerable pilot testing, we found the “quadratic equation scenario” (see
Fig. 1), which we developed, to be suitable for our purposes.
Students were asked to solve (x + 2)(2x + 5) = 0, and then to check

their answer. One student, Carrie, wrote the following (line numbers
have been added):
(x + 2)(2x + 5) = 0 Line 1
∴ 2x 2 + 5x + 4x + 10 = 0 Line 2
∴ 2x 2 + 9x + 10 = 0 Line 3
∴ (2x + 5)(x + 2) = 0 Line 4
∴ (2x + 5) = 0 and (x + 2) = 0 Line 5
∴ 2x = −5 and x = −2 Line 6
∴ x = −5/2 and x = −2 Line 7
Check: Put x = −5/2 in (2x + 5), and put x = −2 in (x + 2).
Thus, when x = −5/2 and x = −2, (2x + 5)(x + 2) is equal to
0 ×0 which is equal to 0.
Since 0 is on the right-hand side of the original equation, it fol-
lows that x = −5/2 and
x = −2 are the correct solutions.
Comment fully on Carrie’s responses.
Fig. 1 The quadratic equation scenario
In Fig. 1, line numbers were shown to assist respondents to comprehend what

they were expected to do. Use of the scenario not only helped us identify students
who harbored a “0 × 0 = 0” misconception, but also any of the three other common
misconceptions to be found in the scenario. The respondents’ task was to comment
on the solutions presented by a hypothetical student, “Carrie,” when solving the
equation (x + 2)(2x + 5) = 0, and then to consider the appropriateness of Carrie’s
method for checking the solutions that she obtained. Each response to the quadratic
equation scenario (in Fig. 1) was given a score from 0 to 4, depending on how many
of the following four points were noted:
• Lines 2, 3, and 4 were unnecessary, since the left-side is already factored in
Line 1.
• In Lines 5 through 7, the word “or,” and not “and,” should have been used.
• For the check, the solutions should have been substituted into Line 1.
• For the check, each solution should have been substituted into both parentheses
in the initial equation.
Each response to each of the eight equations and eight inequalities shown in Table 1
was scored 0 (if it was incorrect) or 1 (if it was correct), and up to 4 was allocated for
the quadratic equation scenario. Thus, each student received a total score out of 20.
Study Design, and Results
As well as gathering pre- and post-intervention data, the study design incorporated
a retention component aimed at checking whether students retained correct concep-
tions, or misconceptions, over a period of between 6 and 12 weeks. An important
aspect of the intervention program was the creation of an environment in which all
students would reflect metacognitively on the strategies that they used when they
attempted to solve equations and inequalities (Clements and Ellerton 2009).
In this chapter the results of the analyses of post-intervention/retention data are
only briefly summarized. The focus of this chapter is on what the 12 classes of
students, totaling 328 students, knew about the eight equation/inequalities pairs—
shown previously in Table 1—at the start of the semester in which they were en-
rolled in AT2, and how they responded, at that same time, to the quadratic equation
scenario.
Our challenge was to create a method which would help students develop a re-
lational understanding of algebra content concerned with elementary equations and
inequalities that they should have already acquired as a result of their having taken
Algebra 1 and Algebra 2 at high school. In fact, the instructors could dedicate only
a relatively small amount of AT2 class time (a total of about three hours, including
pre-testing) to improving the students’ knowledge and skills with respect to elemen-
tary equations and inequalities. That was because the focus of the AT2 course was
expected to be on more advanced topics.
Population and Sample Considerations
Between 2006 and 2009 almost all AT2 classes were taught by the same two in-
structors (Instructor A and Instructor B), both of whom were well qualified and
experienced mathematics educators. For most of the semesters there were only two
sections for AT2, one of which was taught by Instructor A and the other by In-
structor B. In each section there were always between 25 and 30 students. For each
semester, placement of students into the two AT2 classes was done by departmental
administrators, without consultation with the instructors. Initial testing suggested
that for each semester the two AT2 sections had very similar mathematics profiles.
Mean scores on initial tests given to sections during the first AT2 class in a semester
were never very different, and these initial mean scores remained similar throughout
the period 2006–2009. During that period the highest mean score for a section out of
a possible 20 (16 for the equations and algebraic inequalities and 4 for the quadratic
equation scenario) was 8.0 and the lowest was 6.1, with an overall mean of 6.5. The
highest individual score of any of the 328 students was 16 (out of the possible 20),
and the lowest was 0.
From a “population” perspective, there is no reason to suppose that the stu-
dents taking AT2 could be regarded as representative of any well-defined group,
except perhaps of students at the particular university who were preparing to become
middle-school mathematics specialists. The university at which the study occurred

has a strong reputation in mathematics education, and it is likely that most of the 328
students were better qualified, and mathematically more knowledgeable, than “typi-
cal” teacher-education students preparing to be middle-school mathematics teachers
at many other US universities or colleges.
Results
Summaries of performances on each of the eight equations/inequalities pairs are

shown in Table 2.
Table 2 Percentages correct, 328 mathematics teacher-education students on eight equa-

tion/inequalities pairs
Equation Number (and %) “Corresponding” Number (and %)
Correct (n = 328) Algebraic Inequality Correct (n = 328)
9(x − 1) = 0 321 (98%) 4(x − 1) > 0 210 (64%)

1 1
=3 268 (82%) >4 4 (1%)
x x
x2 = 9 74 (23%) x2 > 4 16 (5%)
9 4
x= 69 (21%) x> 1 (0%)
x x
x2 + 6 = 0 69 (21%) x2 + 2 > 0 53 (16%)
4(x + 1) = 4(x − 3) 173 (53%) 9(x + 1) > 9(x − 2) 77 (23%)
(x − 3)(x − 2) = 0 194 (59%) (x − 3)(x − 1) > 0 2 (1%)
x + 5 = 8 − (3 − x) 71 (22%) x + 3 > 6 − (3 − x) 109 (33%)
For the 328 AT2 students, the mean score on the 16 tasks (making up the 8 equa-
tion/inequalities pairs) was 5.4 out of 16. For the equations, about 47% of responses
were correct, and for the algebraic inequalities, about 18% of responses were cor-
rect. If responses to 9(x − 1) = 0 and 4(x − 1) > 0 were not taken into account,
then 40% of responses for the equations and 11% of responses for the inequalities
were correct. Yet, most students indicated that they were certain they were right, or
they thought they were right, for all 16 tasks. As one student would subsequently
write, in a reflection: “This was an interesting exercise because I thought I knew
the correct answers, but in fact I did not.” Most students experienced a reality check
in the sense that they quickly discovered that they had obtained wrong answers to
equations and inequalities that they were certain, or they thought, they knew how to
solve correctly.
Table 3 summarizes performances of the 328 students on the quadratic equation
scenario, and Table 4a provides an overview of results of our quantitative and qual-
itative analyses of student responses to four of the eight equation/inequality pairs.
Interestingly, no student drew a sketch graph, for any inequality, in the working
spaces provided on the pencil-and-paper instrument.
Table 3 Summary of numbers (and percentages) of 328 students identifying the four “errors” in
the quadratic equation scenario (from Fig. 1)
Identified Stated “and” Stated Carrie should Identified the Mean Score (/4) (and
Unnecessary should be “or” have Substituted in 0 × 0 = 0 SD), Quadratic Equation
Work (Lines 5–7) Line 1 Misconception Scenario
(Lines 1–3)
176 (54%) 36 (11%) 28 (9%) 107 (33%) 1.07, SD = 0.45
Table 4a Summary of quantitative and qualitative analyses of student responses to four equa-
tion/inequality pairs
Equation or % of Most Common Errors Comments
Inequality Responses (included if at least 10% of all
Correct students gave that response)
9(x − 1) = 0 98% This was by far the easiest of all the

tasks. Most students obtained the
equivalent equation 9x = 9, but 5 of
the 328 students divided both sides by
9 to get x − 1 = 0.
4(x − 1) > 0 64% • x < 1 (14% of all responses) These results are consistent with
• 1, or another number (11%) Vaiyavutjamai’s (2004) analyses of her
• x > 2 (10%) inequalities data.
1
=3 82% • 3 (10% of all responses) When working was shown, multiplying
x both sides by x was common, as was
“cross-multiplication.”
1
>4 1% • x < 1/4 (31% of all The small amount of working that was
x responses) shown suggested that students merely
• x > 1/4 (24%) treated the inequality as if it were an
• 1/4 (23%) equation.
x2 = 9 23% • 3 (68% of all responses) Working was rarely shown.

x2 > 4 5% • x > 2 (32% of all responses) Once again, working was rarely
• x > ±2 (14%) shown. In no case was a number line,
• 2, or ±2 (13%) or a sketch of a Cartesian graph, shown
• x > 3 (10%) in working space.
9
x= 21% • 3 (59% of all responses) Students tended to write: x 2 = 9,
x x = 3.
4
x> 0% • x > 2 (35% of all responses) Students tended to write: x 2 > 4, and
x • x > ±2 (15%) then x > 2, or x > ±2, or simply 2, or
• 2, or ±2 (11%) ±2. No number line or graph was
• 3 (10%) drawn by any student.
Table 4b provides an overview of results of our qualitative analyses of responses

to the remaining four equation/inequality pairs. Once again, no student drew a graph,
for any inequality, in the working spaces.
Table 4b Quantitative and qualitative analyses of responses to the remaining four equa-
tion/inequality pairs
Inequality Responses (included if 10%, or more
which were than 10%, of all students
Correct gave that response)
√ √
x2 + 6 = 0 21% • −6 or i 6 (33% of all Most students wrote x 2 = −6
responses)
√ √ and then, without further
• ± −6 or ±i 6 (14%) comment, offered something
•0√(11%) √ which, presumably, they
• 6 or ± 6 (10%) thought was the solution to the
equation.
x2 + 2 > 0 16% • “No solution,” or the “null Students tended simply to write
set” (16%
√ of all responses)
√ down a “solution” without
• x > i 2 or ±i 2 (16%) showing working, except for
• A single number (10%) x 2 > −2. No student sketched
a graph. Except for those who
gave a correct answer, there
appeared to be little evidence
of holistic thinking about the
inequality’s meaning.
4(x + 1) = 4(x − 3) 53% • 0 (11% of all responses) Often it was not clear from
• A single number other than students’ written answers
0 (15%) whether they realized there was
• 4 = −12, or −4 = 12 (10%) no solution. Statements like
“will not work” were marked
correct.
9(x + 1) > 9(x − 2) 23% • “No solution” (23% of all 34% of students did not
responses) proceed beyond writing down
• 27 > 0, or 9 > −18 (12%) 9x + 9 > 9x − 18.
(x − 3)(x − 2) = 0 59% • x 2 − 5x + 6 (13% of all Evidence from the “quadratic

responses) equation scenario” would
• 3 (10%) suggest that the thinking of
• 2 (10%) many of those who gave
“correct” solutions (e.g., “3,
2”) was guided by a serious
misconception.
(x − 3)(x − 1) > 0 1% • x > 3, x > 1 (29% of all The “x > 3, x > 1” response
responses) was, almost certainly, based on
• 3, 1 (15%) a faulty translation of the null
• x > 3 (13%) factor law for equations to the
realm of inequalities.
x + 5 = 8 − (3 − x) 22% • 0 (25% of all responses) 21% of students did not

• ∞ (11%) proceed beyond writing down
• 0 = 0 (10%) x + 5 = 5 + x.
Table 4b (Continued)
Inequality Responses (included if 10%, or more
which were than 10%, of all students
Correct gave that response)
x + 3 > 6 − (3 − x) 33% • 0 (17% of all responses) This was the only inequality in
• x > 0 (10%) an equation/inequality pair for
which a greater number of
correct answers were given
than for the corresponding
equation. 18% of students did
not proceed beyond writing
down x + 3 > 3 + x.
Conclusions in Relation to the Prospective Teachers’ Knowledge

of Algebraic Inequalities
In this section we reach four conclusions from our analyses of data relating to al-
gebraic inequalities. These conclusions relate to: (a) students’ understandings and
misunderstandings of relationships between equations and algebraic inequalities;
(b) the tendency of students to think that an inequality has just one numerical an-
swer; (c) the tendency to manipulate algebraic expressions without paying due at-
tention to meaning-making; and (d) confusion about what to do when attempting to
solve a non-trivial algebraic inequality.
Over-Emphasising Relationships Between Equations and Inequalities The

students obviously recognized that there was a strong relationship between solv-
ing equations and solving corresponding inequalities. That was a positive result,
because when one is solving an inequality one needs to think about how the solu-
tions to that inequality relate to the solution(s) to the corresponding equation. How-
ever, our analyses revealed that relationships between solving algebraic inequalities
and corresponding equations had become so much emphasized in the teaching and
learning of inequalities that they had become a destructive rather than a constructive
force.
Consider, for example, the pre-instruction data summarized in Table 4a in rela-
tion to the inequality x > x4 . None of the 328 students solved this inequality cor-
rectly. Despite the fact that the students had been encouraged to “show their work-
ing,” none of them drew a number line or showed a graph. Students tended simply to
“multiply both sides of the equation by x” to obtain what they, incorrectly, thought
was an equivalent inequality x 2 > 4. To solve this inequality they mentally solved
the equation x 2 = 4 (rarely was this equation actually written down by a student).
They then wrote down their solutions to this equation with the “>” sign placed
where the “=” sign would “normally” be placed. The students’ attempts to solve
x 2 = 4 (see Table 4a) revealed that most students missed “−2” and this was re-
flected in the fact that 35% of students gave x > 2 as their sole solution for x > x4 .
Even students who knew that there were two solutions to x 2 = 4 often concluded
that the solution for x > x4 was x > ±2 (15% of all students gave that answer).
Thinking that an Inequality has Just One Number in its Truth Set Vaiyavutja-
mai and Clements (2006b) reported numerous transcripts of interviews with students
who, even though they wrote “x > 2” for the solution to an inequality, thought that
the actual inequality had just one numerical solution, namely “2.” In this present
study, 21% of the prospective middle-school teachers of mathematics stated that the
solution for x > x4 was 2, or 3, or ±2.
Manipulations Without Meaning-Making Although there were some encourag-

ing signs that some students were prepared to think about meanings behind equa-
tions and inequalities, most of the students seemed to be unwilling, or unable, to go
beyond mere symbol manipulation (c.f., Linchevski and Sfard 1991). Although it
was encouraging to note that over 50% of the students indicated that the equation
4(x + 1) = 4(x − 3) had no solution, it was discouraging to observe that only 23%
of students were willing or able to think their ways through 9(x + 1) > 9(x − 2).
With that inequality, 34% of the students did not proceed beyond writing down
9x + 9 > 9x − 18. Many of them would write, in subsequent written reflections,
that they simply did not know what to do with such an inequality. That unwilling-
ness, or inability, to think holistically about an equation was strongly revealed in
data associated with the equation/inequality clever tasks in Tables 4a and 4b. Thus,
while 59% of responses to (x − 3)(x − 2) = 0 were deemed to be correct, only 1%
of responses to the corresponding inequality, (x − 3)(x − 1) > 0, were correct. Yet,
all 328 students had previously taken courses in which equations and algebraic in-
equalities were part of the curriculum, and all of them had also studied quadratic
functions. One of the encouraging signs from the data analyses summarized in Ta-
bles 4a and 4b was that more students gave a correct answer for x + 3 > 6 − (3 − x)
than for x + 5 = 8 − (3 − x).
Not Knowing How to Proceed with Inequalities When we analyzed students’

written responses to the equations/inequalities pairs it became obvious to us that
whereas the students had some ideas—albeit often imperfect ideas—about how to
solve equations, they rarely knew how to proceed when faced with inequalities like
x > x4 or (x − 3)(x − 1) > 0. Other than to recommend and sometimes illustrate
graphical approaches, textbook writers and algebra educators have rarely been par-
ticularly helpful on this matter—although Dobbs and Peterson (1991) did attempt
to do something about it. With x > x4 , for example, a suitable approach might be
to consider separately the cases for x > 0, and x < 0, and then to combine the
results. Thus, for x > 0, x > x4 is equivalent to x 2 > 4, which is equivalent to
x > 2. For x < 0, x > x4 is equivalent to x 2 < 4, and −2 < x < 0. Combining,
one finds that if x can be any non-zero number then the solution set for x > x4
is {x : −2 < x < 0} ∪ {x : x > 2}. This approach can be supported by the use of
graphical calculators, or by sketch graphs. Teaching students to interpret correctly
representations that appear on the screen of graphical calculators can be just as dif-
ficult as assisting them to learn appropriate algorithms (such as the one suggested
for x > x4 ), or to draw and interpret appropriate sketch graphs. The axiomatic ap-
proach and the graphical representation approaches need to be learned by students.
These approaches become most useful when one is used to complement the other,
and students are helped to reflect on meaning, and to translate from one approach to
another (Pimm 1995).
Most of the prospective middle-school mathematics teachers treated inequalities
as equations that happened to have inequality signs rather than equals signs. Most
(but not all) of them experienced difficulty with tasks for which holistic thinking,
rather than mere algebraic manipulation, was called for.
Prospective Teachers’ Knowledge in Relation to Quadratic

Equations
At the beginning of every semester during the period 2006–2009 we asked our AT2
students to find all real numbers that could replace x and make x 2 = 9 true. With ev-
ery class, on every occasion, more than 50% of the students gave one possible value
of x, and usually only between 20% and 30% included −3 in their answer (see the
overall result for the 328 students in Table 4a). In the data summarized in Table 4b it
can be seen that 79% of the 328 prospective mathematics teachers did not know that
x 2 + 6 = 0 has no real-number solutions. Almost 60% of the prospective teachers
stated that x = 3 and x = 2 were solutions to (x − 3)(x − 2) = 0, but data from the
quadratic equation scenario indicated that many of these students also thought that
the x in (x − 3) equalled 3 and, simultaneously, the x in (x − 2) equalled 2.
The inspiration for the design of the quadratic equation scenario arose from our
knowledge of the results of our own research with pre-service teachers in the United
States (Clements and Ellerton 2006) and of the results of research by Lim (2000), in
Brunei Darussalam, and Vaiyavutjamai (2004), in Thailand. The work by Lim and
Vaiyavutjamai deserves to be known by all algebra educators, and algebra classroom
teachers.
Lim (2000) noticed that some teachers, when teaching children how to solve
equations like (x − 3)(x − 2) = 0, “expanded the parentheses” on the left side to get
the equivalent equation x 2 − 5x + 6 = 0, and then re-factored the left side before
applying the null factor law. These teachers informed Lim that it was important for
students to write the equation in “standard form” before attempting to solve it. Thus,
our quadratic equation scenario featured a fictitious student, Carrie, attempting to
solve (x + 2)(2x + 5) = 0. She expressed this in equivalent form, 2x 2 + 9x + 10 =
0, before re-factoring and applying the null factor law. Almost half (46%) of our
prospective teachers of mathematics saw nothing strange about that.
Readers who feel that this could not happen in a nation like the United States
of America may wish to consult http://www.algebrahelp.com (Mishanski 2010).
According to the webpage, Algebrahelp.com “is a collection of lessons, calcula-
tors, and worksheets created to assist students and teachers of algebra.” Here is a
printout of the first 19 lines of the method that is shown for solving the equation
(x − 2)(x − 3) = 0. There are actually 57 lines in the recommended solution, but
the last 38 lines are not shown here because of space limitations. On the same web-
page, the recommended solution for (x − 2)(x − 2) = 0 also has 57 lines, with the
solution being given as x = {2, 2}.
Simplifying
(x + −3)(x + −2) = 0
Reorder the terms:
(−3 + x)(x + −2) = 0
Reorder the terms:
(−3 + x)(−2 + x) = 0
Multiply (−3 + x) ∗ (−2 + x)
(−3(−2 + x) + x(−2 + x)) = 0
((−2 ∗ −3 + x ∗ −3) + x(−2 + x)) = 0
((6 + −3x) + x(−2 + x)) = 0
(6 + −3x + (−2 ∗ x + x ∗ x)) = 0
(6 + −3x + (−2x + x 2 )) = 0
Combine like terms: −3x + −2x = −5x
(6 + −5x + x 2 ) = 0
Solving
6 + −5x + x 2 = 0
Solving for variable ‘x’.
Factor a trinomial.
(2 + −1x)(3 + −1x) = 0 . . .
Bad News, Good News and Some Concluding Comments
Bad News
The AT2 prospective teachers were not far away from becoming fully qualified,
endorsed middle-school teachers of mathematics, yet 41% of them did not solve
(x − 3)(x − 2) = 0 correctly, and 46% of them did not feel moved to comment that
Lines 1–3 in Carrie’s response (in the quadratic equation scenario) were unneces-
sary. Only 11% of them recognized that the correct connective in Lines 5–7 was “or”
(and not “and”), and 91% did not notice, or did not comment on, the fact that when
checking her solutions, Carrie did not substitute in the original equation in Line 1.
The serious misconception that x = −2 in (x + 2) and, simultaneously, x = −5/2
in (2x + 5), which led Carrie to conclude that since 0 × 0 = 0 both her solutions
must be correct, was noticed by only one-third of all students.
It is likely that the “errors” featured in the quadratic equation scenario are made
by school students in many nations, and research is needed to see how prevalent
the errors are among prospective and even practicing teachers. We have, in fact,
analysed responses by many experienced and qualified US middle-school teachers
of mathematics to the quadratic equation scenario, and found that they fare only
marginally better than the 328 prospective teachers in identifying Carrie’s errors.
Elsewhere (Clements and Ellerton 2009), we have provided evidence that unless
AT2 students are made aware of the kinds of errors featured in Table 3 and in Ta-
bles 4a and 4b, they are likely to begin their mathematics teaching careers guided
by the same misconceptions that became evident when AT2 students responded to
the tasks we asked them to do at the beginning of their AT2 courses.
Good News
All of the 328 participating AT2 students became aware of the misconceptions that
they displayed during the initial round of tests, and almost all of them were able to
demonstrate retention of newly-acquired knowledge, skills and concepts at the end
of the AT2 course. Their improvement was brought about by their active participa-
tion in what we have called a 5-R intervention program which featured the following
five components:
1. There was an initial reality check, by which students came to realize that their
thinking about the area under consideration had been guided by fossilized mis-
conceptions. This often generated confusion among students who began to real-
ize that they did not know what they thought they knew.
2. Then there was a review component, by which student misconceptions were
identified and corrected by an instructor, with students being guided toward ap-
propriate conceptions.
3. Students then reflected by making written statements on how they had previously
thought about the concepts. They were also expected to comment on any new, or
revised, or extended conceptual understandings that they were in the process of
developing.
4. Then followed a period of between 6 and 12 weeks when students revisited their
conceptual understandings. The intention, here, was to ensure that the students’
new understandings were appropriate, stable and coherent.
5. The final component was when students were assessed for retention. Without
students having been fore-warned, tests parallel to the original tests were admin-
istered, the aim being to see if the students had acquired and retained accurate
conceptions, and whether they could apply, with appropriate confidence, their
new understandings in relevant problem-solving or problem-posing contexts.
The 5-R intervention model is illustrated in Fig. 2, which is taken from Clements and
Ellerton (2009). Each of the five ordered treatment components—Realize, Review,
Reflect, Revisit and Retain—requires action on a learner’s part. As a consequence
Fig. 2 Schematic summary

of 5-R intervention model
(from Clements and Ellerton
2009)
of each intervention component, students are depicted (see Fig. 2) as reaching new
conceptual phases (or “C-phases”) in their development of relational understandings
that are internalized in the form of accurate, rich, linked concept images. The con-
ceptual phases are termed “Certainty (Misplaced),” “Confusion,” “Construction,”
“Conceptualization,” “Coherence,” and “Confidence.”
In their written reflections, many students noted that they had rarely encoun-
tered equations like x + 5 = 8 − (3 − x), and simply did not know what to do
when they arrived at a statement like “x + 5 = 5 + x.” Similarly, with respect
to 4(x + 1) = 4(x − 3), they said they had never before been asked to interpret a
proposition like “4 = −12.” However, after their thinking about statements such as
these had been identified and, when necessary, “straightened out,” most of the stu-
dents became confident that they would be able to deal with such problems in the
future. On the retention test they demonstrated that this confidence was well placed.
Pre-intervention and post-intervention/retention percentages correct for the 328
students on the 16 questions (parallel questions were used for post-instruction/re-
tention assessment) are shown in Table 5. For 15 of the 16 questions (Question 5
was the exception), the improvement was substantial. Although the study design did
not feature control groups, evidence that the 5-R intervention program was the main
factor influencing improvement is given in Clements and Ellerton (2009).
Some of the comments made by students in their written reflections provided
poignant commentary on standard approaches in school and college algebra. One
student was moved to reflect, directly, on what had happened to her in relation to
equations and inequalities:
One thing that stands out to me is the pre-test we took during the first class. Once we got
the test back graded and went over the answers I realized that I made a lot of stupid silly
mistakes. I didn’t recognize that x + 1 > x − 2 is always true and other similar problems.
Now, I actually look at the problem and analyze it to see if what the problem is asking
makes sense.
Another student wrote:

During middle school and high school, algebra was one of my favorite subjects. I liked
being able to follow rules and to easily find a solution to any problem. However, I now
know the reason I was good at algebra was because I could memorize and remember rules.
I had very little understanding to back up these rules and I had no idea why they worked.
This course has helped me to build this understanding and to explain why the rules worked.
Table 5 Percentages correct, for 328 mathematics teacher-education students on 16 pre-teaching

and 16 parallel retention tasks at pre-intervention and retention stages
Pre-Teaching Pre-Intervention, Retention Retention,
Question % Correct Question % Correct
(n = 328) (n = 328)
1 1
1. =3 82% =2 97%
x x
2. x = 9
2 23% x = 16
2 96%
9 4
3. x = 21% x= 88%
x x
4. x 2 + 6 = 0 21% x2 + 3 = 0 93%
5. 9(x − 1) = 0 98% 5(1 − x) = 0 97%
6. 4(x + 1) = 4(x − 3) 53% 3(x − 1) = 3(x + 3) 98%
7. (x − 3)(x − 2) = 0 59% (x + 3)(x + 4) = 0 95%
8. x + 5 = 8 − (3 − x) 22% 2x + 5 = 10 − (5 − 2x) 95%
1 1
9. > 4 1% >3 55%
x x
10. x 2 > 4 5% x2 > 9 77%
4 1
11. x > 0% x> 43%
x x
12. x 2 + 2 > 0 16% x2 + 5 > 0 89%
13. 4(x − 1) > 0 64% 4(x − 3) > 0 95%
14. 9(x + 1) > 9(x − 2) 23% 9(x + 5) > 9(x − 1) 97%
15. (x − 3)(x − 1) > 0 1% (x + 3)(x − 2) > 0 61%
16. x + 3 > 6 − (3 − x) 33% x + 2 > 7 − (5 − x) 96%
It has forced me to solve the problems on my own without relying on the formula. This way
of thinking gives meaning to problems rather than just mindlessly plugging numbers into a
formula in order to get an answer.
Another student wrote:

Once I got the results I was shocked at how low my score was. I had completely forgotten
the basic principles. These basic rules date back to middle-school years. It took me a while
to appreciate that, yes, my answers were not correct. I will not make these mistakes again.
So far as Carrie’s attempt to solve the quadratic equation was concerned, I simply did not
recognise Carrie’s mistakes. Although I now can see that she did not need to distribute the
terms in the parentheses at the start, I’m still not completely sure what she did wrong when
she wrote 0 × 0 = 0 in her check.
The final comment in this last statement draws attention to the fact that sometimes
the instructors had to talk with students individually to help them overcome miscon-
ceptions. The student who wrote the last excerpt was struggling to recognise why
the statement “0 × 0 = 0” in the check indicated a serious lack of understanding.
It was only after one of the instructors talked individually with the student that the
misconception was straightened out.
Student Confidence Considerations
On the pre-intervention test, for each of the eight equations, all 328 students indi-
cated that either they were certain they were right or they thought they were right.
They were less confident on the inequalities, but usually the students thought that
they were right. For all eight inequality tasks, 293 of the 328 students indicated that
they were certain they were right, or they thought they were right, or they thought
they had a 50-50 chance of being correct. This initial confidence was usually mis-
placed, suggesting that many students “did not know that they did not know.”
On the post-intervention/retention tests, most students indicated that they were
confident they were correct on 14 of the 16 tasks. This time their confidence was
vindicated. The two exceptions were for 1/x > 3 and for x > 1/x. On the retention
test, only 91 of the 328 students indicated that they were certain their answers were
correct on these two tasks. In fact, 60 of those 91 students did give correct responses
to those tasks. Most of the other 237 students were not confident that their answers
were correct for the two tasks, and in most cases these students gave incorrect re-
sponses. At least, these students now knew that they did not know.
Concluding Comments
As stated earlier, our development of the clever tasks reported in this chapter was
assisted by two major factors. First, we had many years of experience in teaching
and researching school algebra and algebra courses for prospective teachers (e.g.,
Clements and Ellerton 2006), and that experience provided us with relevant peda-
gogical content knowledge. And, second, we had been directly associated with the
little-known research of Lim (2000) and Vaiyavutjamai (2004), and it was that re-
search which inspired not only the quadratic equation scenario but also our decision
to link carefully selected equations and inequalities as pairs, where in seven cases
out of eight, a pair was regarded as forming a composite clever task.
Using the eight equation/inequality pairs enabled us to gather, quickly, rich data
relating to the students’ algebra knowledge, skills and conceptions. Analyzing those
data placed us in a strong position to do something about helping the students to
take control of their situations, so that they could improve their knowledge, skills
and conceptions relating to linear and quadratic equations and inequalities.
The more we worked on this project, the more we came to realize how difficult
school algebra can be for many learners. Our work, using the 5-R model, has at
least helped us to get prospective middle-school teachers to the stage where they
have strong conceptions related to equations and inequalities, which of itself is an
achievement (de Castro 2004). Furthermore, they have come to know that they are
learning correct algebra, and that previously some of their algebra content knowl-
edge was inaccurate, even though they had believed otherwise. Our use of the clever
tasks was important in helping the prospective teachers upgrade their knowledge,
and to replace misplaced confidence with appropriate confidence so far as their al-
gebra knowledge was concerned.
References
Allen, F. B., Douglas, E. C., Richmond, D. E., Rickart, C. E., Swain, H., & Walker, R. J. (1965).
First Course in Algebra: Teacher’s Commentary, Part 1. Stanford, CA: School Mathematics
Study Group.
Blanco, L. J., & Garrote, M. (2007). Difficulties in learning inequalities in students of the first year
of pre-university education in Spain. Eurasia Journal of Mathematics Science & Technology
Education, 3, 221–229.
Boero, P., & Bazzini, L. (2004). Inequalities in mathematics education: The need for complemen-
tary perspectives. In M. J. Høines & A. B. Fuglestad (Eds.), Proceedings of the 28th Conference
Bergen, Norway: Bergen University College.
Booth, L. R. (1984). Algebra: Children’s Strategies and Errors. Windsor, Berks, UK: NFER-
Nelson.
Clement, J. (1982). Algebra word problem solutions: Thought processes underlying a common
misconception. Journal for Research in Mathematics Education, 13, 16–30.
Clements, M. A., & Ellerton, N. F. (2006). Improving prospective mathematics teachers’ under-
standing of the solution process for quadratic equations in the form (x − a)(x − b) = 0. In H.
S. Dhindsa, I. P. Cheong, I. J. Kyeleve, O. Chukwu, & J. S. H. Quintus Perera (Eds.), Shaping
the Future of Science, Mathematics and Technical Education (pp. 287–296). Gadong, Brunei
Darussalam: Universiti Brunei Darussalam.
Clements, M. A., & Ellerton, N. F. (2009). A model for improving prospective mathematics teach-
ers’ algebra content knowledge. Brunei International Journal of Science and Mathematics
Education, 1(1), 68–84.
Clements, M. A., Evans, K. M., Green, L. J., Smith, G. C., & Watterson, G. A. (1967). Mathematics
for Today and Tomorrow, Book 1. Melbourne, Australia: Sun Books.
de Castro, B. (2004). Pre-service teachers’ mathematical reasoning as an imperative for codified
conceptual pedagogy in algebra: A case study in teacher education. Asia Pacific Education
Review, 5(2), 157–166.
Dobbs, D., & Peterson, J. (1991). The sign-chart method for solving inequalities. Mathematics
Teacher, 84, 657–664.
Fujii, T. (2003). Probing students’ understanding of variables through cognitive conflict: Is the
concept of a variable so difficult for students to understand. In N. A. Pateman, B. J. Dougherty,
& J. Zilliox (Eds.), Proceedings of the 2003 Joint Meeting of PME and PMENA (Vol. 1, pp.
49–65). Honolulu: University of Hawai’i.
Hart, K. M. (Ed.) (1981). Children’s Understanding of Mathematics (pp. 11–16). London: Murray.
Karp, A. (2007). Interview with Henry Pollak. The International Journal for the History of Math-
ematics Education, 2(2), 67–89.
Kieran, C. (2004). The equation/inequality connection in constructing meaning for inequality sit-
uations. In M. J. Høines & A. B. Fuglestad (Eds.), Proceedings of the 28th Conference of
the International Group for the Psychology of Mathematics Education (Vol. 1, pp. 143–127).
Bergen, Norway: Bergen University College.
Kieran, C. (2007). Learning and teaching algebra at the middle-school through college levels:
Building meaning for symbols and their manipulation (A project of the National Council of
Teachers of Mathematics). In F. K. Lester (Ed.), Second Handbook of Research on Mathematics
Teaching and Learning (pp. 707–762). Charlotte, NC: Information Age Publishing.
Küchemann, D. E. (1981). Algebra. In K. M. Hart (Ed.), Children’s Understanding of Mathemat-
ics: 11–16 (pp. 102–119). London: John Murray.
Lim, T. H. (2000). The teaching and learning of algebraic equations and factorisation in O-level
mathematics: A case study. Unpublished M.Ed dissertation, Universiti Brunei Darussalam.
Lim, T. H., & Clements, M. A. (2000). Language and algebra in a Bruneian context. In M. A.
Clements, H. H. Tairab, & K. Y. Wong (Eds.), Science, Mathematics and Technical Education
in the 20th and 21st Centuries (pp. 155–164). Gadong, Brunei Darussalam: Universiti Brunei
Darussalam.
Linchevski, L., & Sfard, A. (1991). Rules without reasons as processes without objects—The case
of equations and inequalities. In F. Furinghetti (Ed.), Proceedings of the 15th Annual Meeting
for the International Group for the Psychology of Mathematics Education (Vol. 1, pp. 317–
324). Assisi, Italy: International Group for the Psychology of Mathematics Education.
MacGregor, M. E. (1991). Making Sense of Algebra: Cognitive Processes Influencing Comprehen-
sion. Geelong, Australia: Deakin University.
Mishanski, J. (2010, March 27). Step-by-step answers. Retrieved from http://algebrahelp.com.
Pimm, D. (1995). Symbols and Meanings in Mathematics. London: Routledge.
Schmidt, W., Tatto, M. T., Bankov, K., Blömeke, S., Cedillo, T., Cogan, L., Shin, I. H., Houang,
R., Feng, J. H., Paine, L., Santillan, M., & Schwille, J. (2007). The Preparation Gap: Teacher
Education for Middle School Mathematics in Six Countries (MT21 Report). East Lansing, MI:
Michigan State University Center for Research in Mathematics and Science Education.
Tsamir, P., & Bazzini, L. (2002). Algorithmic models: Italian and Israeli students’ solutions to
algebraic inequalities. In A. D. Cockburn & E. Nardi (Eds.), Proceedings of the 26th Conference
Norwich, UK: International Group for the Psychology of Mathematics Education.
Tsamir, P., Almog, N., & Tirosh, D. (1998). Students’ solutions to inequalities. In A. Oliver
& K. Newstead (Eds.), Proceedings of the 22nd Conference of the International Group for
the Psychology of Mathematics Education (Vol. 4, pp. 129–136). Stellenbosch, South Africa:
International Group for the Psychology of Mathematics Education.
Vaiyavutjamai, P. (2004). Factors influencing the teaching and learning of equations and in-
equations in two government secondary schools in Thailand. Unpublished PhD thesis, Uni-
versiti Brunei Darussalam.
Vaiyavutjamai, P., & Clements, M. A. (2006a). Effects of classroom instruction on students’ un-
derstanding of quadratic equations. Mathematics Education Research Journal, 18(1), 47–77.
Vaiyavutjamai, P., & Clements, M. A. (2006b). Effects of classroom instruction on student per-
formance on, and understanding of, linear equations and linear inequalities. Mathematical
Thinking and Learning, 8(2), 113–147.
Vaiyavutjamai, P., Ellerton, N. F., & Clements, M. A. (2005). Students’ attempts to solve two ele-
mentary quadratic equations: A study in three nations. In P. Clarkson, A. Downton, D. Gronn,
M. Horne, A. McDonagh, R. Pierce, & A. Roche (Eds.), Building Connections: Research,
Theory and Practice (pp. 735–742). Sydney, Australia: Mathematics Education Research
Group of Australasia.
The Algebraic Nature of Fractions: Developing
Relational Thinking in Elementary School
Susan B. Empson, Linda Levi, and Thomas P. Carpenter
Abstract The authors present a new view of the relationship between learning
fractions and learning algebra that (1) emphasizes the conceptual continuities be-
tween whole-number arithmetic and fractions; and (2) shows how the fundamen-
tal properties of operations and equality that form the foundations of algebra are
used naturally by children in their strategies for problems involving operating on
and with fractions. This view is grounded in empirical research on how algebraic
structure emerges in young children’s reasoning. Specifically, the authors argue that
there is a broad class of children’s strategies for fraction problems motivated by the
same mathematical relationships that are essential to understanding high-school al-
gebra and that these relationships cannot be presented to children as discrete skills or
learned as isolated rules. The authors refer to the thinking that guides such strategies
as relational thinking.
The research summarized in this chapter was supported in part by a grant from the National
Science Foundation (ESI-0119732) and a Faculty Research Assignment from the University of
Texas. The opinions expressed in this chapter do not necessarily reflect the position, policy, or
endorsement of the National Science Foundation or UT. The authors would like to thank Jeremy
Roschelle, Walter Stroup, Scott Eberle, and Brian Katz, along with two anonymous reviewers, for
very helpful reviews of this chapter as it was being written and Annie Keith and her students for
helping us to understand the potential for developing algebraic thinking.
S.B. Empson ()
College of Education, University of Texas at Austin, Austin, USA
L. Levi
Cognitively Guided Instruction (CGI) Professional Development Initiatives, Teachers
Development Group, Madison, USA
T.P. Carpenter
Department of Curriculum and Instruction, University of Wisconsin-Madison, Madison, USA

410 S.B. Empson et al.
Of what use . . . is it to be able to see the end in the beginning? (Dewey 1974, p. 345).
Fractions and algebra are two topics in school mathematics that are considered
critical to the curriculum and difficult to learn (National Council of Teachers of
Mathematics 1998, 2000). Students’ misconceptions and procedural errors for frac-
tions and algebra, for example, have been well documented (Kerslake 1986; Matz
1982; Sleeman 1984; Stafylidou and Vosniadou 2004). Moreover, high-school stu-
dents’ poor performance in algebra has been blamed on their weak proficiency in
fractions. According to a recent Math Panel report, for instance, the ability to per-
form fraction computations easily and quickly is one of the most critical prerequi-
sites for algebra (U.S. Department of Education 2008).
We see the relationship between fractions and algebra differently. If there is an
obstacle to learning algebra, it begins to form as children learn basic arithmetic.
As a direct result of typical approaches to instruction in the U.S., American stu-
dents tend to understand arithmetic as a collection of procedures, rather than in
terms of conceptual relationships or general properties of number and operation. By
the time the problem is exposed as children learn fractions, it is fairly entrenched,
and it is only exacerbated by the fact that fractions are taught in isolation from
whole numbers and that fraction operations are taught as a collection of proce-
dures. Concrete materials and models may help children make critical connections
(Lesh et al. 1987), but our take on the types of connections that are most fruit-
ful for understanding fractions represents a departure from earlier lines of think-
ing.
In this chapter we present an alternative view on the relationship between frac-
tions and algebra that (1) emphasizes the conceptual continuities between whole-
number arithmetic and fractions; and (2) shows how the fundamental properties
of operations and equality that form the foundations of algebra are used naturally
by children in their strategies for problems involving operating on and with frac-
tions. We ground this view in research on children’s thinking to illustrate how al-
gebraic structure emerges in young children’s reasoning and can, with the help of
the teacher, be made explicit. Specifically, we argue that there is a broad class of
children’s strategies for fraction problems motivated by the same mathematical re-
lationships that are essential to understanding high-school algebra and that these
relationships cannot be presented to children as discrete skills or learned as iso-
lated rules. We refer to the thinking that guides such strategies as relational think-
ing.
These arguments are based on our research over the last 14 years, in which we
have been studying how to provide opportunities for students to engage in rela-
tional thinking in elementary classrooms and how to use relational thinking to learn
the arithmetic of whole numbers and fractions. We have focused on understand-
ing children’s conceptions and misconceptions related to relational thinking, how
conceptions develop, how teachers might foster the development and the use of re-
lational thinking to learn arithmetic, and how professional development can support
the teaching of relational thinking. This research has included design experiments
with classes and small groups of children (e.g. Falkner et al. 1999; Empson 2003;
Koehler 2004; Valentine et al. 2004), case studies (Empson et al. 2006; Empson and
The Algebraic Nature of Fractions: Developing Relational Thinking in Elementary School 411
Turner 2006), and large-scale studies (Jacobs et al. 2007); and it has been synthe-
sized in two books (Carpenter et al. 2003; Empson and Levi 2011).
In this chapter, we illustrate elementary school children’s use of relations and
properties of operations as a basis for learning fractions and argue that relational
thinking is a critical foundation for learning algebra. We first define relational think-
ing and then we discuss how the use of relational thinking supports the development
of children’s understanding of arithmetic. At the same time we challenge the notion
that an invigorated focus on fractions in the middle grades is the key to equipping
students to learn algebra meaningfully (Hiebert and Behr 1988; U.S. Department of
Education 2008). Instead, we argue that the key can be found in helping children
to see the continuities among whole numbers, fractions, and algebra. Finally, we
suggest that a model of the development of children’s understanding of arithmetic
that is based upon a concrete to abstract mapping is too simplistic. We propose in-
stead that developing computational procedures based on relational thinking could
effectively integrate children’s learning of the whole-number and fraction arithmetic
in elementary mathematics, in anticipation of the formalization of this thinking in
algebra.
What Is Relational Thinking?
Relational thinking involves children’s use of fundamental properties of operations

and equality1 to analyze a problem in the context of a goal structure and then to
simplify progress towards this goal (Carpenter et al. 2003; see also Carpenter et
al. 2005; Empson and Levi 2011). The use of fundamental properties to generate a
goal structure and to transform expressions can be explicit or it can be implicit in
the logic of children’s reasoning much like Vergnaud’s (1988) theorems in action.
For example, to calculate 12 + 34 a child may think of 34 as equal to 12 + 14 and
reason that 12 plus another 12 is equal to 1, then plus another 14 is 1 14 . In a study by
Empson (1999), several first graders reasoned this way when given a story problem
involving these fractional quantities. This solution involves anticipatory thinking, a
construct introduced by Piaget and colleagues (Piaget et al. 1960) to characterize
the use of psychological structures to coordinate a goal with the subgoals used to
accomplish it; thinking can involve several such coordinations. These students rec-
ognized that they could decompose 34 into 12 + 14 , and that if they decomposed it
this way, they could regroup to add 12 + 12 . In other words, they transformed 34 to
2 + 4 in anticipation of adding 2 + 2 . This solution involved thinking flexibly about
1 1 1 1
3
both the quantity 4 and about the operation, taken into account concurrently rather
than separately as a series of isolated steps. Their thinking can be represented by the
1 Essentially, we are referring here to the field properties and basic properties of equality (Herstein
1996; see also Carpenter et al. 2003; Empson and Levi 2011).
following equalities:

1 3 1 1 1 1 1 1 1 1
+ = + + = + + =1+ =1 .
2 4 2 2 4 2 2 4 4 4
Although the first graders did not represent their reasoning symbolically in this way,
their solution is justified in part by the implicit use of the associative property of
addition, which we have represented explicitly here to highlight the logic of the
their thinking.
Relational thinking is powerful because the applicability of fundamental proper-
ties such as the associative property of addition and the distributive property of mul-
tiplication over addition cuts across number domains and into the domain of algebra
where one reasons about general quantities rather than specific numbers. Consider
the expression 7a + 4a. A basic algebraic skill is to simplify this expression to 11a,
by application of the distributive property of multiplication over addition:
7a + 4a = (7 + 4)a = 11a.
The same property that justifies this transformation can also be used to justify that
70 + 40 = 110 and 75 + 45 = 11
5 :
70 + 40 = 7 × 10 + 4 × 10 = (7 + 4) × 10 = 11 × 10 = 110,
7 4 1 1 1 1 11
+ = 7 × + 4 × = (7 + 4) × = 11 × = .
5 5 5 5 5 5 5
Yet addition of whole numbers and addition of fractions are taught in isolation from
each other in the elementary curriculum, and they are often taught by rote, without
reference either to the underlying properties or the process of deciding how and
when to use a property. For example, to add fractions children are taught to first find
a common denominator and then add the two numerators; many children remember
this process as a series of steps to execute. They are not encouraged to draw on
their understanding of the distributive property either to derive or to explain this
procedure. Many children are therefore simply not prepared later to explicitly draw
on the appropriate properties to justify why 7a + 4a is 11a, but 7a + 4b is not 11ab.
Children learn arithmetic with understanding when they are encouraged to use
and develop their intuitive understanding of the properties of number and operation.
Our research has led us to recast the meaning of learning with understanding in
terms of thinking relationally: to understand arithmetic is to think relationally about
arithmetic, because the coherence of operations on whole numbers and fractions is
found at the level of the fundamental properties of operations and equality. Teaching
arithmetic in general and fractions in particular primarily as a set of procedures
fails to introduce children to the powerful reasoning structures that form the basis
of our number system. On the other hand, if children enter algebra with a well
developed ability to think relationally about operations, they are prepared to learn
to reason meaningfully about and carry out transformations involving generalized
expressions through the explicit application of algebraic properties. In the following
section we show how these properties emerge and can be developed in the context of
carrying out number operations involving fractions and we discuss their connections
to learning algebra.
Use of Relational Thinking in Learning Fractions

Children’s difficulties learning fractions have been well documented (Kerslake
1986; Stafylidou and Vosniadou 2004). The difficulty, however, may be in how frac-
tions are taught rather than how intrinsically easy or hard they are to understand.
Indeed, a conclusion we draw from our research is that fractions are not unduly
difficult if instruction develops children’s capacity for relational thinking.
A focus on relational thinking can transform fractions into a topic that children
understand by drawing on and reinforcing the fundamental properties that govern
reasoning about both whole-number and fraction quantities and operations. Chil-
dren use relational thinking in their solutions to story problems (e.g., Baek 2008;
Carpenter et al. 1998; Empson et al. 2006) and open number sentences (Carpenter
et al. 2005). Teachers can cultivate children’s use of relational thinking by using
a combination of these types of problems. In this chapter we focus on children’s
relational thinking in the context of solving story problems.
Understanding Fractional Quantities Through Relational Thinking
Before children can learn to operate on or with fractions, they need to under-
stand fractional as quantities. Because a fraction is defined by the multiplicative
relationship between its two terms, a mature understanding of fractions as quan-
tities is relational in nature. Young children can construct a relational understand-
ing of fractions by solving and discussing Equal Sharing problems (Empson 1999;
Empson and Levi 2011; Streefland 1993).
To solve a problem about equally sharing quantities, such as two pancakes shared
among three children, children must partition the quantities equally and completely.
Children’s earliest, non-relational strategies often involve partitioning the pancakes
into halves. In this example the two pancakes would yield four halves. A child
using this strategy might then try to distribute the four halves into three groups.
When the child discovers that there is a half left over, the child may then parti-
tion the extra half into half, and then partition each of those parts into half again,
continuing until the parts get too small to partition. This solution is not relational
in that it lacks anticipatory thinking. The child knows that it is necessary to parti-
tion the pieces to share them, but approaches the problem one step at a time, par-
titioning into halves without anticipating how the resulting parts are going to be
shared.
Children begin to think relationally about fractional quantities when they begin
to reason about the relationship between partitions into equal and exhaustive shares
Fig. 1 Child’s strategy for

sharing 2 pancakes equally
among 3 children,
demonstrating emerging
relational understanding of
fractions
and the number of sharers. To solve two pancakes shared by three children, a child
could decide to completely share the first pancake with all three children, and then
to share the second pancake in the same way (Fig. 1). Alternatively, a child who
began by distributing one half to each person might then decide to partition the left
over half into three equal parts. In either case, a child who thinks about the number
of people sharing and at the same time how to partition the things to be shared is in
the process of developing a relational understanding of fractions.
These strategies implicitly use several important mathematical relationships. For
ease of illustration, we concentrate on the strategy in which the child partitions
each whole candy bar into thirds. Although young children are unlikely to use the
following notation to represent their reasoning, it follows this logic:
1 1 2
2 ÷ 3 = (1 + 1) ÷ 3 = 1 ÷ 3 + 1 ÷ 3 = + = .
3 3 3
This reasoning embodies the knowledge that three one-thirds make a whole pancake
and that one pancake divided among three people yields one-third of a pancake to
each. It also suggests an intuitive understanding of how a “distributive-like property”
can be applied to division.2
A fully operationalized and explicit understanding of fractions as relational quan-
tities develops gradually. Most basic to this understanding is that unit fractions are
created by division or partitioning and that unit fractions are multiplicatively related
to the whole:
1 1
1÷n= and × n = 1. (1)
n n
Multiple opportunities to combine unit-fraction quantities in solutions to Equal
Sharing problems and to notate these solutions—such as “1 third and 1 third equals
2 This property can be represented as a ÷ c + b ÷ c = (a + b) ÷ c, which is equivalent to a × 1c +

b × = (a + b) × 1c . It is sometimes referred to as the right distributive property of division over
1
c
addition. On the other hand, a ÷ (b + c)is not the same as a ÷ b + a ÷ c; that is, there is no left
distributive property of division over addition.
2 thirds”— lead to the following more generalized relational understanding:
1 m
m× = . (2)
n n
The conceptual connections between children’s pictorial and symbolic representa-
tions of fractional quantities require prolonged attention to develop in a flexible,
integrated way (Empson et al. 2006; Saxe et al. 1999). These relationships are ini-
tially grounded in children’s informal knowledge of partitioning quantities, such as
cupcakes and sandwiches. Children arrive at the generalized relational understand-
ing represented by (1) and (2) above as the result of repeated opportunities to create,
represent, and reason about these relationships in various interlinked forms over an
extended period of time.
Understanding these basic relationships is absolutely critical to children’s abil-
ity to reason with understanding about fraction operations and computations. Con-
sider the case of Holly, a fifth grader who had been exposed to fraction instruction
throughout her school career but did not understand fractions as relational quanti-
ties. She had learned that fractions involved partitioning wholes into parts, but she
did not understand the relation between the parts and the whole. Fractional parts, to
her, were entities unrelated to whole numbers. These limitations in her understand-
ing were exposed in her solution to the following problem:
Jeremy is making cupcakes. He wants to put 12 cup of frosting on each cupcake. If he makes
4 cupcakes for his birthday party, how much frosting will he use to frost all of the cupcakes?
To solve the problem, Holly drew the picture in Fig. 2 and decided the answer was
“four halves.” Upon further questioning, it became clear that Holly did not see how
these quantities could be combined; she insisted the answer was four halves and four
halves only. It seemed instead that the entire circle partitioned in half represented
the fraction 12 for Holly, and it would have been nonsensical to combine them (akin
to asking, “How much is 4 apples?”). For her, fractions existed separately from other
numerical measures.
Fig. 2 Holly’s written work for figuring four groups of half each, suggesting a non-relational
understanding of fractions. (The 4 over what looks like 12 is Holly’s way of writing 4 halves.
She appears to be trying to remember syntactic features of the numeral and confounding “12”
with “1/2”)
Contrast Holly’s solution to a third grader’s solution to the following problem.

Mr. W has 10 cups of frog food. His frogs eat 12 a cup of frog food a day.
How long can he feed his frogs before his food runs out?
Fig. 3 A third grader’s

written work for figuring 10
groups of one half each,
showing a relational
understanding of fractions as
quantities
The third grader, John, represented each cup of frog food with a rectangle, then
divided each rectangle in half and notated “ 12 ” on each half to show how much
food Mr. W’s frogs could eat in a day (Fig. 3). He then counted these to arrive at
an answer of 20 days. Unlike Holly, John used a relational understanding of the
quantity 2 × 12 = 1 to construct a solution. John’s solution represents a big step
forward over Holly’s. He might have gone further in his use of relational thinking
by grouping the half cups in order to figure the total number of days more efficiently.
For example, he could have reasoned that 2 half cups are one cup, 4 half cups are 2
cups and so on, until he reached the number of half cups in 10 whole cups. He also
could have reasoned directly that 20 groups of 12 are the same as 10 groups of 1.
This type of reasoning, which takes into account both a relational understanding
of fractional quantities and relations involving the operation of multiplication, is
illustrated in the cases in the following section.
Use of Relational Thinking to Make Sense of Operations Involving

Fractions
As children come to understand fractions as relational, they start to use this under-
standing to decompose and recompose quantities for the purpose of transforming
expressions and simplifying computations. These manipulations are done purpose-
fully and draw on (a) children’s intuitive understanding of fractional quantities as
relational described above and (b) children’s relational understanding of operations
cultivated in the context of whole-number reasoning and problem solving.
Children’s strategies for multiplication and division word problems involving
fractions can draw on and reinforce their growing understanding of the multiplica-
tive nature of fractions. At the same time, the use of such problems supports the
emergence of relational thinking about operations as children attempt to figure out
how to make operations more efficient. Children’s thinking becomes more antici-
patory in that they begin to make choices about how to decompose and recompose
fractions in the context of a goal structure that relates operations and quantities. This
Table 1 Combining groups using fundamental properties of multiplication

Equation representing Fundamental and other generalized properties of arithmetic
child’s thinking on which child’s thinking is based
8× 3
8 = 8 × (3 × 18 ) Fractions represented as multiples of unit fractions
= 8 × ( 18 × 3) Commutative property of multiplication
= (8 × 18 ) × 3 Associative property of multiplication
=1×3=3 Inverse and identity properties of multiplication
anticipatory thinking signals the purposeful use of fundamental properties of oper-

ations and equality and is in contrast with algorithmic thinking about operations in
which the goal structure can be summarized as “do next.”
A pivotal point in the growth of children’s understanding is reached when chil-
dren begin to use relational thinking to make repeated addition or subtraction of
fractions more efficient by applying fundamental properties of operations and equal-
ity in their strategies for combining quantities. The emergence of relational thinking
about operations in this context is facilitated by the need to combine several groups
of equal size. For example, in one of the cases that follows, a fifth-grade student
wanted to figure eight groups of three eighths each. The child reasoned that eight
groups of one eighth each equals one, so three such groups would be three. This
reasoning makes implicit use of the commutative and associative properties of mul-
tiplication (Table 1).
As children’s understanding of fractions grows, basic relationships as illustrated
in Table 1 serve as building blocks in more sophisticated relational thinking strate-
gies. These strategies draw upon a variety of these properties in ways that are an-
ticipatory rather than algorithmic and in ways that demonstrate a well connected
understanding of number and operation. Most notably, these strategies are driven
by each child’s understanding and therefore cannot and should not be reduced to a
generalized series of steps for all children to follow. In fact, a teacher would be hard
pressed to explicitly teach these strategies, because each step is embedded in a goal
structure that is specific to each child’s relational understanding of the operations
and quantities for a given problem. In the long run, this relational understanding of
number and operations results in an efficiency in learning advanced mathematics,
such as algebra.
To illustrate the types of relational thinking that elementary students are capable
of using, we discuss two strategies generated by fifth and sixth graders in different
classrooms. The teachers in these classrooms tended to place responsibility for gen-
erating and using conceptually sound strategies on each individual student.3 This
approach to instruction does not typify instruction in U.S. classrooms, and so the
strategies we describe here are not representative of the current performance of U.S.
children in the upper elementary grades (e.g., Hiebert et al. 2003). However, they are
3 We have observed patterns both in the types of relational thinking used and how it develops, which
are beyond the scope of this chapter (see Empson and Levi 2011).
representative of the types strategies that evolve in classrooms such as these—even

if these classrooms are rare—and provide a study of the possibility of integrating
fractions and algebra in the upper elementary grades.
Each problem involved division with a remainder to be taken into account in the
quotient. For each case, we describe the strategy and then note how children used
fundamental properties of operations and equality in their solutions.
Case 1: Measurement division.
The first case comes from a combination fourth- and fifth-grade class, working
on the following measurement division problem:
It takes ____ of a cup of sugar to make a batch of cookies. I have 5 12 cups of sugar. How
many batches of cookies can I make?
The students were given a variety of number choices for the divisor. In order of
difficulty, these choices were 12 , 14 , 34 , and 38 . Several students, including Jill, chose
to work with 38 of a cup of sugar.
Jill began her strategy by drawing upon the basic multiplicative relationship de-
scribed above to generate familiar groupings of three eighths that would simplify
the calculation (Table 1). She said she knew that 8 three-eighths would be 3, which
meant that 4 three-eighths would be half that much, or 1 12 , and 12 three-eighths
would therefore be 4 12 (Fig. 4). At this point, she knew that she needed only 1 more
cup to use up all 5 12 cups. Again Jill used the relationship between 38 and 3 as a
reference point. She said that because 8 three-eighths was 3, a third as many three-
eighths would be a third as much, or 1. That is, ( 13 × 8) × 38 is 1, and 13 × 8 is 83 or
2 23 . She concluded that she could make a total of 12 + 83 batches, which would be
equal to 14 32 batches.
If we unpack Jill’s description of her solution, we see that it involved setting
subgoals that were readily solved using familiar relations. The solution of one sub-
Fig. 4 Jill’s written work for

her strategy to solve 5 12
divided by 38 , suggesting
implicit use of fundamental
properties of operations and
equality
goal provided a springboard for the next. Fundamental properties of operations and
equality were implicit in the solution of each of the subgoals. Jill’s ultimate goal
was to find how many 38 cups it would take to make 5 12 cups. She started with an
overarching view of the problem that facilitated the formulation of a series of sub-
goals. The 5 21 cups could be partitioned into parts that would be easily divided by 38 .
Then the parts could be combined.
Jill’s first subgoal was to identify a multiple of 38 that would give her a whole
number that she might subsequently use as a building block to find how many 38
cups it took to make 5 12 cups. Drawing implicitly on the kind of thinking described
in Table 1, she started with the equation 8 groups of 38 is 3.
Because she had only accounted for 3 of the 5 12 cups of flour in the problem, Jill
now had to find how many 38 cups it took to make 2 21 cups. She recognized that she
could use the equation involving 8 groups of 38 to make another 1 12 cups and that
would leave exactly one cup to deal with. Essentially she used the multiplicative
property of equality and the associative property of multiplication to transform the
equation 8 × 38 = 3 as follows:

1 3 1
× 8× = × 3,
2 8 2

1 3 1
×8 × =1 ,
2 8 2
3 1
4× =1 .
8 2
The next subgoal was to find how many 38 cups it took to make the remaining one
cup. Jill also used the equation 8 × 38 = 3 as the basis for addressing this subgoal.
She again used the multiplicative property of equality and the associative property
of multiplication to transform the core equation as shown below.
3
8× = 3,
8

1 3 1
× 8× = × 3,
3 8 3

1 3
× 8 × = 1,
3 8
8 3
× = 1.
3 8
Note Jill might have simply used the reciprocal relation between 83 and 38 for this
calculation, but she continued to build off of the equation 8 × 38 = 3. Although we
believe it is likely that she did not intend to generate the reciprocal relationship be-
tween 83 and 38 , we find its emergence here significant, because it illustrates how
algebraic relationships can emerge fairly naturally in the context of children’s rela-
tional reasoning. Problems such as this one provide experience with this relation.
Finally, Jill put the parts together using the additive property of equality and the
distributive property.
3 3 8 3 1 1
8× +4× + × =3+1 +1=5
8 8 3 8 2 2
and

3 3 8 3 8 3
8× +4× + × = 8+4+ ×
8 8 3 8 3 8
2 3
= 14 × .
3 8
The kinds of thinking implicit in Jill’s strategy are directly related to the kinds
of thinking that are involved in solving algebra problems with meaning. She started
with a primary goal—to find how many 38 -cups were in 5 12 cups—which subse-
quently guided the formulation of her subgoals. To make progress on solving the
problem, she transformed the primary goal into a series of subgoals for which she
had a ready solution. This practice is fundamental to high-school algebra in which a
series of properties of operations and equality are often used to simplify a complex
equation. For example, to solve a linear equation in one unknown, students set sub-
goals that entail finding successively simpler equations that are closer to the goal of
finding an equation of the form x = a number. As in the above example, the sub-
goals are met by repeated application of fundamental properties of operations and
equality. Similarly, the goal of solving a quadratic equation is transformed into sub-
goals of solving simple linear equations by applying a corollary of the zero property
of multiplication (a × b = 0, if and only if a = 0 or b = 0).
To address each of the subgoals, Jill essentially constructed and transformed re-
lationships of equality using fundamental properties of operations and equality in
ways that were strikingly similar to the thinking used in constructing and solving
equations in formal treatments of algebra. She drew on anticipatory thinking in
transforming the equations into equations that could be put together to solve the
problem. In other words she consistently constructed and transformed equations in
ways that brought her closer to the solution of the basic problem. Again, that is
essentially what solving algebra equations is all about.
Case 2: Partitive division.
Our second case involves a sixth-grade boy, Keenan, who solved the following
problem:
Two thirds of a bag of coffee weighs 2.7 pounds. How much would a whole bag of coffee
weigh?
This problem involves partitive division and differs from the previous division
problem in that the goal is to find out how much per group rather than to find out how
many groups. Keenan’s strategy included the transformation of quantities for the
Fig. 5 Keenan’s strategy to

solve 2.7 divided by 23 ,
suggesting implicit use of
fundamental properties of
operations and equality
purpose of simplifying calculations as well as the flexible use of several fundamental

properties of operations and equality (Fig. 5).
To start, Keenan recognized that the problem was a division problem and wrote
2.7 ÷ 23 . He remarked, “Two divided by 23 is going to be really easy, all I really need
to worry about is the seven tenths. Seven tenths divided by 23 isn’t easy to think
about so [long pause] if I make them both thirtieths, it would be easier.” He notated
his thinking so that it read:
21 20
÷
30 30
and then said, “21 thirtieths divided by 20 thirtieths is just the same as 21 divided
by 20 which is one and one twentieth.” He notated his answer:
21 20 1
÷ =1 .
30 30 20
Keenan then said, “Now all I have to do is 2 divided by 23 , which is 3.” When asked
how he knew that so quickly he said, “2 divided by 13 would be 6 since you have 3
groups of 13 in each 1, so 2 divided by 23 would be 3 since 23 is twice as big as 13 .” He
then extended his notation as follows:
2 1 2 1 1
2.7 ÷ =1 +2÷ =1 +3=4 .
3 20 3 20 20
This strategy incorporates several instances of relational thinking. Keenan began
by decomposing 2.7 into .7 + 2. This choice involved anticipatory thinking in that
he analyzed the problem to see what relationships he might draw upon to simplify
his calculations, rather than simply begin to execute a series of steps to solve the
problem. He used the commutative property of addition and a “distributive-like”
property to simplify the division. Although he did not notate this step, his thinking
could be represented as:
2 2 2 2 2
2.7 ÷ = (2 + .7) ÷ = (.7 + 2) ÷ = .7 ÷ + 2 ÷ .
3 3 3 3 3
Of note is his correct use of this distributive-like property for division. This division
relationship is generalizable and can be justified with the distributive property of
multiplication over addition in conjunction with the inverse relationship between
multiplication and division (see footnote 2).
Next Keenan facilitated the computation of .7 ÷ 23 by transforming .7 into 21

30 and
2
3 into 20
30 and then using these transformed quantities as follows:
2 21 20 1
.7 ÷ = ÷ = 21 ÷ 20 = 1 .
3 30 30 20
Again, Keenan used anticipatory thinking to produce equivalent fractions for the
purpose of simplifying the division.
Keenan then computed 2 ÷ 23 . This computation appeared to be routine for him;
however, he justified it as follows:

2 1 1 1
2÷ =2÷ 2× =2÷ ×2 = 2÷ ÷2
3 3 3 3
with an associative-like property of division. Again he used a generalizable princi-

ple, similar to the distributive-like principle he used above, that could be justified
with formal properties but is rooted in a relational understanding of fractional quan-
tities and division.
Like Jill, Keenan had a unified view of the entire problem and its parts. This view
allowed him to set subgoals to address the parts individually with the understanding
that the answers to the problems addressed by these subgoals could be reassembled
into the whole. As was the case with Jill, Keenan drew on anticipatory thinking and
a fluid understanding of how expressions and equations could be transformed. Once
again the parallels with the kind of thinking used in symbolic treatments of algebra
are striking.
Discussion of Cases
The strategies used by these two elementary aged children to divide fractions illus-
trate the power of relational thinking and its algebraic character. Children’s thinking
in these examples was anticipatory in that their strategies were driven by a goal
structure premised on relational thinking. These strategies contrast with the goal
structure in the execution of standard algorithms as they are typically learned which
can be summarized as “do next.”
Further, the thinking displayed by these children resembles the “competent rea-
soning” that proficient mathematical thinkers use to compare rational numbers, as
reported by Smith (1995). Based on an analysis of 30 students’ solutions to order
and equivalence problems, Smith argued that competent reasoning is characterized
by the use of strategies that exploit the specific numerical features of a problem
and often apply only to a restricted class of fractions. These strategies were reliable
and efficient. He contrasted this reasoning with the use of generalized, all purpose
strategies—such as conversion to a common denominator to compare fractions—
which the students in his study tended to use as a last resort.4 These findings led
Smith to conclude “the analysis of skilled reasoning with rational numbers should
. . . move beyond a focus on particular strategies to examine the character of the
broader knowledge system that has those strategies as components” (p. 38). We are
proposing that this system consists of children’s informal algebra of fractional quan-
tities and that it is expressed in children’s relational thinking.
From this perspective, Jill and Keenan represent students who are in the process
of developing a knowledge base for reasoning about fractions in ways that can be
characterized as proficient and competent. Moreover, this knowledge base is inte-
grated with properties of whole-number operations and relations and anticipates the
algebra of generalized quantities, typically taught in the eighth or ninth grade. We
are not suggesting that standard algorithms for fraction operations have no place in
developing fluency with number operations. Proficient thinkers use them when they
see no way to exploit the number relationships in a problem. However, we are ar-
guing that when children are supported to develop relational thinking in elementary
school, their knowledge of generalized properties of number and operation becomes
explicit and can serve as a foundation for learning high-school algebra in ways that
mitigate the development of mistakes and misconceptions.
A Conjecture Concerning Relational Thinking as a Tool in

Learning New Number Content
Jill and Keenan used fundamental properties of operations and equality and other
notable relationships, such as 38 × 8 = 3, as tools in their strategies to divide frac-
tions. The use of these relationships was coordinated within a goal structure and is
a hallmark of relational thinking. In this section we discuss a critical and perhaps
surprising implication of a focus on relational thinking in the elementary curriculum
with respect to the role of other types of tools such as concrete materials and mod-
els in facilitating the development of children’s understanding of number operations
involving fractions (as well as decimals and integers—which are beyond the scope
of this chapter).
Some approaches to teaching for understanding emphasize the use of concrete
materials such as base-ten blocks or fraction strips to model abstract relation-
ships (e.g., Van de Walle 2007). The use of such materials has at times been
seen as a universal remedy to children’s difficulties in understanding mathemat-
ics. Several studies have shown, however, that concrete materials alone are insuffi-
cient at best and at worst, ineffective (Brinker 1997; Resnick and Omanson 1987;
4 This approach to numerical reasoning is not unique to children. Dowker (1992) reported that
professional mathematicians prefer to approach computational estimation in the same ways, that
is, by exploiting specific numerical features of a problem rather than using a generalized algorithm
that works in all cases.
Uttal et al. 1997). In a review of this research, Sophian (2007) noted that manipula-
tives are symbols themselves and how they map to mathematical notation and pro-
cesses is best appreciated by those who already understand the mapping (p. 157).
Teachers can show students how to manipulate these materials to perform calcula-
tions involving fractions just as they show students how to manipulate symbols to
perform calculations. Some students may remember steps involving materials more
easily than they remember symbolic algorithms, but in neither case are they neces-
sarily reasoning about the relationships involved in each step or more globally in the
problem.
In contrast, when manipulatives and other types of models are used as tools to
think with, rather than to simply generate an answer, they can play a critical role in
the development of children’s understanding (Carpenter and Lehrer 1999; Koehler
2004; Martin and Schwartz 2005). For example, the images that children create and
reason about as they partition quantities such as cupcakes and sandwiches in their
solutions to story problems can help children conceptualize fractions in terms of
basic relationships such as 23 = 2 × 13 (Empson and Levi 2011).
Keeping in mind this valuable use of visualizing tools, we propose a shift in rel-
ative emphasis as the curriculum turns to advanced number operations: Children’s
use of relational thinking can and should drive the development of new content
and concrete materials and models should be used to support the emergence of re-
lational thinking. Jill’s and Keenan’s strategies for division of fractions described
above cannot be mapped in any straightforward way onto the manipulation of con-
crete materials and so do not appear to represent an abstraction of their operations
on concrete materials. Instead, these strategies (1) incorporated a relational under-
standing of fractions and (2) were planned and executed (sometimes in an emergent
sense) on the basis of each child’s understanding of fundamental properties of oper-
ations and equality.
One fairly popular way to introduce fraction multiplication, for example, is by
using an area model (Izsák 2008). Its advantages are that it is generalizable—it can
be used to multiply any two fractions—and it is “concrete” so children can “see”
the multiplication. To multiply 14 × 23 using this model, for example, a rectangular
unit is divided into thirds and two of the thirds are shaded. Then the rectangle is
divided into fourths orthogonally to the original partition into thirds. Based on this
partitioning, one fourth of the rectangle is shaded. The intersection of the shaded
parts (Fig. 6) represents the product of one fourth and two thirds. The model might
be used to develop understanding of multiplication of fractions, but the use of this
model is easily proceduralized, especially if it is introduced before children have
had opportunities to make and integrate relational connections between quantities
and operations. (See teachers’ own difficulties with the proceduralization of this
model, reported in Izsák 2008.)
Using relational thinking, children might approach a problem such as this one in
any number of ways employing strategies that involve the application of generalized
properties of arithmetic. For example, a child could reduce the calculation to oper-
ating on unit fractions, by applying the distributive property of multiplication over
addition. A child might say, “a quarter of 13 is 12
1
so a quarter of 23 has to be 12
1
plus
Fig. 6 Area representation of

the product of 14 × 23
1 2
12 which is 12 .”
This thinking could be formally represented:

1 2 1 1 1 1 1 1 1 1 1 2
× = + = × + × = + = .
4 3 4 3 3 4 3 4 3 12 12 12
A child could also transform the calculation into an easier one through an implicit
use of the associative property of multiplication. The reasoning might be “a quarter
of 13 is 12
1
so a quarter of 23 has to be 12
1 2
times 2 which is 12 :”

1 2 1 1 1 1 1 2
× = × ×2 = × ×2= ×2= .
4 3 4 3 4 3 12 12
These strategies represent the same types of relational thinking that we saw in
children’s strategies for division and mirror the types of relational thinking that
children use in whole-number multiplication (Baek 2008). They are two examples
of possible strategies that are driven by relational thinking instead of the potentially
rote use of a concrete model. With some experimentation, the reader should be able
to generate strategies for the multiplication of any two fractions that incorporate the
same fundamental properties of operations and equality and are robust and general-
izable.
In summary, our conjecture is that if instruction is focused on developing rela-
tional thinking with whole numbers throughout the early grades, the role of concrete
materials in introducing and developing understanding of operations on fractions
and decimals will likely change. Concrete materials would be used to support the
development of relational thinking rather than simply as tools to calculate answers
or justify algorithms. Further, encouraging children to construct and use procedures
based on relational thinking would help them to integrate learning number opera-
tions across different number domains.
Conclusion
The kinds of activity and thinking illustrated in this chapter are not isolated exam-
ples, and they do not represent mathematics that should be reserved for only a lim-
ited number of students or as supplementary enrichment (Carpenter et al. 2003). The
results of a recent study by Koehler (2004) document that young children of a wide
range of abilities are able to learn to think about relations involving the distribu-
tive property and that instruction that focuses on relational thinking as illustrated in
these examples supports the learning of basic arithmetic concepts and skills.
In this chapter we have argued that a focus on relational thinking can address
some of the most critical perennial issues in learning fractions with understanding.
One of the defining characteristics of learning with understanding is that knowledge
is connected (Bransford et al. 1999; Carpenter and Lehrer 1999; Greeno et al. 1996;
Hiebert and Carpenter 1992; Kilpatrick et al. 2001). Not all connections, however,
are of equal value, and we propose that our conception of relational thinking can
sharpen mathematics educators’ conceptions of what learning with understanding
looks like. Students who engage in relational thinking are using a relatively small
set of fundamental principles of mathematics to establish relations. Thus, relational
thinking can be seen as one way of specifying the kinds of connections that are pro-
ductive in learning with understanding. We have presented several such connections
made by children in elementary grades in the context of generating strategies for
problems involving multiplication and division of fractions.
We have further argued that relational thinking is a critical precursor—perhaps
the most critical—to learning algebra with understanding, because if children under-
stand the arithmetic that they learn, then they are better prepared to solve problems
and generate new ideas in the domain of algebra. However, relational thinking is
almost entirely neglected in typical U.S. classrooms with the unfortunate result that
children experience all types of learning difficulties as they move beyond arithmetic
into learning algebra. Some proposed solutions focus on a renewed emphasis on
prerequisite skills (e.g., U.S. Department of Education), while others emphasize the
use of concrete materials and models (e.g., Lesh et al. 1987). We have presented
an alternative view of how to address these difficulties that centers on cultivating
children’s implicit use of fundamental properties of the real-number system to solve
arithmetic problems, to better align the concepts and skills learned in arithmetic and
algebra. At the heart of this view is the reciprocal relationship between arithmetic
and algebra as it is revealed in children’s reasoning about quantity.
References
Baek, J. (2008). Developing algebraic thinking through explorations in multiplication. In C. E.
Greenes & R. Rubenstein (Eds.), Seventieth Yearbook: Algebra and Algebraic Thinking in
School Algebra (pp. 141–154). Reston, VA: National Council of Teachers of Mathematics.
Bransford, J. D., Brown, A. L., & Cocking, R. R. (Eds.) (1999). How People Learn: Brain, Mind,
Experience, and School. Washington, DC: National Academy Press.
Brinker, L. (1997). Using structured representations to solve fraction problems: A discussion of
seven students’ strategies. Paper presented at the annual meeting of the American Educational
Research Association Annual Meeting, March, Chicago, IL.
Carpenter, T. P., & Lehrer, R. (1999). Teaching and learning mathematics with understanding. In
E. Fennema & T. A. Romberg (Eds.), Mathematics Classrooms that Promote Understanding
(pp. 19–32). Mahwah, NJ: Erlbaum.
Carpenter, T. P., Franke, M. L., Jacobs, V., Fennema, E., & Empson, S. B. (1998). A longitudinal
study of invention and understanding in children’s use of multidigit addition and subtraction
procedures. Journal for Research in Mathematics Education, 29(1), 3–20.
and Algebra in the Elementary School. Portsmouth, NH: Heinemann.
Carpenter, T. P., Franke, M. L., Levi, L., & Zeringue, J. (2005). Algebra in elementary school:
Developing relational thinking. ZDM, 37(1), 53–59.
Dewey, J. (1974). The child and the curriculum. In R. Archambault (Ed.), John Dewey on Educa-
tion: Selected Writings (pp. 339–358). Chicago: University of Chicago Press.
Dowker, A. D. (1992). Computational estimation strategies of professional mathematicians. Jour-
nal for Research in Mathematics Education, 23, 45–55.
Empson, S. B. (1999). Equal sharing and shared meaning: The development of fraction concepts
in a first-grade classroom. Cognition and Instruction, 17(3), 283–342.
Empson, S. B. (2003). Low-performing students and teaching fractions for understanding: An in-
teractional analysis. Journal for Research in Mathematics Education, 34, 305–343.
Empson, S. B., & Levi, L. (2011). Extending Children’s Mathematics: Fractions and Decimals.
Portsmouth, NH: Heinemann.
Empson, S. B., & Turner, E. E. (2006). The emergence of multiplicative thinking in children’s
solutions to paper folding tasks. Journal of Mathematical Behavior, 25, 46–56.
Empson, S. B., Junk, D., Dominguez, H., & Turner, E. E. (2006). Fractions as the coordination of
multiplicatively related quantities: A cross-sectional study of children’s thinking. Educational
Studies in Mathematics, 63, 1–28.
Falkner, K. P., Levi, L., & Carpenter, T. P. (1999). Children’s understanding of equality: A foun-
dation for algebra. Teaching Children Mathematics, 6, 231–236.
Greeno, J. G., Colliins, A. M., & Resnick, L. (1996). Cognition and learning. In D. C. Berliner &
R. C. Calfee (Eds.), Handbook of Educational Psychology (pp. 15–46). New York: Macmillan.
Herstein, I. N. (1996). Abstract Algebra. Hoboken, NJ: Wiley.
Hiebert, J., & Behr, M. (1988). Introduction: Capturing the major themes. In J. Hiebert & M.
Behr (Eds.), Number Concepts and Operations in the Middle Grades (pp. 1–18). Hillsdale, NJ:
Lawrence Erlbaum.
Hiebert, J., & Carpenter, T. P. (1992). Learning mathematics with understanding. In D. Grouws
(Ed.), Handbook of Research on Mathematics Teaching and Learning (pp. 65–97). New York:
Macmillan.
Hiebert, J., Gallimore, R., Garnier, H., Givvin, K. B., Hollingsworth, H., Jacobs, J. et al. (2003).
Teaching mathematic in seven countries: Results from the TIMSS 1999 video study (NCES
2003-013). Washington, DC: U.S. Department of Education, National Center for Education
Statistics.
Izsák, A. (2008). Mathematical knowledge for teaching fraction multiplication. Cognition and In-
struction, 26, 95–143.
Jacobs, V., Franke, M. L., Carpenter, T. P., Levi, L., & Battey, D. (2007). Professional develop-
ment focused on children’s algebraic reasoning in elementary school. Journal for Research in
Mathematics Education, 38, 258–288.
Kerslake, D. (1986). Fractions: Children’s Strategies and Errors. A Report of the Strategies and
Errors in Secondary Mathematics Project. England: NFER-Nelson Publishing Company, Ltd.
Kilpatrick, J., Swafford, J., & Findell, B. (2001). Adding It Up: Helping Children Learn Mathe-
matics. Washington, DC: National Academy Press.
Koehler, J. (2004). Learning to think relationally and using relational thinking to learn. Unpub-
lished doctoral dissertation, University of Wisconsin-Madison.
Lesh, R., Post, T., & Behr, M. (1987). Representations and translations among representations in
mathematics learning and problem solving. In C. Janvier (Ed.), Problems of Representation in
the Teaching and Learning of Mathematics (pp. 33–40). Hillsdale, NJ: Lawrence Erlbaum.
Martin, H. T., & Schwartz, D. (2005). Physically distributed learning: Adapting and reinterpreting
physical environments in the development of fractions concepts. Cognitive Science, 29, 587–
625.
Matz, M. (1982). Towards a process model for school algebra errors. In D. Sleeman & J. S. Brown
(Eds.), Intelligent Tutoring Systems (pp. 25–50). New York: Academic Press.
National Council of Teachers of Mathematics (1998). The Nature and Role of Algebra in the K-14
Curriculum. Washington, DC: National Academy Press.
Piaget, J., Inhelder, B., & Szeminska, A. (1960). The Child’s Conception of Geometry (E. A. Lun-
zer, Trans.). New York: Basic.
Resnick, L., & Omanson, S. (1987). Learning to understand arithmetic. In R. Glaser (Ed.), Ad-
vances in Instructional Psychology (Vol. 3, pp. 41–96). Hillsdale, NJ: Lawrence Erlbaum.
Saxe, G., Gearhart, M., & Seltzer (1999). Relations between classroom practices and student learn-
ing in the domain of fractions. Cognition and Instruction, 17, 1–24.
Sleeman, D. H. (1984). An attempt to understand students’ understanding of basic algebra. Cogni-
tive Science, 8, 387–412.
Smith, J. P. (1995). Competent reasoning with rational numbers. Cognition and Instruction, 13(1),
3–50.
Sophian, C. (2007). The Origins of Mathematical Knowledge in Childhood. New York: Lawrence
Erlbaum Associates.
Stafylidou, S., & Vosniadou, S. (2004). The development of students’ understanding of the numer-
ical value of fractions. Learning and Instruction, 14, 503–518.
Streefland, L. (1993). Fractions: A realistic approach. In T. P. Carpenter, E. Fennema, & T.
Romberg (Eds.), Rational Numbers: An Integration of Research (pp. 289–325). Hillsdale, NJ:
Lawrence Erlbaum.
United States Department of Education (2008). The Final Report of the National Mathematics
Advisory Panel. Washington, DC: Author.
Uttal, D. H., Scudder, K. V., & DeLoache, J. S. (1997). Manipulatives as symbols: A new perspec-
tive on the use of concrete objects to teach mathematics. Journal of Applied Developmental
Psychology, 18, 37–54.
Valentine, C., Carpenter, T. P., & Pligge, M. (2004). Developing concepts of proof in a sixth-
grade classroom. In R. Nemirovsky, A. Rosebery, J. Solomon, & B. Warren (Eds.), Everyday
Day Matters in Science and Mathematics: Studies of Complex Classroom Events (pp. 95–118).
New York: Routledge.
Van de Walle, J. (2007). Elementary and Middle School Mathematics: Teaching Developmentally
(6th ed.). Boston: Pearson.
Vergnaud, G. (1988). Multiplicative structures. In J. Hiebert & M. Behr (Eds.), Number Concepts
and Operations in the Middle Grades (pp. 141–161). Hillsdale, NJ: Lawrence Erlbaum.
Professional Development to Support Students’
Algebraic Reasoning: An Example
from the Problem-Solving Cycle Model
Karen Koellner, Jennifer Jacobs, Hilda Borko, Sarah Roberts,

and Craig Schneider
Abstract The Problem-Solving Cycle (PSC) model of mathematics Professional

Development (PD) seeks to enhance teachers’ knowledge and skills in a variety of
domains. In this paper we consider how participating in the PSC program, with a
specific focus on algebra, impacted one teachers’ instructional practice. We explore
The Professional Development program featured in this chapter is one component of a larger
research project entitled Supporting the Transition from Arithmetic to Algebraic Reasoning
(STAAR). STAAR is supported by NSF Proposal No. 0115609 through the Interagency
Educational Research Initiative (IERI). The views shared in this chapter are ours, and do not
necessarily represent those of IERI. Portions of this chapter were presented at the American
Educational Research Association Annual Meeting, San Francisco, 2006. The authors would like
to acknowledge Eric Eiteljorg’s role in our many years of discussions about this case study, and
his early efforts to get these analyses off the ground. We also thank Ken Bryant for his enthusiastic
participation in our PD, and his unwavering commitment to improving mathematics teaching and
learning. This chapter is a revised version of an article published in ZDM—International Reviews
on Mathematical Education, 37(1), 43–52. DOI 10.1007/BF02655896.
K. Koellner ()
School of Education, Hunter College, New York, NY, USA
J. Jacobs · C. Schneider
J. Jacobs
C. Schneider
H. Borko
School of Education, Stanford University, Stanford, USA
S. Roberts
Department of Curriculum and Instruction, Iowa State University, Ames, USA

430 K. Koellner et al.
the nature of change in the teacher’s classroom, as well as the connection between
these changes, the focus of our PD, and the teacher’s intentions. Our analyses are
based on a set of videotaped classroom lessons collected over two years, along with
interviews and written reflections. We conclude that there was a close match be-
tween the teacher’s personal goals for improvement and our program goals, and no-
table shifts in his algebra instruction that was likely to have supported his students’
algebraic reasoning.
Introduction
Researchers agree that mathematics instructional practices geared to meet the needs
of all learners look very different from the traditional approach of lecturing and
teaching algorithms through a rote memorization process (Kaput 2007). Such teach-
ing takes into account the complexities of supporting mathematical reasoning. It re-
quires, first and foremost, careful attention to students’ thinking, which then guides
instructional decisions built on students understanding, such as selecting appropriate
tasks, promoting mathematical communication, and considering multiple represen-
tations.
Attention to the needs of all learners is particularly important for the do-
main of algebra. In the United States, there is an increasingly widespread com-
mitment to ensuring that all students successfully complete a course in algebra
before entering high school (National Mathematics Advisory Panel 2008). Stu-
dents’ difficulties in learning formal algebra are well documented (Kieran 2007;
National Research Council 1998), yet many teachers continue to use teaching meth-
ods that focus on rote memorization and algorithmic approaches to solving algebra
problems without supporting the development of deeper procedural and conceptual
understanding (Ball et al. 2001).
Moses and Cobb (2001) offer important insights into instructional approaches
that provide access to algebra for all students. For example, they reported that for
students to find algebra valuable and engaging and to participate in discussions, it
is critical that the teacher select tasks that are relevant to their lives and have more
than one solution strategy. Other researchers similarly encourage teachers to foster
classroom discussions, particularly in their algebra lessons—including small group,
whole class, and partner discussions (Winicki-Landman 2001). Structuring activi-
ties to encourage students to talk with their peers can elicit productive conversations
about unique representations and multiple solution strategies (Cnop and Grandsard
1998). Setting as instructional priorities the selection of appropriate tasks, the pro-
motion of student-driven communication, and the encouragement of multiple repre-
sentations all have been suggested as ways to provide students with a better grasp
of algebraic concepts (NCTM 2000).
Despite the increased emphasis on algebra in K-12 education, it appears that
most teachers have weak and fragile understandings of how to teach algebra ef-
fectively (Even 1993; Nathan and Koedinger 2000; Hadjidemetriou and Williams
Professional Development to Support Students’ Algebraic Reasoning 431
2002). Thus, not surprisingly, the enhancement of teachers’ professional knowl-

edge about algebra and the teaching of algebra is widely seen as a key compo-
nent of the effort to support students’ algebraic reasoning (Blanton and Kaput 2005;
Kieran 2007; NCTM 2000). The Problem-Solving Cycle (PSC) model of Profes-
sional Development (PD), which is the focus of this chapter, builds on that premise;
its central aim is to help teachers enhance their professional knowledge and instruc-
tional practices to support all students’ development of algebraic reasoning.
The Problem-Solving Cycle Model of Professional Development
The PSC was developed and implemented as part of the Supporting the Transi-
tion from Arithmetic to Algebraic Reasoning (STAAR) project.1 Our major fo-
cus was on strengthening teachers’ professional knowledge of central algebra
concepts, and helping them explore ways of fostering their students’ algebraic
thinking. A situative perspective on cognition and learning provides the concep-
tual framework that guided the design of the PSC. Scholars within a situative
perspective argue that knowing and learning are constructed through participa-
tion in the discourse and practices of a community, and are shaped by the con-
texts in which they occur (Greeno 2003; Lave and Wenger 1991) With respect
to PD, situative theorists focus on the importance of creating opportunities for
teachers to work together on improving their practice, and locating these learn-
ing opportunities in the everyday practice of teaching (Ball and Cohen 1999;
Putnam and Borko 2000).
Three design principles derived from a situative framework are central to the PSC
model: establishing a professional learning community, using video from teachers’
own classrooms to provide a meaningful context for learning, and establishing com-
munity around video. For more extensive discussions of our conceptual framework
and design principles, please see Borko et al. (2005, 2008), Jacobs et al. (2007),
Koellner et al. (2007).
The PSC consists of a series of three interconnected workshops in which teachers
engage in a common mathematical and pedagogical experience, organized around
a rich mathematical task (see Fig. 1). This common experience provides a structure
within which teachers can build a supportive community that encourages reflection
on mathematical understandings, student thinking, and instructional practices.
Workshop 1 focuses on the mathematics content knowledge needed to teach the
focal problem. The majority of time is devoted to teachers collaboratively solv-
ing the problem, debriefing the mathematics in their solution strategies, and mak-
ing connections between solutions. In addition, teachers develop lesson plans for
teaching the problem to their students. Between Workshop 1 and 2, each teacher is
videotaped implementing his or her lesson.
1 The PD program is one component of STAAR and was partially supported by NSF Proposal No.
0115609 through the Interagency Educational Research Initiative (IERI).

Fig. 1 The problem-solving

cycle model of professional
development
The central purpose of Workshop 2 is to help teachers consider instructional skills

and strategies associated with problem-based teaching (Lampert 2001). In particu-
lar, teachers are guided to think deeply about the role they played in teaching the
selected problem to their students. Activities are designed around a specific peda-
gogical topic and associated video clips from one or more of the teachers’ lessons.
Workshop 3 addresses skills and strategies that comprise the core practice of
“learning about student understanding” (Grossman et al. 2009). Teachers consider
ways to elicit, attend to, and build on students’ mathematical understandings by
studying video clips and other artifacts from their lessons that depict student think-
ing. By considering the various forms of mathematical reasoning their students ap-
plied to the problem and the different solution strategies they used, teachers have
the opportunity to gain further insight into the mathematical concepts entailed in
the problem and students’ learning of those concepts.
The PSC as Implemented in the STAAR Project
Although the PSC model is intended to be applicable to any domain of mathematics,

as part of the STAAR project our content focus was on algebra. Our research and
PD program as part of the STAAR project began in the summer of 2003 with a 2-
week algebra institute. Over the next 2 academic years we held monthly, full-day
PD workshops (7 workshops per year). In Fall 2003 we held workshops focused on
pedagogical practices associated with algebra, and in Spring 2004 we conducted the
first iteration of the PSC using a rich algebraic problem. We conducted two more
iterations during the 2004–05 academic year with an algebraic focus.
Eight middle school mathematics teachers who attended the summer institute
then took part in the program during the 2003–04 academic year. In 2004–05, seven
teachers continued working with us and three additional teachers joined the pro-
gram. Each new teacher was a colleague of one of the current participants.
We had explicit mathematical, instructional, and process foci for each iteration
of the PSC. Our mathematical foci can be seen in the problems we selected. In all
3 iterations, we used problems adapted from Driscoll’s (1999) book Fostering Alge-
braic Thinking: A Guide for Teachers, Grades 6–10 (see Table 1). The three prob-
Table 1 Three iterations of the problem-solving cycle
Workshop 1: Doing the Problem and Planning Workshop 2: Analyzing the Teacher’s Role Workshop 3: Analyzing Student Thinking
the Lesson
Iteration 1: Problem Statement A cube with edges of Teachers watched video clips from 3 lessons Teachers looked at students’ posters from one
Painted length 2 centimeters is built from centimeter and discussed issues related to: lesson and discussed the different solution
Cubes cubes. If you paint the faces of this cube and strategies, ways of thinking, and mathemati-
then break it into centimeter cubes, how many • Launching the lesson cal errors.
cubes will be painted on three faces? How • Closing the lesson Teachers watched a video clip from 1 lesson
many will be painted on two faces? On one • Issues related to language and discussed issues related to:
face? How many will be unpainted? What if • What questions or probes might you use to
the edge has a length different from 2? What move the student forward • Helping students through mathematical
if the length of the large cube is 3 cm? 50 cm? misconceptions
n cm? • Developmental probes to push student
Math Focus: Finding patterns, identifying and thinking
comparing different types of functions (linear,
quadratic, cubic, constant) using multiple rep-
resentations, recursion, and variables.
Iteration 2: Problem Statement: Snakes have colored rings Teachers watched 6 video clips from several Teachers looked at students’ posters from sev-
Snakes in that follow a changing pattern. As snakes grow lessons and discussed issues related to: eral lessons and discussed where their think-
Snakewood older, the pattern extends in a systematic way. ing appeared to be on the right or wrong track.
It starts with a white ring. Then a black ring • deciding when to provide explanations and Teachers watched 4 video clips from several
develops in the middle of the white ring. In ask leading questions, lessons and discussed issues related to student
Professional Development to Support Students’ Algebraic Reasoning
the next stage, the same thing happens with • considering students’ reactions to teachers’ thinking, including:
each white ring, but the black ring stays the comments and questions,
same. What do the 4th and 5th stages look like? • letting students follow their own line of rea- • how students move from recursive patterns
What is the relationship between the numbers soning. to formal rules,
of white and black rings in each stage? How • student misconceptions.
can you find the number of white and black
rings at any stage?
Math focus: Identifying patterns, recursion,
and exploring quadratic functions using mul-
tiple representations.
433
434
Table 1 (Continued)
Workshop 1: Doing the Problem and Planning Workshop 2: Analyzing the Teacher’s Role Workshop 3: Analyzing Student Thinking
the Lesson
Iteration 3: Problem Statement: A building is 12 stories Teachers watch 3 clips and considered the Teachers view 3 clips and considered the fol-
Skyscraper high and is covered entirely by windows on teacher role and following issues: lowing issues:
Windows all 4 sides. Each floor has 38 windows on it.
Once a year, all windows are washed. The cost • teacher questioning • analysis of student solution strategies
for washing the windows is $2.00 for each • listening to student thinking • pre-requisite knowledge
first-floor window, $2.50 for each second-floor • appropriate mathematical questions to • assessment and finding evidence of particu-
window, $3.00 for each third-floor window, move students forward from where they lar student conceptions
and so on. How much will it cost to wash the started
windows of this building? What if the building
is 30 stories tall? n stories tall?
Math focus: Identifying patterns, making con-
nections to strings of numbers, exploring
quadratic functions, and recursion.
Note: All problems are adapted from Driscoll (1999)

K. Koellner et al.
lems highlighted a variety of critical topics related to algebraic reasoning includ-

ing describing and generalizing patterns and functional relationships, differentiating
among functions, assessing the notion of equality, using multiple representations to
model and solve problems, and recognizing and generating multiple solution strate-
gies. Additionally, the problems all met the following criteria: (1) address multiple
mathematical concepts and skills, (2) are accessible to learners with different lev-
els of mathematical knowledge, (3) have multiple entry and exit points, (4) have an
imaginable context, (5) provide a foundation for productive mathematical commu-
nication, and (6) are both challenging for teachers and appropriate for students.
Our instructional foci can be most clearly identified by considering our second
and third workshops. As Table 1 shows, Workshop 2 discussions focused on ana-
lyzing instructional practices such as launching and closing the lesson to support
student learning; posing questions to move students forward in their thinking; de-
ciding when to provide explanations, ask leading questions, and let students follow
their own line of reasoning; and listening to students’ thinking. Workshop 3 discus-
sions centered on analyzing student thinking; they addressed core practices such as
comparing students’ solution strategies, and understanding and working with stu-
dents’ preconceptions and misconceptions. These instructional foci provided a lens
when selecting artifacts of practice (including video clips and student work) to an-
chor discussions throughout the workshops.
Our process foci emphasized the importance of establishing and maintaining a
safe environment for communication, explaining and justifying one’s own thinking,
and actively processing one another’s ideas (Borko et al. 2005; Clark et al. 2005).
With each successive iteration of the PSC, as the teachers’ content knowledge and
analytic skills increased and their community strengthened, we encouraged them to
probe more deeply into relevant and challenging ideas (Borko et al. 2008).
Prior Research on the Development and Impact of the PSC
As part of the STAAR project, we utilized a design experiment approach (Cobb

et al. 2003; Design-Based Research Collective 2003) to study and refine the PSC
model. We collected and analyzed a large amount of data on processes involved in
developing and enacting the model, and on the impact of the PD experience on the
teachers’ professional knowledge and instructional practices.
Several of our initial analyses focused on the processes of developing and enact-
ing the model. For example, we explicated the specific knowledge strands that are
foregrounded during the three PSC workshops, and the opportunities STAAR partic-
ipants had to expand their professional knowledge within each workshop (Koellner
et al. 2007). Additionally, we analyzed discussions around video and found that
the teachers talked in an increasingly focused, in-depth, and analytic manner about
specific issues related to teaching and learning (Borko et al. 2008).
We also examined project’s impact by analyzing teacher interviews and written
reflections. These self-report data indicate a strong impact on specific, targeted areas
of the participants’ professional knowledge, including: mathematics content (e.g.,

the importance of working on rich tasks and generating multiple solution strategies),
methods for improving classroom discourse (e.g., how to conduct group work and
foster conversations about mathematics), and ways of encouraging student think-
ing (e.g., giving students authority, building on students’ thinking, using tasks that
promote student thinking) (Jacobs et al. 2007).
Impact of the PSC on Instructional Practice: A Case Study

Analysis
In this chapter, we consider how participating in the PSC program, with a specific
focus on algebra, impacted one teachers’ instructional practice. Specifically, our
research questions focus on the nature of change in the teacher’s classroom during
the period of time that he participated in our PD program, as well as the connection
between these changes, the focus of our PD, and the teacher’s intentions with regard
to instructional change.
The focus of our case study analysis is Ken Bryant.2 We selected Ken in order
to provide a detailed example of the experiences and practices of one teacher who,
while not necessarily typical of all the teachers who participated in the STAAR
project, demonstrates the PSC’s potential for supporting teachers’ instructional
change. An in-depth analysis of one teacher’s movement towards improving his
practice such as this may be a useful contribution to understanding the factors that
foster change in mathematics teaching. As Shulman (1983) noted, “For the prac-
titioner concerned with process, the operational detail of case studies can be more
helpful than the more confidently generalizable virtue of a quantitative analysis of
many cases.” (p. 495)
Methods
Ken Bryant
Ken Bryant had taught both 5th and 6th grade for 3 years at the time our PD program
commenced. He taught sixth grade mathematics for both years of our PD program.
Ken worked in a K-8 school in a medium-sized, suburban school district. Minority
students comprise approximately half of the district’s population, and over a quarter
of the students are eligible for free or reduced lunch. At Ken’s school, the percent-
ages of minority students and students eligible for free or reduced lunch are similar
to those of the district. Ken had completed a Master’s degree in Education, and near
the end of our program, began working towards his principal’s license.
2A pseudonym.
Throughout our study, Ken maintained a consistent and outspoken desire to learn
and was deeply committed to improving his understanding of mathematics content
and instruction. Near the beginning of the program, he reflected on both his increas-
ing content knowledge (“I’m being overwhelmed with AH-HA moments”) and his
motivation to improve his instruction (“I would like to know how to ask questions
that get groups to talk about math, not recess”). At the end of the first year, Ken’s
continuing desire to improve his practice was evident in his written reflections about
what he had learned. His list included “the need for change in my practice” and “how
to have a critical perspective on the actions in my classroom.”
Ken maintained a perfect attendance record at our PD activities for the two years
of the project. During an interview at the conclusion of the project he remarked that
not only did the STAAR program change his practice, but it inspired him to talk
with his principal about implementing some of the elements of the PSC model at
his school. Shortly after the conclusion of our PD and research program, Ken won
a “teacher of the year” award presented by his district. The fact that our research
group had no part in nominating or judging teachers for this award suggests that his
efforts to improve were noticed and respected by colleagues in his school and across
the district.
Data Sources
The primary data source for this study is the set of videotapes of Ken’s classroom
instruction. As part of the PD program we videotaped each teacher conducting a
lesson using the problem selected for each iteration of the PSC. We also videotaped
additional lessons at regular intervals during the 2003–04 and 2004–05 school years.
We always filmed using two cameras; one with a wireless mike attached to the
teacher, and the other with a table mike near a group of students. In this way we
were able to document the teachers’ instructional moves as well as small group
student activities.
In Ken’s case, we videotaped 14 mathematics lessons over two years: 10 in 2003–
04 and 4 in 2004–05. For the analyses in this case study, we categorized these
lessons into three groups:
(1) lessons taught at the beginning of the PD program, prior to initiation of the PSC
model (pre-PSC, n = 5)
(2) lessons taught as part of the PSC PD, using the focal problems (PSC, n = 3),
and
(3) lessons taught during the time frame of the PSC, but that were not part of the
PD program (post-PSC, n = 6)
We conducted audiotaped interviews with the teachers following each videotaped
lesson, asking them to reflect on various aspects of the lesson. We also interviewed
the teachers on a number of other occasions during the PD program including the
beginning of the program, the end of each academic year, the conclusion of the
program, and as a post-program follow-up. We collected written reflections as fi-

nal activities in many of the PD workshops. Both the interviews and reflections
prompted teachers to describe their experiences in the PD program, consider their
pedagogical goals, and reflect on the impact of the program on their instructional
practices. Ken’s interviews and written reflections served as secondary sources for
our case study analysis.
Data Analysis
As noted above, we drew upon multiple data sources to analyze the impact of the
PSC on Ken’s instructional practices. We also used various analytic techniques. Our
primary set of analyses entailed coding Ken’s videotaped lessons, and then search-
ing for patterns in the codes for the three sets of lessons (pre-PSC, PSC, and post-
PSC). To help interpret these patterns, we examined Ken’s interviews and written
reflections, identifying comments that related to coded features of the videotaped
lessons. In addition, we use a vignette analysis to look deeply into a small por-
tion of one of Ken’s algebra lessons in order to illustrate particular aspects of his
teaching. The specific procedures for these three analytic techniques (quantitative
coding, supplementary use of interviews and reflections, and vignette analysis) are
described below.
Coding Ken’s Videotaped Lessons
While it is a fairly straightforward matter to videotape a teacher on a regular basis,

it is no small feat to create a reliable and valid instrument that taps into the precise
nature of the instructional practice documented on tape. Our research team used the
Learning Mathematics for Teaching (LMT) project’s recently developed instrument,
Quality of Mathematics Instruction (QMI) (LMT 2006). The QMI enables the anal-
ysis of videotaped classroom observations with respect to how teachers use their
mathematical knowledge in the classroom. We decided on this instrument for both
theoretical and pragmatic reasons.
From a theoretical perspective, the QMI is a good fit with our research goals and
methods. Specifically, the intentions of the developers matched our intentions for
using the instrument, and the constructs measured by the instrument overlap with
those that our research project intended to tap into. Our research provides something
of a “test” case for the QMI by implementing it with a group of coders distinct
from, but in consultation with, the instrument developers. At the same time, because
we used the QMI to track the instructional patterns in one teacher’s lessons over
time, we can attest only to the instrument’s adaptability to this context, not to its
generalizability across teachers or correlation with other measures.
Pragmatically speaking, the QMI was immediately available and has established
psychometric properties. LMT researchers have used the QMI to code 90 elemen-
tary and middle school mathematics lessons. They established adequate reliability
and are conducting a series of analyses to assess the validity of the instrument;
preliminary results are promising (LMT 2006). In addition, the LMT researchers
provided a detailed coding glossary, offered a workshop on use of the QMI to our
project team, and were available for ongoing consultation as we adapted the instru-
ment and worked to achieve reliability.
Applying the QMI to Ken’s lessons, we selected a subset of codes that most
closely matched our research questions and PD goals. For example, we did not apply
codes that described the mathematical content of the lesson because, in Ken’s case,
the content was largely determined by his district guidelines and was not a critical
issue for our research agenda. In a few cases, we were unable to establish adequate
reliability for codes that initially appeared promising.
Establishing Interrater Reliability
Our method of coding closely matches that used by the instrument developers (LMT
2006). Coders first parsed each videotaped lesson into 5-minute segments. Within
each segment, coders noted whether the actions indicated by each code were present.
Following the recommendation of the LMT researchers, we coded in pairs; each
pair included at least one mathematics educator. We used a multi-step procedure, as
follows:
1. Four members of our research team coded a lesson separately and then met to
discuss disagreements and points of confusion, modify the coding category, and
reconcile our coding decisions. We did this for several lessons, until we devel-
oped a workable coding scheme.
2. To achieve reliability, the two members of a pair coded 3 lessons separately and
reconciled their coding. They compared their agree-upon coding decisions to
those of another pair and calculated the percentage of agreement between the
pairs. Reliability of at least 80% was reached on all of the codes we report on in
this chapter.
3. Each remaining lesson was coded by one of the pairs.
Supplementing with Interviews and Reflections
Several members of the research team carefully examined transcripts from all of
Ken’s interviews and copies of his written reflections, to identify instances in which
Ken referred to topics captured by the QMI codes. In particular, we looked for com-
ments Ken made about his instructional goals, intentions, or changes in these areas.
We incorporated a representative selection of these comments into the results sec-
tion, in order to provide Ken’s own interpretations of his classroom behaviors.
Vignette Analysis
The QMI allowed us to look at general trends across Ken’s lessons, and not content-
specific trends. Because our PD was focused on algebra, and the three PSC lessons
that Ken conducted were all algebra lessons, a closer examination of those lessons
seemed warranted in order to understand how Ken supported students’ algebraic
reasoning. We focused on Ken’s final PSC lesson, in order to gauge the sorts of
learning opportunities he provided at the end of the PD program. From this les-
son, we identified a relatively short teaching episode that incorporates many of the
features captured by the QMI that are aligned with our PD program goals.
We constructed a vignette based on this episode with the intention of capturing
specific details of Ken’s practice while preserving the complexity and richness of the
classroom context (Miles and Huberman 1994). The vignette provides an example
of many of the QMI codes “in action” and also depicts what Ken’s lessons looked
like after two years of participation in the PD. Within the vignette, we summarize
the events that took place during this episode, include transcripts of conversations
that highlight important mathematical and instructional elements, and provide an
analysis of the events with respect to both our coding and our PD goals.
Results and Discussion
Patterns Drawn from QMI Coding and Analysis
We discuss patterns in Ken’s lessons in three areas: (1) knowledge of mathematics

for teaching, (2) core practices for problem-based teaching, and (3) core practices
for eliciting and building on student thinking. Within each of these areas, we con-
sider the QMI codes that depict patterns of teaching that map onto the goals of our
PD program, in both expected and unexpected ways. It is important to recall that the
subset of QMI codes we selected matched our PD focus and represent practices in
which we anticipated changes.
When interpreting the coded data, we looked for two types of patterns that we
hypothesized might occur: differences between Ken’s pre-PSC lessons and post-
PSC lessons, and differences between Ken’s PSC and non-PSC lessons (pre-PSC
and/or post-PSC). We did not run inferential statistical tests; rather we present raw
averages and caution readers to consider these data in light of our small sample size.
Knowledge of Mathematics for Teaching
A primary focus of our PD program, and foregrounded in Workshop 1 of the PSC,

is expanding teachers’ knowledge of algebra for teaching. However, identifying
changes in knowledge by analyzing videotaped lessons presents a formidable chal-
lenge. We elected to use two of the QMI codes—the teacher’s use of technical and
general mathematical language—to delve into this topic. Segments of Ken’s lessons
were coded as including technical language when he accurately used mathemati-
cal terms and concepts such as “variable” or “function.” Segments were coded as
including general language when Ken expressed mathematical ideas and concepts
Fig. 2 Average percentage of

segments that included
technical language and
general language
using care and precision in his language such as discussing how many dollars are
needed “for each” window.
In Ken’s PSC lessons, a higher percentage of segments contain technical and
general mathematical language (61 and 77 percent of segments, on average, respec-
tively) compared to his non-PSC lessons (see Fig. 2). The fact that technical and gen-
eral mathematical language occurred more frequently in Ken’s PSC lessons than his
non-PSC lessons is in line with our PD goal of increasing teachers’ understanding of
key mathematical concepts, specifically in the domain of algebra (Borko et al. 2005;
Jacobs et al. 2007), and may be due to our explicit focus on the algebraic concepts
underlying the PSC problems in the PD workshops conducted prior to participants’
teaching of each problem. The expectation was that, as a result of this focus, teach-
ers’ mathematical grounding in the problem would be very solid when they imple-
mented it in their classrooms.
Using appropriate mathematical language was something that Ken was cognizant
of, at least when teaching a PSC lesson. In an interview after teaching his first PSC
lesson (in February 2004), Ken remarked, “Something that I’ve been saying a lot
lately is that algebra is much like a language, and they need to learn how to read it
and write it and speak it.” Ken’s focus on language often was in the service of help-
ing his students to explain and justify their reasoning, even if the students struggled
with the appropriate mathematical words in their own explanations. One strategy
that he attempted to use in these situations was to paraphrase or revoice student
ideas. Ken noted:
I try to paraphrase things for them. [For example,] ‘This is what you said and this is what
I got from that. Is that right?’ And maybe I shouldn’t do that because I might be twisting
their words around a little bit, but my hope is that [by paraphrasing] other people understand
what that one person is trying to say.
The fact that Ken’s language in non-PSC lessons did not change as substantially
over time suggests that without highly focused PD work on selected problems, he
is likely to use a similar amount of technical and general mathematical language in
his lessons on a day-to-day basis as he did prior to participation in the PD program.
Core Practices of Problem-Based Teaching
Another primary focus of our PD program, and foregrounded in Workshop 2 of the

PSC, is problem-based teaching. Two aspects of problem-based teaching that we
highlighted, and that are captured by several codes on the QMI instrument, are in-
structional format and instructional practices. Within instructional format, we coded
for (1) the organization of Ken’s lessons into whole group, small group, and indi-
vidual activities and (2) the allocation of time to reviewing, introducing a topic,
working on a task, and closing the lesson. Within instructional practices we coded
for Ken’s use of mathematical representations.
Instructional Format Throughout the PD program, teachers’ solving of algebra

problems was largely done in small groups, and then solutions were shared with
the whole group. We intentionally strived to establish sociomathematical norms in
which authority resided in the community of teachers rather than with the PD facili-
tator (Clark et al. 2005). From the onset of the program, Ken was strongly impacted
by the process of doing mathematics in small groups, and he vowed to incorpo-
rate more groupwork in his own lessons. At the end of the 2-week summer algebra
institute that initiated the PD program, Ken reflected:
I began to rethink my teaching study. . .Usually I keep all the kids in rows separated and
all facing the front. . . To me it seems that I have the fewest discipline problems [this way],
especially with kids talking while I am talking. Then it occurred to me that [in this PD
program] we are talking nearly all the time. But we are talking about math! Fascinating!
How can I get this kind of atmosphere in my class? Well this I haven’t quite figured out yet
but the goal is there and that is what I am going to shoot for this fall. (July 2003)
Over the two-year period that Ken took part in our PD program, he gradually
reduced the amount of time his students spent working as a whole class (see Fig. 3).
At the beginning of the PD (pre-PSC) on average 76% of Ken’s lesson segments
included whole group work. During Ken’s PSC lessons, on average 70% of the time
Note: Percentages do not add to 100% because the categories are not mutually exclusive. If there
was a shift in categories during a 5-minute segment, coders marked both categories as present.
Fig. 3 Average percentage of segments that included whole group, small group, and individual
work
was spent in whole group work, and in Ken’s post-PSC lessons, on average, 63% of
the time was spent in whole group work.
Corresponding to the drop in whole group time, Ken increased the amount of
time his students worked in small groups. In Ken’s PSC lessons, almost half of
the lesson segments included small group work. The amount of time students spent
working individually remained the same in Ken’s pre-PSC and PSC lessons, but
dropped roughly in half during his post-PSC lessons.
Ken’s shift from having his students sit in rows with “all students facing the white
board” to having them sit in groups happened in the beginning of the 2003 school
year, immediately following the STAAR summer algebra institute. Ken recognized
that groupwork initially was difficult for some students: “Working in groups has
been something that’s been pretty tough because the kids haven’t been used to doing
that in mathematics.” It was at this time that Ken decided to focus on establishing
group norms in his classroom. He explained in an October 2003 interview, “I try to
really stress the importance of working together and cooperating with the people that
are in your group. And I really try to stress that everyone in a group is a resource.”
Establishing norms and instructional practices aligned with a more student-
centered class was a gradual process. In interviews throughout the school year, Ken
consistently talked about his desire for all groups to be engaged and discussing
mathematics. In January 2004 he experimented with grouping students based on
their test scores, creating groups with one strong student who could serve as a leader.
He commented in an interview that month, “[The students] kind of understand that
they have a responsibility when they’re a leader. They need to lead the discussion
and get started by asking questions or whatnot. That is something we’re working
on.”
By they end of the PD program, Ken used groups in increasingly unique ways
to promote in-depth, student-led conversations. For example, in his final PSC les-
son, Ken walked around the room listening carefully to each group’s ideas. He then
suggested that two of the groups have a conversation about their strategies for de-
riving the formula (i.e., determining an incremental growth pattern). In an interview
shortly after the lesson, Ken explained, “[The groups] were totally going about [the
problem] in different ways, yet they had nearly identical formulas in the end. So
this seemed to be a good time to bring the two groups together.” In this way, Ken
promoted across-group conversations and processing of multiple strategies within
small group work time.
Ken also made a substantial shift in the allocation of time during his mathematics
lessons. In his pre-PSC lessons, the vast majority of class time was spent reviewing
(or “warming up” or going over homework) (see Fig. 4). As he watched his video-
tapes, Ken noted with displeasure the time he devoted to reviewing and set a goal
that students spend more time working on mathematical problems. Specifically, he
wanted to spend at least half of his lesson time providing students with the opportu-
nity to work in small groups.
Almost three-quarters (72%) of Ken’s pre-PSC lesson segments, on average,
were devoted to reviewing. By contrast, 11% of the segments, on average, involved
introducing the major task and 27% included student work time. No lesson segments
were coded as closure or providing a summary of the mathematics.
Note: Percentages do not add to 100% because the categories are not mutually exclusive. If there
was a shift in categories during a 5-minute segment, coders marked both categories as present.
Fig. 4 Average percentage of segments that included review, introducing, work time, and closure
Videotapes of Ken’s PSC lessons clearly indicate a shift away from reviewing
toward a greater focus on the mathematical work of the lesson. In those lessons his
review time dropped to 27%, on average, whereas the time he spent introducing the
problem, having students work on the problem, and closing the lesson all increased
markedly.
In his post-PSC lessons, Ken spent more time reviewing than in his PSC lessons,
but less than in his pre-PSC lessons. Ken spent a relatively short amount of time
introducing problems, but he maintained a strong commitment to providing students
with time to work on the problems, and he devoted more time to closing the lessons.
These two changes in Ken’s practice—having students work in groups and pro-
viding more student work time—are closely related and are aligned with the format
of our PD workshops, particularly when teachers solved mathematics problems. Af-
ter two years in the PD program Ken reflected:
I felt like I learned so much and I was able to make huge gains just from working with people
and really having a lot of time [to solve problems]. And it wasn’t just sit in a class, listen
to a lecture, [and] get some practice problems for homework, like I’ve been teaching for
so long. It really changed my own idea about how to deliver instruction for math. Giving
discovery time, instead of just having someone model [how to solve the problem.] (June
2005)
Reflecting again a year later, he explained:

Cooperative grouping, in general, I didn’t do it before [the STAAR program]. I thought I
did, but it was more of a seating arrangement than anything else. And now, I think it’s letting
kids learn in their groups. Kind of backing off from guided instruction and letting them do
more discovery in their groups. (Spring 2006)
Instructional Practices One aspect of instructional practice that was emphasized

in our PD is the importance of using mathematical representations such as manipula-
tives, visuals, tables and figures. Four codes on the QMI instrument capture various

selecting representations,
deploying representations,
multiple representations, and
making links
aspects of how representations can be used in mathematics lessons. We found a dis-

tinct contrast between Ken’s PSC lessons and his non-PSC lessons on all four codes
(see Fig. 5). The PSC lessons included a higher percentage of segments, on aver-
age, in which Ken appropriately selected representations, deployed representations,3
used multiple representations, and made links between various representations.
In his first PSC lesson in February 2004, Ken had students use cubes to model
the problem situation. In an interview after the lesson, he reflected on the value of
students using representations saying, “[The cube] is this thing that they are holding
onto and working with. This object is giving the mathematics meaning.” In addition
to using cubes, Ken expected that creating a table would help students organize their
data, more easily notice patterns, and ultimately generalize the patterns as equations.
He explained,
I was hoping that they would come to that realization that they needed a table. I ended
up showing the class [how to make a table] because otherwise they would have just
kept building bigger and bigger cubes without being able to organize the information
systematically. . . . I wanted them to be able to hold their understanding [in the table.]
A comparison of Ken’s pre-PSC and post-PSC lessons reveals some change in

his use of representations. The percentage of segments in which he appropriately
selected and deployed representations increased slightly, and the percentage of seg-
ments in which he used multiple representations or made links between them de-
creased slightly. It seems likely that, outside of his PSC lessons, the focus on repre-
sentations fluctuated from lesson to lesson and may have been driven by nuances in
the curriculum.
Learning About and Using Students’ Understanding
A final focus of our PD program, and foregrounded in Workshop 3 of the PSC, is

building on students’ understanding. In our PD program, we explicitly focused on
3 The code Deployed Manipulatives is actually part of QMI section 3, but we include it here because
of its connection to the other codes involving manipulatives and similar data patterns.

requests for students to
describe or explain
three broad topics related to student understanding that are well-captured by the
QMI instrument: (1) prompting students to describe, explain and justify their ideas;
(2) interpreting student productions; and (3) using students’ errors. All three topics
were ones to which Ken was particularly attuned; he frequently mentioned in his
interviews and reflections that he was working toward improvements in these areas.
Encouraging teachers to have students describe, explain, and justify their math-
ematical thinking was a central premise throughout our PD program (Clark et al.
2005). Two QMI codes examined the number of times the teacher requested that
students describe or explain their ideas; in other words, they focus on the teacher’s
attempts to have students generate this type of information (and not whether stu-
dents were actually able to provide a description or explanation). Ken appeared
less focused on requesting students to describe their ideas in his PSC lessons than
his non-PSC lessons (see Fig. 6). At the same time, in his post-PSC lessons, over
two-thirds of the segments on average contained a request for a description. Ken’s
requests for student explanations remained essentially unchanged across the three
lesson categories, although there were slightly more requests for explanations in the
PSC lessons than the pre- or post-PSC lessons.
An examination of Ken’s interviews and reflections throughout the PD program
suggests that he internalized the goal of helping students to voice their ideas, al-
though we see only minimal evidence from the “describe” and “explain” codes.
During a PD workshop in December 2003, Ken wrote, “[In my classroom] I would
like to see more thought-provoking, student-engaging questions and activities that
really get kids talking about math.” Then again, in October 2004, Ken wrote, “My
teaching goal is to better facilitate student-led discussions.” During the 2004–05
school year, Ken made it more of a priority to have students come to the front of
the room and share their ideas after working in small groups. As a final reflection in
May 2005, he commented, “I realized the importance of talking about our thinking,
and giving kids the opportunity to share their ideas.”
An analysis of the codes interpreting student productions and using students’ er-
rors shows a notable difference between Ken’s PSC lessons and non-PSC lessons in
the degree to which he actively built on students’ comments and errors (see Fig. 7).
In his PSC lessons Ken devoted more time to both of these ways of using students’

Interpreting student
productions and using
students’ errors
ideas to develop the mathematical frame of the lesson. Ken also did more interpret-
ing of student productions in his post-PSC lessons compared to his pre-PSC lessons.
At the beginning of the PD program Ken reflected, “When I help a group or a
student, I tend to give them too big of a clue. I need to work on just guiding kids in
the direction of their own thinking” (July 2003). Over the next two academic years,
Ken strived to allow his students to move in the direction of their own thinking,
rather than asking everyone to think (and solve problems) in exactly the same way.
In an interview conducted after a videotaped lesson near the end of the 2004–05
school year Ken explained,
It feels so much better to get them to a level of understanding on something that they’ve
kind of got ownership of. I mean, those were their ideas. It was so much more valuable to
them to be able to go from their own perspective or the way that they thought about it to the
end, rather than my way of thinking about it. (May 2005)
Vignette Analysis: Ken’s Skyscraper Windows Lesson
In order to understand how Ken worked to support his student’s algebraic reasoning
by the end of the PD program, we consider his final PSC lesson in a more in-depth
manner. In this way, we can juxtapose some of the suggested findings from the QMI
coding with a particular algebra lesson. As with all of the PSC lessons, Ken had
taken part in a full-day PD workshop specifically focused on solving and planning
to teach this problem (Workshop 1). Because of the in-depth PD experience related
to the problem, and the fact that this lesson takes place after two years of PD, we
expect that it should showcase Ken’s instructional changes and highlight his best
efforts to support students’ algebraic thinking.
In this section of the chapter, we have created a vignette to depict Ken’s teaching
of the Skyscraper Windows problem (Driscoll 1999) (see Table 1). Ken modified
the problem in order to make it accessible to his sixth graders, by chunking it into
more manageable components over the course of two days. On the first day, students
worked in small groups to determine the cost of washing all of the windows on a
1-story building (in which all of the windows cost the same amount to wash). That
lesson moved back and forth between small group work and whole class discussions.
Table 2 The cost per

window and cost per floor for Number of floors Cost per window Cost per floor
each floor of an 8-story
building 1 $2.00 $76.00
2 $2.50 $95.00
3 $3.00 $114.00
4 $3.50 $133.00
5 $4.00 $152.00
6 $4.50 $171.00
7 $5.00 $190.00
8 $5.50 $209.00
Our vignette centers on the second day of the lesson, when Ken asked his students
to determine the cost of washing the windows on an 8-story building (in which the
cost per window varied by floor).
When we enter the lesson, Ken has just brought his students together after they
had been working in small groups. He stands by the whiteboard and calls on students
to explain their strategies for finding the total cost of washing the windows on all
8 floors. As the discussion unfolds, Ken attempts to help the students clarify their
ideas by asking questions and constructing a table depicting the cost per window for
each floor of the building (see Table 2). The table becomes a particularly important
heuristic in the students’ sharing of responses and making sense of strategies.
In response to Ken’s initial question about the total cost to wash all the windows
in the building, students provide three different answers: $1140, $608, and $209.
Ken works through each of these ideas with the students, beginning with $209.
Ken begins, “Now, I want to talk about “process” [referring to how the students solved the
problem]. Did we all agree about the cost of washing an 8-story building? How much would
it cost?”
One girl responds, “$1140.” Ken records this response on the whiteboard.
Ken asks, “Others? What did other people get?”
Maria shares, “$608.”
Ken writes down the response and smiles, “Maybe you guys don’t agree [on the solution],
but that’s okay.”
Another group reports, “$209.”
Ken again writes the number down and asks, “How did you get that?”
Julia says that she multiplied $5.50 ∗ 38 [the number of windows on each floor].
Ken then asks the class where Julia got $5.50. The answer is not given, but Ken points to
the table showing that $5.50 ∗ 38 = $209. He questions, “Do you think that’s how much
the whole building costs?” prompting the students to carefully consider the meaning of the
number $209.
Several students respond that this amount is for only the 8th floor.
This short exchange illustrates a number of QMI codes including requesting stu-
dent descriptions, using a student’s error, and using a mathematical representation.
We found it interesting that Ken opts to delve into the least sophisticated strategy
first, encouraging students to carefully analyze this idea, understand where the er-
ror occurs, and then move forward. Note how his questioning about Julia’s strategy
enables other students to make sense of her reasoning. Even though the students are
unable to immediately respond to his question about where Julia got $5.50, once
he points to this number on the table, they quickly see that she only accounted for
the windows on the 8th floor. Thus, Ken’s questions, in combination with the visual
representation, lead the class to understand that her group’s strategy was correct for
finding the cost of washing all the windows on the eighth floor, but not for the entire
building.
Ken moves on to the next solution strategy, unpacking how one group of students
arrived at the answer $608. Ken continues questioning and this time he is able to
generate a more elaborated explanation.
Ken asks, “Who gave us $608?”
Maria responds, “To get my answer I took 76 ∗ 8, where it is $76 for each floor.
Another student jumps in to counter this answer, saying, “But the price increases by $.50
each floor, so you can’t just multiply the result by $2 for each window [like on the first
floor].”
Ken revoices the student’s response, “So this would only work for us if each floor was $2
per window? But is that the case [for this problem]?”
Several students respond, “No.”
Ken continues in his effort to revoice the student’s idea, “I like your effort, but that solution
did not work since the price increases $.50 for each floor.”
In this excerpt, Ken assures the students that he values the contributions they
have made. He intentionally promotes a discussion that builds from students’ errors
in order to ensure that all students understand the central concepts in this problem.
Ken encourages students to engage with their classmates’ ideas, and he helps to
interpret and clarify ideas that might not be understood by everyone. In addition,
Ken models general mathematical language that is appropriate for this problem.
Lastly, Ken moves on to the third answer, $1140. He remains in the front of the
room, asking questions to guide the conversation. He erases the first two incorrect
answers ($209 and $608) from the whiteboard. It seems clear that Ken has inten-
tionally chosen to work through these erroneous solutions before delving into an
explanation for the correct response.
Ken asks, “Well, what about this one?” [pointing to $1140]
Michelle shares, “I got what it cost to wash each floor individually, and then I added them
all together.”
Other students nod in agreement.
Ken reiterates, “So, you took $76 + $95 + $114 and so on to get $1140?” [pointing to the
table on the whiteboard].
Michelle and others at her table say, “Yes.”
Ken asks, “Does everyone see how they did that? And, how it is different than the first two
strategies we looked at?”
The preceding discussion of the first two incorrect strategies paved the way for
students to follow the logic behind this relatively brief description of the correct
strategy. Ken again referred to the table in order to clarify what Michelle meant,
making links between what was written on the table and the specific numbers that
she added to get to the total amount. This exchange demonstrates the intersection
of several QMI codes: using a mathematical representation, making links, and inter-
preting a student’s production. Shortly after the conversation took place, Ken asked
the students to work in groups to find the total cost of washing windows on a 30-
story building. Their success on this quite complicated task indicates that most did
follow the reasoning behind the strategies that were presented for an 8-story build-
ing.
From the perspective of our research team, Ken clearly enacted the goal of sup-
porting his students’ thinking about the algebraic patterns in this problem. He used
a variety of instructional techniques that offered all of his sixth graders access to
challenging mathematics in a meaningful context. Furthermore, throughout the les-
son, Ken listened carefully to students’ ideas and supported them to follow their
own line of reasoning, moving them forward in the intended direction by engaging
in productive conversations characterized by appropriate mathematical language,
descriptions and explanations, interpreting of student ideas, building from student
errors, and frequent references to mathematical representations.
Conclusions
There is a broad consensus that major shifts are needed in the way teachers approach
mathematics instruction, particularly algebra teaching for middle school students.
However, many studies have shown that teachers’ conceptions of instructional prac-
tices are resistant to change (Ferrinni-Mundy and Schram, 1997; Jacobs et al. 2006).
The PSC model takes a comprehensive approach to PD, seeking to enhance teach-
ers’ knowledge and skills in a variety of domains. Given the STAAR project’s focus
on algebra, all three iterations of Workshop 1 concentrated on improving teachers’
algebraic content knowledge, Workshop 2s highlighted core practices of problem-
based algebra instruction, and Workshop 3s highlighted eliciting and building on
students’ understanding of algebraic concepts. The PD program that the STAAR
participants engaged in encouraged them to work toward changes on a host of inter-
related topics, all widely seen as key elements in promoting students’ algebraic
learning.
Our data suggests that Ken Bryant internalized many of the PD program’s goals,
and strived to make important changes in his mathematics classrooms. Drawing
from Ken’s interviews and written reflections throughout the project, together with
an analysis of his classroom instruction over two years, it appears that Ken made
steady and intentional progress toward these goals.
We also found that Ken’s PSC lessons looked, in many ways, different from his
non-PSC lessons. Ken may have taught his PSC lessons differently because of the
extensive PD work around those problems, which could have increased his content
and pedagogical content knowledge on specific mathematical topics, and prompted
an emphasis on particular instructional goals. In addition, the PSC problems were all
algebra-based, whereas Ken’s other videotaped lessons varied with respect to their
content focus. It is possible that Ken’s algebra lessons, in general, look different
from his non-algebra lessons. We are also conscious of the fact that our videotaping
included only a small subset of the hundreds of mathematics lessons that Ken taught
over a two-year period.
Although the picture of Ken that we have painted may not be representative of
all aspects of his teaching, it seems reasonable to conclude that there were some
shifts in his instructional practices related to the goals of the program, and that
these changes can be attributed to his participation in the program. Additionally,
Ken appears to have strong intentions for continued pedagogical reflection and im-
provement.
Our initial program of research in the STAAR project suggests that focused and
sustained PD programs such as the Problem-Solving Cycle can influence change in
practice, albeit at a relatively slow pace. Additional research is needed to determine
whether these changes in instructional practice create a classroom environment that
makes mathematics more accessible to all students, and whether the PSC is scalable
and sustainable to the extent that it can make a notable difference in mathematics
teaching and learning. Furthermore, additional investigations are needed to under-
stand the relationship between the PSC model of PD and changes in instructional
practice corresponding to the curriculum and content teachers use on an everyday
basis.
References
Ball, D. L., & Cohen, D. K. (1999). Developing practice, developing practitioners: Toward a
practice-based theory of professional education. In L. Darling-Hammond & G. Sykes (Eds.),
Teaching as the Learning Profession: Handbook of Policy and Practice (pp. 3–32). San Fran-
cisco: Jossey-Bass.
Ball, D. L., Lubienski, S., & Mewborn, D. (2001). Research on teaching mathematics: The un-
solved problem of teachers’ mathematical knowledge. In V. Richardson (Ed.), Handbook of
Research on Teaching (4th ed.). New York: Macmillan.
Blanton, M. L., & Kaput, J. J. (2005). Characterizing a classroom practice that promotes algebraic
reasoning. Journal for Research in Mathematics Education, 36(5), 412–446.
Borko, H., Frykholm, J. A., Pittman, M. E., Eiteljorg, E., Nelson, M., Jacobs, J. K., Clark, K. K., &
Schneider, C. (2005). Preparing teachers to foster algebraic thinking. Zentralblatt für Didaktik
der Mathematik: International Reviews on Mathematical Education, 37(1), 43–52.
Borko, H., Jacobs, J., Eiteljorg, E., & Pittman, M. E. (2008). Video as a tool for fostering productive
discourse in mathematics professional development. Teaching and Teacher Education, 24, 417–
436.
Clark, K. K., Jacobs, J., Pittman, M., & Borko, H. (2005). Strategies for building mathematical
communication in the middle school classroom: Modeled in professional development, imple-
mented in the classroom. Current Issues in Middle Level Education, 11(2), 1–12.
Cnop, I., & Grandsard, F. (1998). Teaching abstract algebra concepts using small group instruction.
International Journal of Mathematical Education in Science & Technology, 29(6), 843–851.
Cobb, P., Confrey, J., diSessa, A., Lehrer, R., & Schauble, L. (2003). Design experiments in edu-
cational research. Educational Researcher, 32(1), 9–13.
Design-Based Research Collective (2003). Design-based research: An emerging paradigm for ed-
ucational inquiry. Educational Researcher, 32(1), 5–8.
Driscoll, M. (1999). Fostering Algebraic Thinking: A Guide for Teachers, Grades 6–10. New York:
Heinemann.
Even, R. (1993). Subject-matter knowledge and pedagogical content knowledge: Prospective sec-
ondary teachers and the function concept. Journal for Research in Mathematics Education, 24,
94–116.
Ferrinni-Mundy, J., & Schram, T. (1997). Recognizing and Recording Reform in Mathematics
Education: Issues and Implications. Reston, VA: National Council of Teachers of Mathematics.
Journal for Research in Mathematics Education Monograph, no 8.
Greeno, J. G. (2003). Situative research relevant to standards for school mathematics. In J. Kil-
patrick, W. G. Martin, & D. Schifter (Eds.), A Research Companion to Principles and Stan-
dards for School Mathematics (pp. 304–332). Reston, VA: National Council of Teachers of
Mathematics.
Grossman, P., Hammerness, K., & McDonald, M. (2009). Refining teaching, re-imagining teacher
education. Teachers and Teaching: Theory and Practice, 15(2), 273–289.
Hadjidemetriou, C., & Williams, J. (2002). Teachers’ pedagogical content knowledge: Graphs from
a cognitivist to a situated perspective. In A. D. Cockburn & E. Nardi (Eds.), Proceedings of
(Vol. 3, pp. 57–64). Norwich, UK.
Jacobs, J., Hiebert, J., Givvin, K. B., Hollingsworth, H., Garnier, H., & Wearne, D. (2006). Does
eighth-grade teaching in the United States align with the NCTM Standards? Results from the
TIMSS 1995 and 1999 Video Studies. Journal for Research in Mathematics Education, 37(1),
5–32.
Jacobs, J., Borko, H., Koellner, K., Schneider, C., Eiteljorg, E., & Roberts, S. A. (2007). The
problem-solving cycle: A model of mathematics professional development. Journal of Mathe-
matics Education Leadership, 10(1), 42–57.
Kaput, J. J. (2007). What is algebra? What is algebraic reasoning. In J. Kaput, D. Carraher, & M.
Blanton (Eds.), Algebra in the Early Grades. London: Taylor & Francis.
F. K. Lester Jr. (Ed.), Second Handbook of Research on Mathematics Teaching and Learning:
A Project of the National Council of Teachers of Mathematics (Vol. 2, pp. 707–752). Charlotte,
NC: Information Age Publishing.
Koellner, K., Jacobs, J., Borko, H., Schneider, C., Pittman, M., Eiteljorg, E., Bunning, K., &
Frykholm, J. (2007). The problem-solving cycle: A model to support the development of teach-
ers’ professional knowledge. Mathematical Thinking and Learning, 9(3), 271–303.
Lampert, M. (2001). Teaching Problems and the Problems of Teaching. New Haven: Yale Univer-
sity Press.
Lave, J., & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation. Cambridge,
UK: Cambridge University Press.
Learning Mathematics for Teaching (LMT). (2006). A coding rubric for measuring the quality of
mathematics in instruction (Technical Report LMT1.06). Ann Arbor, MI: University of Michi-
gan, School of Education.
Miles, M. B., & Huberman, A. M. (1994). An Expanded Sourcebook: Qualitative Data Analysis.
Thousand Oaks, CA: Sage.
Moses, R. P., & Cobb, C. (2001). Organizing algebra: The need to voice a demand. Social Policy,
31(4), 4–12.
Nathan, M. J., & Koedinger, K. R. (2000). Teachers’ and researchers’ beliefs about the develop-
ment of algebraic reasoning. Journal for Research in Mathematics Education, 31, 168–190.
National Research Council (1998). The Nature and Role of Algebra in the K-14 Curriculum: Pro-
ceedings of a National Symposium. Washington, DC: National Academy Press.
Putnam, R., & Borko, H. (2000). What do new views of knowledge and thinking have to say about
research on teacher learning? Educational Researcher, 29(1), 4–15.
Shulman, L. S. (1983). Autonomy and obligation: The remote control of teaching. In L. S. Shulman
& G. Sykes (Eds.), Handbook of Teaching and Policy (pp. 484–504). New York: Longman.
Winicki-Landman, G. (2001). Shhh. . . Let them think. . . Let them talk!. Australian Senior Math-
ematics Journal, 15(2), 30–38.
Using Habermas’ Theory of Rationality to Gain
Insight into Students’ Understanding
of Algebraic Language
Abstract In this chapter we consider students’ use of algebraic language in math-

ematical modeling and proving. We will show how a specific model derived from
Habermas’ construct of rational behavior allows us to describe and interpret several
kinds of students’ difficulties and mistakes in a comprehensive way, provides the
teacher with useful indications for the students’ approach to algebraic language and
suggests further research developments.
Introduction
Habermas’ work has attracted the interest of many educational scholars (see the re-
view of the translation into English of Truth and Justification by Tere Sorde Marti
2004). According to our knowledge, all uses of Habermas’ ideas so far concern gen-
eral issues in education, related to the changing aims and features of schooling in
a changing world, and the framing of participatory action research (in particular,
see Carr and Kemmis 2005; Kemmis 2005, 2006). No specific research develop-
ment concerns mathematics education, though some values and innovative trends in
this field are related to those general issues. However, we think that at least one of
Habermas’ constructs, that of “rational behavior”, is of specific interest for math-
ematics education. Indeed (as we will try to show in this chapter) it allows one to
analyze complex mathematical activities (like conjecturing, proving, modeling) in
a comprehensive way and to deal with them not only as school subjects and sets of
tasks, but also as ways of experiencing mathematics as one of the components of
scientific rationality.
F. Morselli () · P. Boero

Department of Mathematics, University of Genova, Genova, Italy
P. Boero

454 F. Morselli and P. Boero
In this chapter, we will present and discuss some specific ways of using Haber-
mas’ construct of rational behavior in mathematics education in the area of proving
and modeling.
First of all, we will present our adaptation of Habermas’ construct to the special
case of the use of algebraic language in proving and modeling. We will show what
it can bring to the field by comparing it with other analytical tools and elaborations.
Our adaptation will be illustrated through an ad-hoc selection of short emblematic
episodes that will be analyzed by means of the new theoretical tool. Afterwards,
we’ll discuss some relevant episodes from a teaching experiment in order to show
how the theoretical tool can be used by mathematics educators and teachers for
the management and analysis of teaching and learning of algebraic language in the
approach to proving. Finally, we’ll sketch research developments concerning the dy-
namic interplay between the components of rational behavior in the use of algebraic
language and the role of verbal language as a crucial tool for rational behavior in it.
These research developments will add new arguments to the elaboration presented
in Boero et al. (2008) and will relate to the specific functions of verbal language
in mathematical activities. We will also outline educational implications of these
developments.
Habermas’ Construct of Rational Behaviour
According to Habermas’ definition (see Habermas 2003, Chap. 2), a rational be-
havior in a discursive practice can be characterized according to three interrelated
criteria of rationality: epistemic rationality (conscious control of the validity of
statements and inferences that link statements together within a shared system of
knowledge, or theory); teleological rationality (conscious choice and use of tools
and strategies to achieve the goal of the activity); communicative rationality (con-
scious choice and use of communication means within a given community, in order
to achieve the aim of communication).
In a long-term research perspective, we think that Habermas’ construct is a
promising analytic instrument in mathematics education because it connects the in-
dividual and the social by taking into account the epistemic requirements of “math-
ematical truth” in a given cultural context and the ways of ascertaining and commu-
nicating it by means of suitable linguistic tools.
In our previous research we have dealt with an adaptation of Habermas’ construct
of rational behavior in the case of the approach to conjecturing and proving (see
Boero 2006; Morselli 2007; Morselli and Boero 2009). In this chapter we focus
our interest on students’ use of algebraic language in proving and modeling. Our
adaptation of Habermas’ construct will be introduced in the subsequent section.
Using Habermas’ Theory of Rationality to Gain Insight into Students’ Understanding 455
Adaptation of Habermas’ Construct of Rational Behavior to the

Case of the Use of Algebraic Language
The aim of this section is to match Habermas’ construct of rational behavior to the
specificity of the use of algebraic language in modeling and proving.
Algebraic language will be intended in its ordinary meaning: a system of signs
and transformation rules, which is taught in secondary school as a tool to generalize
arithmetic properties, to develop analytic geometry and to model non-mathematical
situations (in physics, economics, etc.). In particular, algebraic language can play
two kinds of roles for modeling (according to Norman’s broad definition: see Nor-
man 1993, and Dapueto and Parenti 1999, for a specific elaboration in the case
of mathematics): a tool for proving through modeling within mathematics (e.g.
when proving theorems of elementary number theory)—internal modeling; or a
tool for dealing with extra-mathematical situations (in particular to express relations
between variables in physics or economics, and/or to solve applied mathematical
problems)—external modeling.
Our interest in considering the use of algebraic language from the perspective of
Habermas’ definition of rational behavior comes from our previous research (Boero
2006; Morselli 2007), which suggests that some of students’ main difficulties in
conjecturing and proving depend on specific aspects (already pointed out in liter-
ature: see next Section) of the use of algebraic language that make it a complex
and demanding matter for students. In particular, we refer to: the need to check the
validity of algebraic formalizations and transformations; the correct and purposeful
interpretation of algebraic expressions in a given context of use; the goal-oriented
character of the choice of formalisms and of the direction of transformations; the re-
strictions that come from the needs of following taught communication rules, which
may contradict private rules of use or interfere with them. Accordingly, we propose
the following three dimensions of rational behavior in the use of algebraic language
in proving and modeling.
Epistemic Rationality
This consists of two requirements:

– modeling requirements, which concern coherency between the algebraic model
and the modeled situation: control of the correctness of algebraic formalizations
(be they internal to mathematics—like in the case of the algebraic treatment of
arithmetic or geometrical problems; or external—like in the case of the algebraic
modeling of physical situations) and interpretation of algebraic expressions.
– systemic requirements, which concern the use of algebraic language and methods.
In particular, these requirements concern the manipulation rules (syntactic rules
of transformation) of the system of signs usually called algebraic language, as
well as the correct application of methods to solve equations and inequalities.
Teleological Rationality
This consists of the conscious choice and finalization of algebraic formalizations,

transformations and interpretations that are useful to the aims of the activity. It in-
cludes also the correct, conscious management of the writer-interpreter dynamics
(Boero 2001): the author writes an algebraic expression, transforms it and interprets
it according to the goal of the activity.
Communicative Rationality
In the case of algebraic language, we need to consider not only communication with
others (explanation of the solving processes, justification of the performed choices,
etc.) but also communication with oneself (in order to activate the writer-interpreter
dynamics). Communicative rationality requires the author to follow not only com-
munity norms concerning standard notations, but also criteria for easy reading and
manipulation of algebraic expressions.
As an example to illustrate some aspects of the relationships and differences be-
tween epistemic and teleological rationality, we can consider the problem of finding
the rectangle with maximum area among those of given perimeter, say 2p. This
example uses algebraic language both in (internal) modeling and in proving.
In order to find and justify the solution, we can express the length of each of the
two longer sides of the rectangle as p/2 + x, consequently the length of each of the
shorter sides is p/2 − x. The expression (p/2 + x)(p/2 − x) represents the area of
the rectangle; by multiplying we get:
Area = p 2 /4 − x 2 .
By checking this expression we see immediately that the maximum is obtained when
x = 0. By interpreting this result we find a square of side p/2 as the rectangle of
maximum area.
By choosing another expression for the area, we call x the length of one side of
the rectangle, consequently the length of the other side will be p − x and the formula
for the area will be:
Area = x(p − x) = px − x 2 .
In this case it is not so evident that the area reaches its maximum when x = p/2
(either analytic geometry tools or calculus tools, or a change of variables, are needed
to get the result).
These two simple examples show what we mean by epistemic rationality and
teleological rationality: both algebraic expressions used to model the area of a rect-
angle of given perimeter are correct (modeling requirements of epistemic rationality
are satisfied in both cases). The systemic requirements of epistemic rationality are
also satisfied (performed algebraic transformations are correct). However, from the
teleological rationality point of view the two expressions are not equivalent: only
the first one yields the result in a straightforward way.
Further examples will be provided in the next Sections.
Relationships with Other Studies on Proving and Modeling and

on the Teaching and Learning of Algebra
The aim of this Section is to link our work on the use of algebraic language in
proving and modeling with some studies that deal with proving, modeling and the
teaching and learning of algebra. A literature review concerning these research do-
mains is out of the scope of this section. What we want to show is how some results
and research perspectives in these domains are connected with some aspects of ra-
tional behavior in modeling and proving considered in our study, or even can be
reinterpreted using our perspective.
Proving
In this Subsection we consider studies on proving and proof that motivated our
initial adaptation of Habermas’ construct of rational behavior to the case of proving
in general (see Boero 2006; Morselli 2007; Morselli and Boero 2009), and other
studies directly related to the topic dealt with in this chapter (the use of algebraic
language in proving).
As Balacheff (1982) pointed out, the teaching of proofs and theorems should
have the double aim of making students understand what a proof is and learn to
produce it. Accordingly, we think that, in mathematics education, proof should be
treated considering both the object aspect (a product that must meet the epistemic
and communicative requirements established in today mathematics—or in school
mathematics) and the process aspect (a special case of problem solving: a process
intentionally aimed at a proof as product—the teleological dimension, in our adap-
tation of the Habermas’ construct).
Dreyfus (1999) points out the great divide that may exist between what is an
acceptable explanation for the teacher and what is an explanation for the students:
“Which aspects of an answer (or solution) are considered most important: compu-
tation, statement of the answer, relationship between computation and answer, pro-
cedural or conceptual? Are the same aspects considered important by the teachers
and by the students (. . . )? How are the students supposed to know what the teacher
considers important?” (p. 93)
For Dreyfus, the problem is that “most students have not been enculturated
into the practice of proving, or even justifying the mathematical process they use”
(p. 94).
Dreyfus’s work has driven our attention on the issues of students’ awareness
about the epistemic, teleological and communicative components of rationality in
proving, and on the possible mismatch between the students’ and the teacher’s views
on the different requirements and their importance.
Stylianides (2007) deals with the crucial question of introducing pupils into the
culture of theorems. In his study, he proposes the following definition of proof that
can be applied in the context of a classroom community at a given time:
“Proof is a mathematical argument, a connected sequence of assertions for or
against a mathematical claim, with the following characteristics:
• It uses statements accepted by the classroom community (set of accepted state-
ments) that are true and available without further justification;
• It employs forms of reasoning (modes of argumentation) that are valid and known
to, or within the conceptual reach of, the classroom community; and
• It is communicated with forms of expression (modes of argument representation)
that are appropriate and known to, or within the conceptual reach of, the class-
room community”.
(Stylianides 2007, p. 291).
We can interpret this definition as related to two aspects of rationality in proving:
the epistemic side for the first and the second characteristics; and the communicative
side, for the third one.
Concerning proving with algebraic language, we will try to show how our tool
can account in an original way for some phenomena (already considered in current
literature) regarding students’ learning to prove using algebraic language, in partic-
ular those concerning:
– students’ behaviors classified in terms of “proof schemes” (see Harel and Sowder
1998): for instance, our tool can account for external conviction proof schemes
(which are frequent among novices) in terms of dominant concerns for commu-
nicative rationality in the use of algebraic language, which hinder the necessity
of satisfying the requirements of epistemic rationality and teleological rationality.
We also note that Harel’s construct of “necessity”, a basic component of his DNR
perspective underlying proof schemes elaboration, can be connected to epistemic
rationality;
– the relationships between argumentation and proof (see Duval 1991, 2007; Pede-
monte 2007, 2008 for different positions on the subject): our tool can provide a
comprehensive perspective to situate the epistemic validity of proof, interpreted
in terms of epistemic rationality (in particular, its systemic requirements), in rela-
tion with the problem-solving character of proving, interpreted in terms of teleo-
logical rationality. This perspective can be very useful when students learn to use
algebraic language in conjecturing and proving because it allows the teacher to
distinguish between what must be taught as rules and criteria to follow strictly (on
one side, the syntactic rules of transformation of the algebraic expressions; on the
other, the correspondence between the algebraic expressions and the mathemati-
cal situation they should represent), and what must be managed by students in a
flexible and creative way (the choice of suitable algebraic representations of the
problem situation, the exploration of the algebraic expressions in order to foresee
possible, useful transformations to perform, etc.).
Modeling
Concerning modeling, we think that our tool provides insight into some aspects of
the modeling cycle (Blum and Niss 1991) and related students’ difficulties: in the
choice of the mathematical model; in its mathematical implementation; and in the
control of the validity of the model through the comparison of the results (derived
from its mathematical implementation) with the modeled situation. We will demon-
strate all these aspects through later example problems (the “Bomb problem” and
the “Spring problem”).
Teaching and Learning of Algebra
A complete overview of literature on teaching and learning algebra is beyond the

scope of this chapter. At the same time, we underline that our approach is differ-
ently oriented, since it starts from reflection on proving and modeling and turns to
algebraic language as a proving and modeling tool. We will try to show how some
educational studies on algebra contain elements that are linked to the three dimen-
sions of rationality we previously outlined.
Many studies refer to the crucial passage between arithmetic and algebra, point-
ing out continuities, false continuities and discontinuities (Filloy and Rojano 1989;
Bednarz and Janvier 1996), as well as the need for a careful study of transposition
issues (Chevallard 1984). In these cases, the focus is on epistemic aspects (some
rules that are still valid for arithmetic are no longer valid for algebra), but also tele-
ological (algebra can solve more problems than arithmetic). We also refer to studies
concerning an early approach to algebra, where algebraic thinking is progressively
constructed in the child as both an instrument and as an object of thinking in arith-
metic situations (Malara and Navarra 2003). Here epistemic and teleological ratio-
nality are again at issue.
Another strand of research concerns the use of algebra when solving word prob-
lems and its comparison with the use of arithmetic tools. Most students were found
to prefer arithmetic methods when solving word problems (Stacey and MacGregor
1999). Here we see a link to teleological rationality, since in most cases the use of
algebra would be more appropriate for solving the problem.
Studies that focus on how students use algebraic tools for proving include those
that deal with difficulties in using algebra, or with students’ reluctance to use algebra
in proving. Some studies even suggest that the use of algebraic manipulations, when
not coupled with understanding and control, may result in a mere “symbol push-
ing”. For instance, Weber and Alcock (2004) distinguish between semantic proof
production (when the prover uses instantiations of mathematical concepts to guide
the formal inferences that he or she draws) and syntactic proof production (when the
prover draws inferences by manipulating symbolic formulae in a logically permis-
sible way). The authors note that the two approaches require a different conceptual
understanding and also note that the syntactic approach often leads to proofs that are
convincing, but not explaining, while semantic proofs are more likely to have also
an explanatory power. Similarly, Douek (1999) found that university students of-
ten relied to semantically-rooted arguments when proving. Furinghetti and Morselli
(2009) report the case of a university student who, faced with a statement in number
theory to be proved, is keen to use algebra, but fails to choose a useful algebraic
representation for the situation. These studies suggest the importance of taking into
account not only the epistemic dimension, but also the teleological one.
More generally, Arzarello et al. (1994, 1995) deal with algebraic formulas as
thinking tools. They point out the importance of choosing the representation that is
most suitable to the task, and also that, once the formula is chosen, it may suggest
something new to the reader. They discuss these two uses as follows: “the first way
of thinking with a formula is transforming (part of) its intension manipulating it ac-
cording to its (supposed) extension; the second way is discovering a new (supposed)
intension, without doing formal manipulations, but looking at a new (supposed) ex-
tension (in a possibly new frame)” (Arzarello et al. 1994; p. 114).
We see here a reference to teleological and communicational rationality in the
first way, and communicative rationality in the second way.
Among the “general” studies on the teaching and learning of algebra, we mention
the one developed by Radford and Puig (2007). They note that the difficulties en-
countered by students when learning algebra are related to the meaning of signs and
the syntax of the algebraic language: “More specifically, students’ difficulties are
often connected to: (1) the understanding of the distinctive manner in which simple
signs (e.g. “x”, “n”) and compounded signs (e.g. “2 + 5” or “x + 17”) stand for the
objects that they represent, and (2) the grasping of the sense of the operations carried
out on those signs.” (p. 146). These two kinds of difficulties are linked to epistemic
rationality. Adopting a historical perspective, the authors note that “contemporary
algebra, with all its concepts, is the product of a lengthy historical-cultural process
that the students encounter in the highly complex social institution that we call the
school. To learn algebra is not to construct the objects of knowledge (for they have
already been constructed) but to make sense of them” (p. 152). In their perspective,
learning algebra is a cultural and social process. This is coherent with our frame,
since the perspective of rationality allows us to take into consideration the social
and cultural dimensions of learning. The authors also note that solving equations
through algebraic symbolism rests on a mode of diagrammatic thinking. This means
that “symbolic algebraic thinking requires the cognitive ability to switch between
verbal and perceptual meanings and to become conscious that the latter is governed
by the shape of expressions whose syntactic complexity may lead to multilayered
perceptual meanings” (p. 160). This suggests a connection between epistemic and
teleological dimensions of rationality.
Radford and Puig ’s study mainly concerns how to make sense of algebraic sym-
bols. Among the studies dealing with this issue, we also consider that of Arcavi
(1994, 2005), who illustrated the symbol sense and pointed out the importance of
three things: knowing how and when to use (or not use) algebra, choosing the best
representation among different possibilities, and reading the symbols. In our per-
spective, these aspects refer to teleological and communicative dimension. We feel
that our approach may frame all the different aspects mentioned by Arcavi into the
three dimensions of rationality, and it places these aspects within the larger frame of
rationality in proving (in mathematical activity). Furthermore, our theoretical tool
leads us to study the different components that may be present at the same time. We
may say that our approach is in line with Arcavi’s elaboration, but we have a more
explicit aim of using the theoretical tool in order to analyze the students’ processes
and to study interconnections between the different components. Furthermore, re-
call that the model we drew from Habermas is a comprehensive frame for studying
mathematical activity as a rational behavior, with a focus on proving. Later, we in-
vestigated whether the same model is viable also for studying specific aspects of
proving, such as the use of algebraic language. Thus, we may say that our elabora-
tion and Arcavi’s are compatible, but different in aim and scope.
Let us consider now the notion of structure sense (Linchevski and Livneh 1999;
Hoch and Dreyfus 2006). Students are said to display structure sense for high school
algebra if they can: “Recognise a familiar structure in its simplest form; deal with a
compound term as a single entity and through an appropriate substitution recognise
a familiar structure in a more complex form [. . . ]; choose appropriate manipulations
to make best use of a structure” (Hoch and Dreyfus 2006, p. 306). From the defini-
tion, structure sense seems to take into consideration mainly aspects of teleological
rationality, (choose appropriate substitutions, perform appropriate manipulations),
together with some aspects of communicative rationality, since it takes into consid-
eration the reading of formulas.
Pierce and Stacey (2001), in their study concerning the use of computer al-
gebra systems, deal with the idea of “algebraic insight”, defined as “the subset
of symbol sense which is needed to solve a problem already formulated math-
ematically. Students need algebraic insight to enter expressions correctly, mon-
itor the solution process and interpret the output as conventional mathematics”
(p. 418). Part of the algebraic insight involves an algebraic expectation, defined
as the analogue of arithmetic estimation. It consists, for example, in expecting that
(2 − x + x 2 )(x 4 − x 3 + 27x − 63) will be a polynomial of degree six. Algebraic
expectation encompasses the recognition of conventions and basic properties, the
identification of structures and key features. We may say that algebraic insight con-
cerns some aspects of rationality: entering expressions correctly is related to epis-
temic rationality, monitoring the solution process is related to epistemic and tele-
ological rationality, and interpreting the output is related to epistemic and commu-
nicative rationality.
In order to justify a new analytic tool in mathematics education, it is necessary to
show how it can be useful in describing and interpreting students’ behavior, orient-
ing and supporting teachers’ educational choices, or suggesting new research devel-
opments. The aim of the following Sections is to provide evidence that these three
aims are achieved by our adaptation of Habermas’ construct to the use of algebraic
language in proving and modeling.
Description and Interpretation of Student Behavior
This Section consists of two parts.

In the first part we provide examples of the use of our adaptation of Habermas’
construct to interpret and discuss short excerpts of students’ texts dealing with the
use of algebraic language in elementary modeling and proving. The aim of this sub-
section is twofold. First, we want to illustrate the use of our construct; paraphrasing
Dreyfus (1999, p. 87), we can say that neither the examples nor the students are rep-
resentative in any sense but have been chosen for illustrative purposes. The choice
of examples, including topics, level of mathematics, and level of students have been
influenced by our own personal bias and experience.
Second, we want to show that our construct is a flexible tool to deal with differ-
ent kinds of problem situations and related students’ difficulties at different school
levels (from grade 8 to university courses), thus including students first learning to
use of algebraic language in modeling and proving, as well as its use by more com-
petent students. Some similarities between the two cases (novices and competent
students) are interesting because they show how a correct and effective use of alge-
braic language in modeling and proving is not an aim that can be attained once and
for all, but that its development requires a long-term maturation and familiarization
with the use of algebraic language in different kinds of tasks.
In the second part of this section, we concentrate on a teaching experiment per-
formed in grade 7, concerning the use of algebraic language in proving in the arith-
metic domain. The purpose of this part is to show how our analytic tool both allows
us to interpret students’ behaviors and to shed light on the different dimensions that
teachers must take into account when guiding students towards an effective use of
algebraic language in an internal modeling activity and in proving.
Habermas’ Analytical Tool: Examples of Analysis of Student

Behavior at Different School Levels
The following examples are derived from a wide corpus of students’ individual writ-
ten productions and transcripts of a posteriori interviews, collected in the last fifteen
years by the Genoa research team in mathematics education. The wide corpus was
collected during the last fifteen years. The first analysis of the corpus, performed at
different times, showed the importance of taking into account dimensions that, later,
we found explicitly treated in Habermas’ theory. This means that our elaboration of
the theoretical tool was fostered by the reflection on the corpus. After our adaptation
of the theoretical tool, we came back to the corpus and checked the reliability of the
tool. We present here a selection of episodes, analyzed through the tool, with the
double aim of illustrating the tool and showing how the tool helps us understand the
episodes.
In particular, we will consider five categories of students:
(a) 8th grade students who are approaching the use of algebraic language in model-
ing physical phenomena;
(b) 9th grade students who are approaching the use of algebraic language in proving;
(c) 11th grade students who are learning to use algebraic language in modeling
physical phenomena;
(d) students who are attending university to become primary school teachers;
(e) students who are attending the third year of the university course in mathemat-
ics.
A common feature for all the considered cases is that the individual tasks require
not only the solution, but also the explanation of the strategies followed to solve the
problem. However, while in cases (a), (d) and (e) the explanation of the strategies
was inherent in the didactical contract already established with the teacher for the
whole course, in the cases (b) and (c) such explanation was only an occasional
request. Each individual task was followed by a posteriori interviews.
Example 1 (The sum of two consecutive odd numbers) The students (22 students,
grade 9) experienced the traditional teaching of algebraic language in Italy: trans-
formation of progressively more complex algebraic expressions in an attempt to
“simplify”. In order to prepare students for the task proposed by the researcher, two
examples of “proof with letters” had been presented by the teacher, one of which
included the algebraic representation of even and odd numbers.
THE TASK: Prove with letters that the sum of two consecutive odd numbers is
divisible by 4.
Here we report some recurrent solutions (in parentheses the number of students
who performed such a solution; note that “dispari” means “odd” in Italian).
E1 (4 students):
In this case, we can observe how the systemic requirements of epistemic rationality
are satisfied (algebraic transformation works well), while the internal modeling re-
quirements fail to be satisfied (the same letter is used for two different numbers).
E2 (8 students):
In this case, both the systemic and the internal modeling requirements of epistemic
rationality are satisfied, but the requirements in teleological rationality are not sat-
isfied: students do not realize that the chosen representation does not allow to move
towards the goal to achieve (because the letter d does not represent in a transparent
way the fact that d is an odd number), thus they do not change it.
E3 (5 students): (or similar sequences)
We can infer from the context (and also from some a-posteriori comments by the
students) that “dc” means “dispari consecutivi” (consecutive odd numbers).
In this case epistemic rationality fails in the first and in the second equality, but
teleological rationality works well: the flow of thought is intentionally aimed at the
solution of the problem; algebraic transformations are used as a calculation device

to prepare the conclusion (divisibility by 4). We may note the presence of commu-
nicative rationality in the use of dc, which is related to private communication (with
oneself).
Example 2 (The product of two consecutive even numbers) The following task had
been preceded by the task from Example 1, performed under the guide of the teacher.
58 university students, attending the third year of preparation to become primary
teachers, performed the activity.
THE TASK: Prove in general that the product of two consecutive even numbers
is divisible by 8
Very frequently (about 55% of cases) students performed a long chain of trans-
formations, with no outcome, like in the following example:
E4:
In this case, we see how both requirements of epistemic rationality are satisfied:
internal modeling requirements (concerning the algebraic modeling of even num-
bers); and systemic requirements (correct algebraic transformations). The difficulty
is in the lack of an interpretation of formulas, led by the goal to achieve (thus in
teleological rationality). The student gets lost, even though the interpretation of the
fourth expression would have provided the divisibility of n(n + 1) by 2 because one
of the two consecutive numbers n and n + 1 must be even. We can also observe how
(in spite of the didactical contract), in general, no substantial sentence precedes or
follows the sequence of transformations (sometimes we find only a few words: “I
use formulas”; “I see nothing”).
In the following case, both the modeling and systemic requirements are not sat-
isfied: the same letter is used for two consecutive even numbers (note that “pari”
means “even” in Italian), and the algebraic transformation is affected by a mistake.
E5:
The student seems to work under the pressure of the aim to achieve: having foreseen
that the multiplication by 2 may be a tool to solve the problem, she tries to justify
it by considering the juxtaposition of two copies of p that generates “2”. Indeed, in
the interview the student said that she had made the reasoning “p is divisible by 2
and thus p 2 is divisible by 4” before completing the expression. In this case we can
see how teleological rationality prevailed over epistemic rationality and hindered it.
We have also found cases like the following one:
E6:
Also in this case, from the a posteriori interview we infer that the lacks in epistemic
rationality probably depend on the dominance of teleological rationality without
sufficient epistemic control:
I have seen that in the case p = 2 things worked well, so I have thought that
putting a multiple 8k of 8 in the general formula would have arranged the situation.
Example 3 (The bomb problem) TASK: A helicopter is hovering over a target.

A bomb is left to fall. Twenty seconds later, the sound of the explosion reaches
the helicopter. What is the relative height of the helicopter over the ground?
The problem was posed to groups of third year mathematics students in seven
consecutive years, and to two groups of 11th grade students (with a scientific–
oriented high school curriculum). The younger students were reminded that the
falling of the bomb happens according to the laws of uniformly accelerated mo-
tion, while the sound moves at the constant speed of 340 m/s. However no formula
was suggested.
The problem is a typical (though elementary) applied mathematical problem,
whose solution needs an external modeling process. In terms of teleological ratio-
nality, the goal should guide students to choose an appropriate algebraic model of
the situation, solve the second degree equation derived from the algebraic model,
and select the good solution (the positive one) by performing a suitable comparison
between the obtained solutions and the problem situation.
The first difficulty students meet is in the time coordination of the two move-
ments: it is necessary to tell the model that the whole time for the bomb to reach
the ground and for the sound of the explosion to reach the helicopter is 20 seconds.
The second difficulty is in the space coordination of the two movements: the space
covered by the falling bomb is the same covered by the sound when it moves from
the ground to the helicopter.
Let us consider some students’ behaviors.
Most students are able to write the two formulas:
E7:
These are standard formulas learned in Italian high schools in grades 10th or 11th , in
physics courses. About 25% of the high school students and 20% of the university
students stick to those formulas without moving further. From their comments we
infer that in some cases the use of the same letters for space and time in the two al-
gebraic expressions generates a conflict that they are not able to overcome. We can
see how general expressions that are correct for each of the two movements (if con-
sidered separately) result in a bad model for the whole phenomenon. Teleological
rationality should have driven formalization under the control of epistemic ratio-
nality; such control should have revealed the lack of the modeling requirements of
epistemic rationality, thus suggesting a change in the formalization. However, such
an interplay between epistemic rationality and teleological rationality did not work
for those students.
In other cases (about 10% of both samples) the coordination of the two times was
lacking, and the idea of coordinating the spaces (together with the formalization of
both movements with the same letters) led to the equation:
E8:
with two solutions1 t = 0, t = 68. Some students were unable to interpret and use
these times (because 68 is out of the range given by the text of the problem), but
other students found the height of the helicopter by multiplying 340 × 68; the fact
that the result is out of the reach of a helicopter did not provoke any critical reaction
or re-thinking, probably because it is normal that school problems are unrealistic!
Some students who introduced the third equation tb + ts = 20 added it to the first
two equations without changing the name of the variable (t).
Less than 60% of students of both samples wrote a good model for the whole
phenomenon:
tb + ts = 20
h = 0.5gtb2 = 340ts
and moved to a second degree equation by substituting ts = 20 − tb or tb = 20 − ts

in the equation:
0.5gtb2 = 340ts
Many mistakes occurred during the solution of the equation (mainly due to the man-
agement of big numbers). Once two solutions were obtained (one positive and the
other negative), in most cases the choice of the positive solution was declared but
not motivated. A posteriori comments reveal that the fact that a negative solution
is unacceptable (given that the other solution is positive!) was assumed as obvious,
without any physical motivation.
In terms of epistemic rationality, three kinds of difficulties arose: first, in the
control that the chosen algebraic model was a good model for the physical situation;
second, in the control of the solving process of an equation with unusual complex-
ity of calculations (big numbers); third (once the valid equation—a second degree
equation—was written and solved), in the motivation of the chosen solution.
In terms of communicative rationality, we can observe how (in spite of the re-
quest to explain the steps of reasoning) very few students from both samples were
able to justify the crucial steps of the solving process. How is it possible to interpret
this kind of difficulty? In some cases the steps were derived from a gradual adapta-
tion of the equations to the need of getting a “realistic” solution. In other cases the
equations were written as if the idea of coordination of the spaces and times of the
phenomenon was supported by an intuition, but no wording followed. A posteriori
interviews revealed that most students who had been unable to justify their choices
were sure about their method only afterwards, when checking the positive solution
and finding that it was “realistic”, thus revealing a lack in teleological rationality
(lack of consciousness about the performed modeling choices). Moreover, a num-
ber of inappropriately obtained solutions were quite realistic. Many students who
produced the correct solution were not able to explain (during the comparison of
solving processes) why the other solutions were mistaken. This suggests lacks in
1 We underline that g is rounded to 10, rather than 9.8.

teleological rationality (motivation of choices with reference to the aim to achieve)

and in epistemic rationality (control of the validity of the steps of reasoning). This
conclusion can be reinforced if we consider the fact that almost all students who
were able to provide a verbal justification for their modelisation were also able to
explain why the other solutions were not acceptable (even if those results were real-
istic).
Example 4 (The spring problem) The problem was proposed to Grade 8 students.
The students were requested to choose, among the following formulas, the one
which best represents the elongation of a spring according to the number N of clips
suspended to it (L is expressed in centimeters).
(I) L = 20 · (1 + N )
(II) L = 20 − 0.2 · N
(III) L = 20 + 0.2 · N
(IV) L = 20 + 20 · N
Students had at their disposal the following table, derived from an experiment:
N L
(number of clips) (centimeters)
0 20
10 21.8
20 23.6
30 26
40 28.2
50 30.3
60 32.2
70 34.1
Students’ background included: the use of algebraic formulas to represent ge-

ometric situations (perimeters and areas of geometric figures); the representation
of linear functions in the Cartesian plane; and the construction of tables like the
one above through direct measurement of the elongation of springs. At this stage
of students’ preparation, it was not expected that they produce a linear model for
the elongation of a spring, but only that they succeed in choosing between different
linear models provided by the teacher.
The aim of the task is relevant to the perspective of approaching mathematical
modeling of physical phenomena, because it is related to the choice of a mathemat-
ical model suitable to fit the behavior of the phenomenon. In terms of the analytic
tool we derived from Habermas’ theory, the aim of the task concerns the external
modeling requirements of epistemic rationality.
Different behaviors were observed. Here, we present the three main categories of
behaviors that were singled out, and we analyze them in terms of our construct.
Some students performed substitutions of values within the formulas without get-
ting any conclusion. Frequently, they also graphed the points thus obtained, without
comparing the graph with a possible qualitative behavior of the phenomenon (or
with a graphic representation of the points derived from the table). It seems that
the “ritual” accomplishment of the usual task of substituting values in order to get
points in the Cartesian plane hindered the aim of finding the good formula and/or of
comparing those values with the experimental values.
Some practices, such as substituting values in order to get points on the Carte-
sian plane, could be useful in order to compare the formulas in terms of modeling
requirements of epistemic rationality, but in this case these practices seem to be
just aimed at plotting the functions. We may say that teleological rationality did not
work (it was not present, or it was badly oriented).
Some students found the best formula by substituting values in each formula and
comparing the points thus obtained in the Cartesian plane with the points derived
from the table. In this strategy, the formulas are compared according to some ex-
ternal modeling requirements and systemic requirements of epistemic rationality: in
this case, teleological rationality drives the process towards the aim to attain.
Other students considered the structure of the formula and concluded that the
second formula cannot work, because it would imply that the length of the spring
decreases when the number of clips increases. In this case, we may say that they
put in action a control in terms of external modeling requirements. Afterwards, they
realized that the first and the fourth formulas are equivalent; we may say that they
performed a control involving the systemic requirements of epistemic rationality
inherent in the use of formulas, performed by substituting some values or by con-
sidering the distributive property. Comparing the third and the fourth formulas, they
chose the third one because the fourth one “produces values that are too big, when
N increases, in comparison with the table”.
Habermas Analytical Tool: Analysis of a Teaching Experiment
The Context of the Study: Description of the Research Project
The following examples are drawn from a research project, started in 2008 and still
going, entitled “Language and argumentation in the study of mathematics from pri-
mary school to university”.2 The project aims at setting up and experimenting with
teaching activities for different school levels (from grade 1 to grade 13), with an
eye on continuity between different school levels, and with a special focus on argu-
mentation, which is seen as a central theme in mathematics education (Hanna and
De Villiers 2008) as well as a crucial competence in the development of citizenship
2 The project is developed by the Mathematics Department of the University of Genova, with the
collaboration of the Regional School Office, and is carried out within the National Project “Lauree
Scientifiche” (MIUR-Confindustria).
(Anichini et al. 2003). The conception, implementation and analysis of the teaching
activities was carried out by a research team made up of university researchers in
mathematics education and school teachers, in a relationship of mutual exchange
and collaboration, according to the Italian paradigm of the Research for Innovation
(Arzarello and Bartolini Bussi 1998).
In this section, we refer to a sequence of teaching activities that was performed
in grade 7 (age of the students: 12–13). The researcher, who took part in the ac-
tivities together with the teacher, was present in the classroom and acted both as
an observer and, occasionally, as a sort of additional teacher, especially during the
classroom discussions. According to the general aim of the project, special care was
devoted to tasks such as “explain your solution, compare your solution with that of
your classmates, choose between two options and justify your choice”. Accordingly,
collaborative group work and mathematical discussion (Bartolini Bussi 1996) were
usual modes of work in the classroom. These choices were linked to the necessity
of establishing a sort of “argumentative attitude” in the classroom. The sequence of
teaching activities was conceived in order to introduce algebra as a proving tool in
the classroom. Students had prior experience in dealing with numerical expressions
and using formulas to express the areas and perimeters of plane figures. They had
no experience using algebra to represent geometric or arithmetic relations.
Next, we present a selection of episodes from the sequence of teaching activities,
with the aim of showing the viability of our construct not only for interpreting stu-
dents’ behaviors, but also for supporting/orienting teachers’ choices. We underline
that the aim of this chapter is not to discuss in detail the nature of the sequence of
activities. Data from the activities serve as a ground for our discussion on the use of
the analytic tool derived from Habermas’ theory. We also underline that our adapta-
tion of Habermas’ model is a tool for the researcher and, in a long-term perspective,
a tool for the teacher in order to better understand students’ performance and guide
them towards adherence to a mathematical rationality.
First Task: Choose a Number. . .
The first task that was proposed to the students is the following:
The teacher proposes the following game: Choose a number, double it, add 5, take away the
chosen number, add 8, take away 2, take away the chosen number, take away 1. Without
knowing the number that you initially chose, is it possible for the teacher to guess the result
of the game? If yes, in what way?
The students worked on the task individually, and afterwards they shared and
compared their solutions, first in small groups and then within a classroom math-
ematical discussion. Many groups found out that the result is always 10, indepen-
dently from the chosen number, and some students even tried to find out some rea-
sons why the result is always 10. See, for instance, this selection of written group
solutions, that goes from seeing that to looking for reasons why:
Group A: Yes, because at the end you always find 11 − 1

Group B: Yes, because it is a mathematical procedure, by means of which you get
always the same result, with any chosen number. The factor that causes that, is the
instruction “Take away the chosen number”.
Group C: With any chosen number, the result is 10 because multiplying by 2 is
equivalent to adding twice the chosen number, the same number that after must
be taken away twice, which gives zero, and doing the other calculations, even in a
different order, you always get 10.
Some students were able to find out a reason why, but were not able to communicate
the reason to their classmates. Students realized that solutions in natural language
were not always efficient in communicating the reasons to others. This paved the
way to the subsequent task, aimed at proposing algebra as a proving tool.
Second Task: Representing the Game
The students were given the following task:

Write the game as an expression, using a different color for the chosen number. Write an
expression that works for any number you choose.
The students solved the task individually, and afterwards they shared and compared
their solutions within a mathematical discussion orchestrated by the teacher. As
we’ll see in the following, two main representations of the game were singled out.
As regards the first question (writing the game in form of expression), the stu-
dent Ric chose to represent the game as an expression, while the student Tor chose
to represent the sequence of calculation to be performed, thus creating a sort of pro-
cedural representation of the game.
Ric’s representation:
Tor’s representation:
As regards the second part of the task (“Write an expression that works for any
number you choose”), Ric proposed a representation of the game as an expression,
while Tor proposed a sequential-algorithmic representation of the game:
Ric’s representation:
Tor’representation:
We may note that the expression of Tor does not satisfy the internal modeling re-
quirements of epistemic rationality, since the same letter, N , is used to represent
the different results of the calculation steps. In contrast, Ric’s expression satisfies
the internal modeling requirements of epistemic rationality. A significant discus-
sion comparing Ric’s and Tor’s representations of the problem was fostered by the
teacher’s question: In your opinion, which of the two representations would be cho-
sen by a mathematician?
Here we provide a transcript from the mathematical discussion (at this point, the
two representations are written on the blackboard, so that all students may refer to
them):
1. Mir: I would choose Ric’s expression, because it is easier to understand and. . .
faster.
2. Teacher: faster. Giam?
3. Giam: I would choose Tor’s expression, because it is more schematic and. . .
any person, even a 6-year-old, may understand it.
4. Teacher: any other idea?
5. Brac: I would choose Tor’s expression, but writing different letters for any re-
sult.
6. Teacher: because, in your opinion, if we always use the same letter. . . you
would change the letter. Why?
7. Brac: because the result changes at each passage, even if the result finally is
always 10, but the result changes at each passage.
8. Teacher: and you think that. . . if I use N , N should stand always for. . .
9. Brac: for the number that I chose at the beginning.

10. Teacher: while N × 2 . . . (many voices: no). So, M, and after M + 5 . . .
equals. . . and so on. . . Using different letters. Mat?
11. Mat: I would choose both. Because one is faster and more immediate. . .
12. Teacher: which one?
13. Mat: while the other one is more. . . schematic.
14. Teacher: Soz, which one would you choose, if you were a mathematician?
15. Soz: I would choose Ric’s, because it is more complicated, but also simpler.
16. Teacher: why is it simpler?
17. Soz: because it is. . . more complicated to understand, but also simpler.
18. Teacher: Chris?
19. Chris: I would choose Tor’s, because it puts into evidence the results. But it
would be better to change the letters, as Brac suggested.
20. Teacher: so the results. . . how could we call them? The intermediate results. . .
the results of each step of calculation. . .
21. Chris: yes.
22. Nav: I would choose Ric’s because the expression you have. . . you put N in
the place of the chosen number, because in mathematics N stands for “number”,
and it is more. . . more. . .
23. Teacher: more coherent with. . .
24. Nav: it follows better the request.
25. Teacher: Ric, could you tell us why do you think that are the advantages of
your representation?
26. Ric: well. . .
27. Teacher: because you wrote the expression is equal to 10, isn’t it?
28. Ric: first, because there are no parentheses, so it is easier, second, because in
the place of the chosen number there is N , which stands for any number, and
then, since we have 10 with any chosen number, you put N and you always get
10.
29. Teacher: and you get 10. But also Tor used the letters. . .
30. Voices: yes, but he goes to new line. . .
31. Alex: for me, Ric’s is more correct, because it is an expression and it [the text]
says that it must work for any number. Ric’s is an expression and works for
any number, while Tor’s is an expression decomposed into operations, so. . .
The text asks you to create an expression, it doesn’t ask you to decompose this
expression.
32. Teacher: so, for you the advantage of Ric’s is that it is an expression.
33. Alex: yes.
34. Teacher: rather than a sequence of operations. So, both used a letter. What you
say is that Ric created a real expression, while. . .
35. Alex: Tor’s is an expression decomposed into operations
36. Teacher: Ash?
37. Ash: I would choose Ric’s expression, because N indicates always the same
number, while in Tor’s expression N is both the chosen number and the results
of the computations.
Lines #1 and #3 show that some students judge Tor’s expression to be a “legitimate”
mathematical representation as well as that of Ric. This led us to analyze the cri-
teria according to which the students evaluate the representations. Lines #1 and #3
suggest that both representations are judged first of all in terms of their efficiency
and easiness in relation to the game. In the students’ view, they are both legitimate,
since they both allow one to perform the game. This can be interpreted, in terms
of our adaptation of Habermas’ construct, as a predominance of the communicative
dimension over the other ones. The students evaluate the representations basically
on the basis of the communicative dimension.
Brac’s comment (#5) brings to the fore also the epistemic dimension: the repre-
sentation created by Tor does not meet the internal modeling requirements of epis-
temic rationality, since the same letter, N , is used to represent the chosen number,
but also the result of the intermediate steps of calculation. The same position is hold
also by other students, who suggest amending the representation given by Tor (#19,
#37).
We may note that, throughout the discussion, the different dimensions of ratio-
nality appear: some students judge the two representations in terms of “easiness” or
“complicatedness” (thus focusing on the communicative dimension), other students
even refer to the adherence to the text (#24, 31). Summing up, we may say that two
streams of discussion intertwine: evaluating the correctness of Tor’s expression (in-
ternal modeling requirements of epistemic rationality) and comparing between the
two representations in terms of comprehensibility (communicative rationality) and
adherence to the text. The adherence to the text may be interpreted in terms of com-
municative rationality (adherence to the norms of the community), but also in terms
of teleological rationality: the students’ aim, in this part of the activity, could be
more focused on the representation of the game than on the search for reasons why
the result is always 10. We note that the choice between the two representations,
and the risk of considering the expression as the “official” way of representing the
game, are crucial points, since the aim of the teacher is not just to make the students
choose Ric’s expression as the “legitimate” way of representing the game, but also
to guide the students to understand why mathematicians usually choose this expres-
sion. In other words, the teacher wants: at first, to work on the correct use of letters
in algebra; after, to work on the way of using algebra, which should be considered
“legitimate” not in terms of adherence to external rules (for a sort of “authoritarian”
legitimacy) but in terms of usefulness and efficacy according to the goal of the activ-
ity. This requires shifting students’ attention from the adherence to the text (which
requires an expression) back to the original goal of the activity (to understand why
the result is always 10).
In the successive part of the mathematical discussion, the representation given
by Tor is amended according to the suggestions of the students. In terms of our con-
struct, Tor’s representation is amended in terms of epistemic rationality, in order to
meet also the modeling requirements. Here we report the amended version:
Tor’s “amended” version:
At this point, the discussion concerns the comparison between two representations:
Ric’s expression and Tor’s amended “sequential” representation. This is quite a cru-
cial point. Indeed, both representations are correct from a mathematical point of
view (thus meeting the modeling requirements of epistemic rationality) and are also
perceived as efficient from a communicative point of view. So the question arises for
the teacher: how is it possible to lead the students to understand that a mathemati-
cian would rather choose Ric’s expression? The point is that the Ric’s expression
is better for the original aim of the task (to understand why it is possible to know
the result of the game). This means, in terms of the model, that Ric’s expression
meets also the teleological requirements. For the teacher, it is important not only to
have the students choose Ric’s representation, but also to lead them to understand
the reasons why it is more suitable. That is, Ric’s expression is a veritable proving
tool, since it allows one to understand why the result is always 10.
In the first part of the discussion, students expose their motivations for the choice
of one of the two representations. Some motivations are still at the communicative
level:
Cler: for me, Ric’s expression is more correct, because, at our age, we just studied
expressions, so it is easier for us.
Mus: for me, it is worthwhile to use Ric’s representation because it is more
schematic and more mathematical.
In some interventions, epistemic and communicative dimensions are intertwined:
Alex: first of all because [Ric’s] follows the text more, and after because it is more
correct.
Some students support Tor’s representation, or highlight the equivalence between
the two representations, as we see in the following comments:
Giam: I mean, for example, Tor’s representation, anybody can do it, and he can
follow all the steps, while in Ric’s, yes you do it, but you don’t really realize
what you are doing. I mean, maybe you do it, but without. . . for example,
the step N × 2, with Tor’s representation you do N × 2, and the result, while
with Ric’s you must go on fast, and maybe you get lost. . .
Brac: we can take both, because in the expression we do both Tor’s and Ric’s,
because in Ric’s, even though it is not so complex and you can do it in your
mind, you do a lot of steps, and it is as if you did Tor’s.
We may note that the issue of the equivalence of the two representations is still at
the epistemic level. In the subsequent part, both observer and teacher intervene in
order to bring to the fore also the teleological dimension:
58. Soz: [to the observer] and you, which one would you choose?
59. Observer: I think that. . . you all said a lot of good things, actually doing one
or the other is the same, and in both cases you get the result, OK? But do
you remember the question of last session? The question was not “what is the
result”, but “will the teacher be able to guess the result?”
60. Cler: just from the expression.
61. Observer: just from the expression, from the initial game. Just knowing the
game, will the teacher be able to guess the result, independently from the chosen
number?
62. Voices: yes.
63. Observer: we said yes, and in the last session you also explained why, right?
64. Chris: because you always get 10.
65. Observer: you said: yes, because you always get 10, and some of you also
explained something more, we also had some motivations why you always get
10.
66. Teacher: do you remember? Brac, you told it, because you told that doing N ×2
means. . .
67. Brac: I mean. . . it is like doing. . . yes, it is like doing N + N .
68. Teacher: N + N in the expression written by Ric, then. . . there is N × 2, Brac,
please go to the blackboard and write N + N under N × 2. Do we all agree that
it is the same thing? And after you write all the expression: +5 − N + 8 . . . And
you already noticed that. . . after N + N , what do I have?
69. Ash: −N .
70. Voices: two times.
71. Teacher: and so?
72. Brac: they all disappear.
73. Teacher: can I understand this, in Tor’s representation?
74. Voices: no.
75. Fag: but, at the end there is +8, so, the two representations are equivalent, but
Ric’s is. . . easier.
76. Teacher: but why is it easier?
77. Giam: because you understand that the chosen number disappears.
78. Teacher: because I can answer. . .
79. Observer: to the original question. Ric’s representation helps us understand

why it is not necessary to know the chosen number to get the result.
80. Teacher: do you understand? I mean. . . in Ric’s representation I understand
why the chosen number doesn’t matter.
81. Observer: I see better that the chosen number disappears.
82. Teacher: [to Brac, who is at the blackboard] underline all the N s. N + N −
N − N . This means that I add N twice, but after I take N away twice. Can I see
this, in Tor’s representation? We don’t see this fact, right?
83. [. . . ]
84. Teacher: so, which expression would a mathematician choose?
85. Voices: Ric’s?
86. Teacher: and actually, we studied the numerical expressions last year, there
should be a reason. . . Now, we have also letters, which stand for numbers.
87. [. . . ]
88. Observer: so, which one a mathematician would choose?
89. Voices: Ric’s.
90. Observer: why?
91. Alex: because you see that it is an expression and you get the result.
92. Observer: because you get the result? Is it for this reason?
93. Voices: no!
94. Ash: because you understand that the chosen number doesn’t matter.
95. Observer: ok, because it helps me to answer the original question.
96. Teacher: the original question was not to tell the result.
97. Observer: the two expressions are both correct, but Ric’s helps me to under-
stand why the chosen number doesn’t matter, why the chosen number disap-
pears.
The discussion concerning the two representations takes place around two main
issues: the equivalence between the two representations (epistemic dimension), and
the reason why mathematicians prefer Ric’s expression (teleological dimension).
We note several things: that the two issues often overlap; that in the first part the
students are more concerned with the equivalence rather than with the choice; and
that the teacher has to deal with complexity of the two overlapping motives of the
discussion. The teacher and the observer intervene so as to bring to the fore also the
teleological dimension, bringing back the students to the original aim of the activity
(to understand why the teacher can always guess the result, even without knowing
the chosen number). All along the discussion, thanks to the mediation of observer
and teacher (see for instance #59, 73), the students shift from evaluations in terms of
correctness and/or comprehensibility to evaluations in terms of efficacy in relation
to the initial task: as we see in lines #77, 94, Ric’s expression is preferred because
it is more useful according to the aim of the activity-understanding why the result is
always 10.
Discussion
In our opinion, the usefulness of a new analytical tool in mathematics education

must be proved through the actual and potential research advances and the educa-
tional implications that it can provide.
Research Advances
In the frame of our adaptation of Habermas’ construct, the distinction between epis-
temic rationality and teleological rationality allows us to describe, analyze and in-
terpret some difficulties in algebra (already pointed out in literature). These diffi-
culties can be accounted for by the students’ prevailing concern for rote algebraic
transformations performed according to systemic requirements of epistemic ratio-
nality over the needs inherent in teleological rationality (see equality E4). Moreover,
the distinction between modeling requirements and systemic requirements of epis-
temic rationality offers an opportunity to study the interplay between the modeling
requirements and the requirements of teleological rationality (see E7); we have also
seen that formalization and/or interpretations may be correct but not goal-oriented
(like in E2 and E4), or incorrect but goal-oriented (like in E5, E6 and E8).
Together with the other dimensions of rationality, communicative rationality al-
lows us to describe and interpret possible conflicts between the private and the stan-
dard rules of use of algebraic language, as well as the ways student try to integrate
them in a goal-oriented activity (see E3).
At present, we are engaged in establishing how the requirements of the three
components of rationality affect the phases of production and interpretation of al-
gebraic expressions. Further research work should be conducted to establish what
mechanisms (related to meta-cognitive and meta-mathematical reflections based on
the use of verbal language, see Morselli 2007) can ensure the control of epistemic
rationality and the intentional, full development of teleological rationality in a well-
integrated way. With reference to this problem, taking into account communicative
rationality in its intra-personal dimension, possibly revealed through suitable expla-
nation tasks and/or interviews) can reveal the role of verbal language (in its math-
ematical register: see Boero et al. 2008, p. 265) in the complex, dynamic relation-
ships between epistemic, teleological and communicative rationality. In particular,
previous analyses (see E3, E4 and Example 3) suggest not only that the request
(related to communicative rationality) to justify performed choices can reveal im-
portant lacks in teleological rationality, but also that the development of a kind of
personal “verbal space of actions” can aid a successful development of the activ-
ity (even if algebraic written traces are not satisfactory from the systemic-epistemic
rationality point of view, like in the case E3). The respective roles of the space of
verbal actions and of the space of algebraic manipulations should be investigated
on the teleological rationality axis. Here Duval’s elaboration about the productive
interplay between different registers in mathematical activities might be borrowed
to better understand and frame what students do (see Duval 1995). Also, the results
by MacGregor and Price (1999) could help highlight the relations, which emerged
from our data, between the production of verbal justifications and the effective use
of algebraic language to achieve the goal of the activity.
Educational Implications
We think that the analyses performed in the previous section can provide both teach-
ers and teacher educators with a set of indications concerning curricular choices as
well as their practical implementations.
Some of our general indications for curricular choices are not new in mathematics
education; we think that the novelty brought by Habermas’ perspective consists in
the coherent and systematic character of the whole set of indications.
First of all, the performed analyses suggest balancing (in the students’ eyes, ac-
cording to the didactical contract in the classroom) the relative importance (in rela-
tionship with the goal to achieve) of:
– production and interpretation of algebraic expressions, versus algebraic transfor-
mations;
– flexible, goal-oriented direction of algebraic transformations, versus rote alge-
braic transformations aimed at “simplification” of algebraic expressions.
These indications are in contrast with the present situation in Italy and in many
other countries, where classroom work is mainly focused on algebraic transforma-
tions aimed at “simplification” of algebraic expressions, and many simplifications
are performed by elimination of parentheses, thus suggesting a unidirectional way
of performing algebraic transformations. In the students’ eyes, the importance of the
formalization and interpretation processes is highly underestimated. The fact that al-
gebraic expressions are given as objects to “simplify” (and not as objects to build, to
transform according to the aim to achieve, and to interpret during and after the trans-
formation process in order to understand if the chosen path is effective and correct or
not) has bad consequences on students’ epistemic rationality and teleological ratio-
nality. As we have seen, many mistakes occur in the phase of formalization (against
the modeling requirements of epistemic rationality), and even when the produced
expressions are correct, frequently students are not able to use them intentionally to
achieve the goal of the activity (against the teleological rationality requirements).
We next consider practical implementation of such curricular choices, so as to
promote teaching algebra as an important tool for modeling and proving. Our anal-
yses suggest that students have a need for constant meta-mathematical reflection
(performed through the use of verbal language) on the nature of the actions to per-
form and on the solving process during its evolution. At present, the only reflective
activity in school involves checking the correct application of the rules of syntac-
tic transformation of algebraic expressions (thus only one component of rational
behavior—namely, the systemic requirements of epistemic rationality—is partly en-
gaged).
Specific indications for the management of classroom situations are exemplified

in the second part of the previous section: the construct derived from Habermas’
theory can work as a tool to interpret and guide students’ work. In particular, the
construct highlights the complex intertwining of the dimensions and, thus, may help
the teacher to guide students towards an efficient use of algebra as a proving and
modeling tool. The construct also highlights the dimensions that the teacher must
manage during the activity, and it may help them to plan didactical choices and on-
the-spot interventions. Indeed, we could even say that our construct is not only a
diagnostic tool (for interpreting students’ behavior and “detecting” their rationality)
but also a tool that may guide teachers’ choices and interventions, thus helping to
promote the development of rational behavior.
Specific tasks, such as the comparison of different strategies or solutions, ac-
companied by well-chosen questions, such as “which one would a mathematician
choose”, as in the 2nd task from the ongoing research project previously discussed,
seem to be effective in fostering the development of a rational way of behaving in
proving and modeling. More specifically, through such tasks the teacher may bring
to the fore also the teleological dimension, which may otherwise remain hidden. At
present, we are also planning and beginning to carry out long-term activities based
on the idea of story narration of guided proof construction as a didactical device to
promote awareness of the components of rational behavior in proving (see Boero et
al. 2010).
References
Anichini, G., Arzarello, F., Ciarrapico, L., & Robutti, O. (2003). Matematica 2001. La matematica
per il cittadino. Attività didattiche e prove di verifica per un nuovo curricolo di Matematica.
Lucca: Mattoni Stampatore. http://umi.dm.unibo.it/italiano/Matematica2001/matematica2001.
html.
Arcavi, A. (1994). Symbol sense: Informal sense-making in formal mathematics. For the Learning
Arcavi, A. (2005). Developing and using symbol sense in mathematics. For the Learning of Math-
ematics, 25(2), 42–48.
Arzarello, F., & Bartolini Bussi, M. G. (1998). Italian trends of research in mathematics education:
A national case study in the international perspective. In J. Kilpatrick & A. Sierpinska (Eds.),
Mathematics Education as a Research Domain: A Search for Identity (pp. 243–262). Dordrecht:
Kluwer.
Arzarello, F., Bazzini, L., & Chiappini, G. (1994). Intensional semantics as a tool to analyse al-
gebraic thinking. Rendiconti del Seminario Matematico dell’Università di Torino, 52(1), 105–
125.
Arzarello, F., Bazzini, L., & Chiappini, G. (1995). The construction of algebraic knowledge: To-
wards a socio-cultural theory and practice. In Proceedings of PME19, The 19th Conference of
the International Group for the Psychology of Mathematics Education. Recife, Brazil (Vol. 1,
pp. 119–134).
Balacheff, N. (1982). Preuve et démonstration en mathématiques au collège. Recherches en didac-
tique des mathématiques, 3(3), 261–304.
Bartolini Bussi, M. G. (1996). Mathematical discussion and perspective drawing in elementary
school. Educational Studies in Mathematics, 31(1–2), 11–41.
Bednarz, N., & Janvier, B. (1996). Emergence and development of algebra as a problem solv-
ing tool: Continuities and discontinuities with arithmetic. In N. Bednarz, C. Kieran, & L. Lee
(Eds.), Approaches to Algebra: Perspectives for Research and Teaching (pp. 115–136). Dor-
drecht: Kluwer.
Blum, W., & Niss, M. (1991). Applied mathematical problem solving, modeling, applications, and
links to other subjects: State, trends and issues in mathematics instruction. Educational Studies
Boero, P. (2001). Transformation and anticipation as key processes in algebraic problem solving.
In R. Sutherland et al. (Eds.), Perspectives on School Algebra (pp. 99–119). Dordrecht (NL):
Kluwer.
Boero, P. (2006). Habermas’ theory of rationality as a comprehensive frame for conjecturing and
proving in school. In Proceedings of PME 30, The 30th Conference of the International Group
for the Psychology of Mathematics Education. Praha, Czech Republic (Vol. 2, pp. 185–192).
Boero, P., Douek, N., & Ferrari, P. L. (2008). Developing mastery of natural language. In L. English
(Ed.), International Handbook of Research in Mathematics Education (pp. 262–295). New
York: Routledge.
Boero, P., Douek, N., Morselli, F., & Pedemonte, B. (2010). Argumentation and proof: A contribu-
tion to theoretical perspectives and their classroom implementation. In Proceedings of the 34th
Conference of the International Group for the Psychology of Mathematics Education (Vol. 1,
pp. 179–209). Belo Horizonte, Brazil: PME.
Carr, W., & Kemmis, S. (2005). Staying critical. Educational Action Research, 13(3), 347–357.
Chevallard, Y. (1984). Le passage de l’arithmétique à l’algébrique dans l’enseignement des math-
ématiques au collège: l’évolution de la transposition didactique. Petit X, 5, 51–94.
Dapueto, C., & Parenti, L. (1999). Contributions and obstacles of contexts in the development of
mathematical knowledge. Educational Studies in Mathematics, 39, 1–21.
Douek, N. (1999). Argumentative aspects of proving of some undergraduate mathematics students’
performances. In Proceedings of PME 23, The 23rd Conference of the International Group for
the Psychology of Mathematics Education. Haifa, Israel (Vol. 2, pp. 273–280).
Dreyfus, T. (1999). Why Johnny can’t prove. Educational Studies in Mathematics, 38, 85–109.
Duval, R. (1991). Structure du raisonnement déductif et apprentissage de la demonstration. Edu-
cational Studies in Mathematics, 22, 233–261.
Duval, R. (1995). Sémiosis et pensée humain. Berne: Peter Lang.
Duval, R. (2007). Cognitive functioning and the understanding of mathematical processes of proof.
In P. Boero (Ed.), Theorems in School: From History, Epistemology and Cognition to Class-
room Practices (pp. 137–161). Rotterdam, The Netherlands: Sense Publishers.
Furinghetti, F., & Morselli, F. (2009). Every unsuccessful solver is unsuccessful in his/her own
way: Affective and cognitive factors in proving. Educational Studies in Mathematics, 70, 71–
90.
Habermas, J. (2003). Truth and Justification. Cambridge (MA): MIT Press.
Hanna, G., & De Villiers, M. (2008). Proof and proving in mathematics education. Discussion
document. ZDM—The International Journal on Mathematics Education, 40, 329–336.
Harel, G., & Sowder, L. (1998). Students’ proof schemes: Results from exploratory studies. In
A. H. Schoenfeld, J. Kaput, & E. Dubinsky (Eds.), Research in Collegiate Mathematics Edu-
cation (Vol. III, pp. 234–283). Providence, RI: American Mathematical Society.
Hoch, M., & Dreyfus, T. (2006). Structure sense versus manipulation skills: An unexpected result.
In Proceedings of PME 30, the 30th Conference of the International Group for the Psychology
of Mathematics Education. Prague, Czech Republic (Vol. 3, pp. 305–312).
Kemmis, S. (2005). Knowing practices: Searching for saliencies. Pedagogy, Culture and Society,
13(3), 391–426.
Kemmis, S. (2006). Participatory action research and the public sphere. Educational Action Re-
search, 14(4), 459–476.
Linchevski, L., & Livneh, D. (1999). Structure sense: The relationship between algebraic and nu-
merical contexts. Educational Studies in Mathematics, 40(2), 173–196.
MacGregor, M., & Price, E. (1999). An exploration of aspects of language proficiency and algebra
learning. Journal for Research in Mathematics Education, 30, 449–467.
Malara, N., & Navarra, G. (2003). ArAl Project. Arithmetic Pathways Towards Favouring Pre-
algebraic Thinking. Bologna: Pitagora.
Morselli, F. (2007). Sui fattori culturali nei processi di congettura e dimostrazione. PhD Thesis,
Università degli Studi di Torino.
Morselli, F., & Boero, P. (2009). Habermas’ construct of rational behavior as a comprehensive
frame for research on the teaching and learning of proof. In Lin, Hsieh, Hanna, & De Villiers
(Eds.), Proceedings of the ICMI Study 19 Conference: Proof and Proving in Mathematics Edu-
cation (Vol. 2, pp. 100–105). Taipei, Taiwan: The Department of Mathematics, National Taiwan
Normal University.
Norman, D. A. (1993). Things that Make us Smart. London: Addison-Wesley.
Pedemonte, B. (2007). How can the relationship between argumentation and proof be analysed?
Pedemonte, B. (2008). Argumentation and algebraic proof. ZDM—The International Journal in
Mathematics Education, 40, 385–400.
Pierce, R., & Stacey, K. (2001). A framework for algebraic insight. In J. Bobis, B. Perry, &
M. Mitchelmore (Eds.), Numeracy and Beyond. Proceedings of the 24th Annual Conference
of the Mathematics Education Research Group of Australasia (MERGA). Sydney, Australia
(Vol. 2, pp. 418–425).
Radford, L., & Puig, L. (2007). Syntax and meaning as sensuous, visual, historical forms of alge-
braic thinking. Educational Studies in Mathematics, 66, 145–164.
Sorde Marti, T. (2004). Editor’s review of Jurgen Habermas’ Truth and Justification.
http://www.hepg.org/her/abstract/63.
Stacey, K., & MacGregor, M. (1999). Taking the algebraic thinking out of algebra. Mathematics
Education Research Journal, 11(1), 25–38.
Stylianides, A. J. (2007). Proof and proving in school mathematics. Journal for Research in Math-
ematics Education, 38, 289–321.
Weber, K., & Alcock, L. (2004). Semantic and syntactic proof production. Educational Studies in
Mathematics, 56, 209–234.
Theoretical Issues and Educational Strategies
for Encouraging Teachers to Promote
a Linguistic and Metacognitive Approach
to Early Algebra
Abstract After an overview of the studies which led to the rise of the study of
early algebra, we sketch our vision of this disciplinary area and of its teaching from
a linguistic and socio-constructive point of view. We take into account the teacher’s
role in the socio-constructive teaching process and stress the importance of reflect-
ing upon the teaching and learning processes in order to reshape the teacher’s ways
of being in the classroom. We dwell upon the strategies enacted and describe the
tools we have shaped: theoretical, for the enculturation of early algebra teachers, and
methodological, which aim at promoting their awareness and control of their action.
We conclude with some considerations about the value of the tools and modalities
we have used, as well as on the factors which determine their efficacy.
Introduction
The idea of giving space to early algebra at K-8 school level (pupils aged 4–14 years)
in association with a socio-constructive practice of teaching seems to be spreading
increasingly. This does not mean that syntactic activities typical of secondary school
should be anticipated at lower school levels, but rather that in elementary arithmetic
more room should be given to activities concerning numbers in relational terms,
so that pupils might be led to compare particular representations with other equiv-
alent representations of the same mathematical object, to detect analogies, and to
A. Cusi () · N.A. Malara · G. Navarra

Mathematics Department, University of Modena and Reggio Emilia, Modena, Italy
N.A. Malara
G. Navarra

484 A. Cusi et al.
generalize and identify properties. In other words, it would be appropriate to revise

arithmetic in a pre-algebraic perspective, with less emphasis on a typical algorith-
mic treatment and setting the ground for the development of algebraic thinking. The
aim is to get students to construct, starting from their early school years, a set of
experiences that make the study of algebra in its formal aspects, meaningful and
justified. In this way, the approach to algebra should be facilitated and the typi-
cal and widespread difficulties students meet when they access higher secondary
school, minimized. At the same time, they should be made aware of the potential of
algebraic language as a tool for thinking.
Several countries have nowadays included this theme, more or less explicitly,
in their national curricula, though this was done in the framework of their specific
cultural and educational features, and recently, comparative studies about the differ-
ent approaches and methods have been carried out (see for instance Kieran 2004;
Cai et al. 2005). Moreover, the introduction of early algebra in the different educa-
tional policies has been promoted by the indication of the British Department for
Education (DFE 1995) and, even more, by those of the National Council of Teach-
ers of Mathematics (NTCM 1998, 2000), together with the exploratory studies of
didactical implementations carried out at research level (to be dealt with later on
paragraph From Traditional Algebra to Early Algebra).
In Europe
In Europe, the initial, embryonic, teaching proposals in the spirit of early algebra
date back to the 70s, as a follow-up on curricula of two different and somehow
opposite tendencies: the psycho-pedagogical trends which underline the importance
of experience and discovery in learning, and the structuralist trend that suggested an
algebrization of the mathematical teaching contents.
In those years, naïve set theory comes to the fore in teaching and the concept
of a binary operation, with its properties, becomes the fundamental basis of the
arithmetic-algebraic area. On the one hand, this approach opens the way to a char-
acterization of the structures of the different number sets; on the other hand, it makes
both the concept of relation and the modelling processes central, due to the fact that
they tend to embed the concept of function among the objects of algebra.
With this approach, starting from primary school, in elementary arithmetic im-
portance is given to the relational aspects of numbers, to the symmetry of equality,
to the recognition of equivalent representations of numbers, to the valuing of arith-
metic properties for ordering numbers. At the same time, room is given to the study
of relationships in realistic contexts and with reference to different number sets,
with the joint detection of variable data for pairs of quantities.1
1 Inthe early ‘70 a noticeable project in Europe was the Hungarian project for primary school
directed by T. Varga, which envisaged activities of this type from the first two years of primary
school.
Theoretical Issues and Educational Strategies for Encouraging Teachers 485
The importance attached to the modelling processes led to a review of the teach-
ing of algebra—up to that moment mainly viewed in purely syntactic terms—as
well as to pose a higher attention to algebraic language as a representational tool.
Pioneer studies carried out by English scholars offer teaching experiences aiming at
generalization by means of realistic situations in several contexts, but also in situa-
tions within mathematics, often playful and even referred to proof (see, for instance
Bell 1976; Bell et al. 1985; Harper 1987).
From Traditional Algebra to Early Algebra
In actual practice, the new views on teaching and learning come into conflict with
the view of traditional algebra. For this and other reasons, diagnostic studies on
pupils’ difficulties were carried out, also taking into account issues related to both
modeling and interpretation of formal expressions. Classical studies in this re-
spect are those of Booth (1984), Kuchemann (1981), Kieran (1989, 1992), Lee and
Wheeler (1989), which point out that many difficulties and blocks in the learning
of algebra result from a teaching of arithmetic essentially centred on the aspects of
calculation and very little focusing on its relational and structural aspects.
This topic was debated during the ICME 6 Congress (Adelaide 1984), and a pro-
posal made to introduce relational, generalization-type and modelling activities in
primary school. However, an important step toward the constitution of early alge-
bra as a disciplinary area was made at ICME 7 (Quebec 1992). In that congress a
proposal was made for objectifying a new area of arithmetic teaching, called pre-
algebra, aimed at the development of ‘pre-concepts’ useful to algebra, i.e. advanced
arithmetic concepts, of a structural type, setting an experiential and conceptual basis
for connecting with more abstract and formal algebraic concepts (Linchevski 1995).
In those years, several scholars pointed out the importance for pupils to acquire
the ‘sense for symbols’ (Arcavi 1994) through a variety of activities which might
help them develop abilities, understanding and ways of feeling that could eventually
lead them to act in a flexible and instinctive way within a system of symbols, to move
around in wider or different systems of symbols and to co-ordinate interpretations
of formulae in various solution worlds (Arzarello 1991; Arzarello et al. 1993; Gray
and Tall 1993, Filloy 1990, 1991, Kaput 1991; Lins 1990). In US, debates about
the algebrization of the K-12 curriculum from kindergarten to secondary school are
undertaken (Kaput 1995).
At the ICME 8 Congress held in Seville (1996) Kieran characterized elementary
algebra through three types of activities2 at increasing levels of complexity, setting
generational activities at the first level, i.e. those through which the objects of alge-
bra can be constructed by linking meanings to experience (Kieran 1998). The second
half of the ‘90s was characterized by a considerable number of experiments on gen-
erational algebraic activities mainly addressing 11–13 year old pupils. Some studies
2 (1) Generational activities; (2) Transformational activities; (3) Global, at meta-level activities.
486 A. Cusi et al.
theorized socio-constructive models of conceptual development in algebra, in which

the influence of the classroom environment on learning, as well as the importance
of the role of the teacher are emphasized, in the framework of a view of algebra as
language (see for instance Da Rocha Falcão 1995; Meira 1990; Radford 2000).
Starting from the year 2000, the issues of early algebra became of increasing
interest in the International community, as shown by studies about early algebra at
the 12th ICMI Study ‘The future of the teaching and learning of algebra’ (Chick
et al. 2001), and other collective studies, such as the forum on early algebra at the
PME 25 (Ainley et al. 2001), the Special Issue on early algebra of the ZDM Journal
(Cai et al. 2005), the book ‘Algebra in Early Grades’ by Kaput et al. (2007) and the
international seminar “Pathways to Algebra” organized by D. Carraher in France
(Evron, June 2008).
All these studies mainly concern issues of implementation of innovative activities
in primary school and analyze pupils’ behaviours and learning. Several studies also
deal with the problem of suitable teacher training with focused interventions on their
professional development (see for instance Carpenter and Franke 2001; Carpenter et
al. 2003; Dougherty 2001; Blanton and Kaput 2001, 2002; Kaput and Blanton 2001;
Menzel 2001).
Our studies are in the line of this last trend and develop within the ArAl3 Project:
teaching sequences in arithmetic to favour pre-algebraic thinking (Malara and
Navarra 2003). In the next paragraph we give an overview of some its basic con-
cepts.
Early Algebra as a Meta-Subject and the ArAl Project

The linguist S. Ferreri (2006) wrote:
You get to the keywords of a discipline through a slow work of foundation of the basic
concepts of the respective disciplinary areas. Appropriating the meaning of words, of some
meaningful words, is a way to stabilize, conceptualize and master a specific knowledge do-
main, as its contents might otherwise remain not grasped. In fact, the word is viewed as
a permanent trace of a construction of knowledge, joint in memory to other pre-existing
words; as a capacity of making explicit a stage of the process of knowledge which is shap-
ing up. Words that show one’s degree of control over knowledge. Words, as portions of
knowledge that can represent itself.
From our perspective, the set of keywords of early algebra does not refer to a
single discipline (either arithmetic or algebra) and to its terms. It defines its limits
by starting from both disciplines but ends up assuming a different identity, mainly
distinct from either. We might talk about a meta-subject whose objects are not ob-
jects, processes, or properties of the two subjects, but rather the genesis of a unifying
language. A meta-language, as such.
3 The term ‘ArAl’ is a synthesis of the terms ‘Arithmetic’ and ‘Algebra’. The Project started in 1998
on the basis of our previous studies (Malara and Iaderosa 1999) at secondary school level (grades
6–8, pupils aged 11–14) and it is designed for primary school in the perspective of a continuity
between the two school levels.
In our view, early algebra is based on these fundamental principles:

• The anticipation of generational pre-algebraic activities at the beginning of pri-
mary school, and even before that, at kindergarten, to favour the genesis of the
algebraic language, viewed as a generalizing language, while the pupil is guided
to reflect upon natural language.
• The social construction of knowledge, i.e. the shared construction of new mean-
ings, negotiated on the basis of the shared cultural instruments available at the
moment to both pupils and teacher.
• The central role of natural language as main didactical mediator for the slow
construction of syntactic and semantic aspects of algebraic language. Verbaliza-
tion, argumentation, discussion, exchange, favour both understanding and critical
review of ideas. At the same time, through the enactment of the processes of
translation, natural language sets up the bases for both producing and interpreting
representations written in algebraic language.
• Identifying and making explicit algebraic thinking, often ‘hidden’ in concepts
and representations in arithmetic. The genesis of the generalizing language can
be located at this ‘unveiling’, when the pupil starts to describe a sentence like
4 × 2 + 1 = 9 no longer (not only) as the result of a procedural reading ‘I multiply
4 times 2, add 1 and get 9’, but rather as result of a relational reading such as ‘The
sum of the product of 4 times 2 and 1 is equal to 9’; i.e. when pupils talk about
mathematical language through natural language and do not focus on numbers,
but rather on relations, that is on the structure of the sentence.
In our project we claim that the main cognitive obstacles to the learning of alge-
bra arise in unsuspected ways in arithmetic contexts and may impact on the develop-
ment of mathematical thinking, mostly due to the fact that many students only have
a weak conceptual control over the meanings of algebraic objects and processes.
Our aim is to make teachers aware and caring about this situation and provide
them with instruments that enable them to design and implement powerful inter-
ventions to face it. In order to be able to do that, the teacher needs to understand
how, and most of all why, the construction of mathematical concepts needs to be
supported by a setting made of solid linguistic and methodological bases, but also
social and psychological ones. Therefore teachers must be able to construct new
meta- competencies and empower their sensitivity in grasping the deep mutual re-
lations between the two subjects, and the seeds of algebraic thinking underlying
arithmetic concepts and representations.
In order better to frame the problematic aspects connected with the role of the
teacher, we briefly discuss this issue in the more general perspective of socio-
constructive teaching.
Socio-Constructive Teaching and Teacher Training

In the teaching of mathematics the socio-constructive model is spreading, since it is
viewed as suitable to educate students (mainly aged between 6 and 14) to work col-
lectively as well as to favour their acquisition of flexibility in thinking. According to
488 A. Cusi et al.
this model, teachers should start their action by devolving to students purposefully
designed problem situations that may bring about the emergence of particular math-
ematical concepts and properties. The core of the model is the view of students as
makers of their own knowledge: it develops through argumentation and exchanges
of ideas, up to the collective systematization of the results obtained and a reflection
upon meanings and role of those results.
The whole-class mathematical discussion plays a central role in the model. In
order to be able to fulfill the task, the teacher should master notions and abilities
that go beyond the mere knowledge of the discipline:
• from a social viewpoint, to be able to create a good interactional context, by
stimulating and guiding the argumentative processes (mediating argumentation
in the words of Schwarz et al. 2004) easing communication, listening, evaluating,
and a capacity of producing a counter-argumentation (Wood 1999);
• to activate socio-mathematical norms that lead to check the acceptability of a
solution, to evaluate different solutions, to appreciate the quality of a solution
(Yackel and Cobb 1996);
• to determine the direction of the discussion in its various phases, filtering stu-
dents’ ideas, so that their attention may be focused on the contents of the teacher’s
views as more relevant and meaningful (Gamoran Sherin 2002);
• to harmonically enact modalities (Anghileri 2006) such as: reviewing (focusing
pupils’ attention on aspects of the activity that may favour the understanding of
the underlying mathematical ideas); restructuring (encouraging students to reflect
upon and clarify to themselves what they have understood, in order to favour both
development and strengthening of mathematical meanings); re-phrasing of stu-
dents’ utterances (re-formulation of what one or more pupils claimed to highlight
and clarify the argumentative processes developed in the classroom); using prob-
ing questions (posing questions in order to investigate on students’ statements,
with the aim of leading them to clarify what they said and favour a development
of their thoughts);
• to involve pupils in metacognitive acts (transactive utterances in the terminology
of Blanton et al. 2003), to enable them to internalize collective argumentative
processes.
The Role of the Teacher’s Reflection
Several researchers underline the value of the teacher’s critical reflections on the
classroom-based processes (Mason 1998; Jaworski 1998; Schoenfeld 1998) and
most of all, of the practice of sharing these reflections (Borasi et al. 1999; Ponte
2004; Jaworski 2003; Malara 2003, 2005; Malara and Zan 2002, 2008; Potari and
Jaworski 2002) for the acquisition of the above described competences.
In particular, Mason (2002), by helping teachers acquire the capacity of carefully
observing themselves in class-based action, suggests the constant practice of the
‘discipline of noticing’, recommending that reflections be shared among colleagues
in order to be validated. Jaworski (2004) points out the efficacy of “communities of

enquiry” (mixed groups made by teachers and researchers) highlighting the fact that
participating in these groups brings about a taking on of identity by the teacher.
On this basis, the hypothesis that is outlined requires a change of perspective by
the teacher. A teacher should re-learn to manage socio-cognitive processes (expe-
rience in the classroom), drawing on the theoretical frameworks proposed to them,
comparing these proposals to their own epistemology, thus being fruitfully and sig-
nificantly enriched in both their own culture and their work in the classroom. In
this way, the teacher may avoid the feeling of powerlessness in front of paradigms
that are too abstract or self-referential to become reliable keys for reading their own
experience, or rather paradigms for an intervention in their own practice.
This hypothesis holds for trainee teachers as well, because it refers to the inter-
twining between methodological and mathematical aspects: a background in math-
ematics is a necessary condition, but not a sufficient one, to become a good mathe-
matics teacher.
Mason (2002) starts up his text on the discipline of noticing with the following
maxim:
I cannot change others; I can work at changing myself. (Page v)
In many respects, the latter strictly links to our discourse. As a matter of fact,
many teachers believe that they can intervene and change their pupils without having
tried consciously to change themselves. In other words: without putting themselves
critically in front of their own practice, and investigating it. Mason also writes:
Working to develop your own practices can be transformed into a systematic and method-
ologically sound process of ‘researching from the inside’, that is, of researching yourself.
(Page xii)
Hence it is a matter of starting up a continuous reflection upon oneself making

use of theoretical supports that may bring about the awareness that a continuous
transformation is needed. The final aim is to get to go beyond the idea that some lit-
tle adjustments (such as changing the textbook, using new technologies, or attending
some training courses) can be enough to produce effective changes in pupils’ learn-
ing.
But, in order to be effective, transformation requires an essential condition: that
we train ourselves to understand in which directions transformation should be pro-
moted. A fruitful exchange between theory and practice may bring the teacher to de-
velop capacities at two levels: at the first level, to grasp signs in all that contributes
to define their own condition, both in the field—in their activity in the classroom—
and in the construction of their own theoretical instruments, at a second level, to
elaborate on the grasped signs so that these signs become part of the foundation of
their cultural background.
The development of a capacity to grasp signs is achieved only through the
teacher’s increasing awareness in learning to transform thousands of occasionally
noticed things into a tool of one’s individual methodology, deriving from a relation
between the capacity of noticing, the motivation to intervene and the acquisition of
instruments that suggest how to intervene.
Concerning the capacity of noticing, Mason adds:
490 A. Cusi et al.
Every practitioner, in whatever domain they work, wants to be awake to possibilities, to

be sensitive to the situation and to respond appropriately. What is considered appropriate
depends on what is valued, which in turn affects what is noticed. . . . noticing what children
are doing, how they respond, evaluating what is being said or done against expectations and
criteria, and considering what might be said or done next. It is almost too obvious even to
say that what you do not notice, you cannot act upon; you cannot choose the act if you do
not notice an opportunity. (Page 7)
Hence the question is: What should the teacher notice? Who would teach him or
her to notice this that?
What we maintain is: the teacher is firstly a mentor of himself, through a con-
tinuous engagement, with the awareness of the support of a reference map as well
as of suitable stimulating tools (continuously and critically reviewed, as we will see
in the next paragraphs), which allow a teacher to start up an exploration of a cul-
tural baggage which is certainly familiar to them, but at the same time, needs to be
re-considered from different viewpoints, through a process that will gradually lead
them to a forma mentis that is profoundly different to that of the previous stage.
In our case, the point of arrival will be a re-reading of one’s conceptions with re-
lation to arithmetic and algebra. The teacher must be active protagonist of his own
development in his approach to early algebra.
In this respect, the main tool our project (ArAl) refers to, is a set of theoretical
constructs, partly drawn from other constructs and partly original, organized in a
Glossary, which we will present here.
The Role of the ArAl Glossary in Teacher Training

The ArAl Glossary is a reference system that allows the teacher gradually to reach
an overall view of early algebra, which merges theory and practice, by approaching
a linguistic view of algebra, within which a convincing control over its meanings
can be constructed together with the pupils.
Each term of the Glossary is a self-sufficient entity, so to speak. The text that
describes it includes some other key terms, the set of which constitutes a more or
less wide Net.
A term of the Glossary may be able to avail itself of a very numerous Net, but be
quoted in few Nets. Vice versa, another term might have a Net with few links, but be
present in many other Nets. Each term, therefore, depending on the numerousness
of its Net, as well as on that of its occurrences, locates the teacher within a double
process of conceptual deepening and extension: deepening of the term through the
relations among the key-terms which appear in its definition; extension, since each
of them is a potential stimulus to read its definition.
The terms of the Glossary may be grouped within five areas:
• GENERAL: didactical mediator, Opaque/transparent (referred to meaning), Re-
lational thinking, Process/Product, Representing/Solving, . . .
• LINGUISTIC: Arguing, Algebraic babbling, Language, Letter, Metaphor, Para-
phrase, Semantic/Syntax, Translating, . . .
• MATHEMATICAL: Formal coding, Additive form, Canonical/Non canonical

form, Multiplicative form, Mathematical Phrase, Function, Unknown, Pseudo
equation, Relation, Equal sign, Structure, Variable, . . .
• SOCIO-DIDACTICAL: Sharing, Collective exchange, Didactical contract, Dis-
cussion, Social mediation, Negotiation, . . .
• PSYCHOLOGICAL: Affective-emotional interference, Perception, . . .
Some terms of the Glossary have more numerous Nets and Occurrences than
others. This attaches to them a status of strong representativeness in the definition
of early algebra from our point of view and enables us to compose a sort of manifesto
through them (in bold case in the text):
The theoretical framework of early algebra supports the hypothesis that
students’ weak control over the meanings of algebra derives from their
ways of constructing arithmetic knowledge from the early years of primary
school.
Algebra should be taught as a new language, one gets to master—through
a set of shared social practices (collective discussion, verbalization, ar-
gumentation)—with modalities that are analogous to those of natural lan-
guage learning: starting from its meanings (semantic aspects) and setting
them gradually in their syntactic structure (a process we called algebraic
babbling).
Crucial elements in this respect are metaphors, didactical mediators in
the achievement of meanings, during the conceptual progression towards
generalization and modeling.
In this view, natural language becomes the most important mediator in the
student’s experience and the main instrument of representation through
which they can illustrate the system of relations (additive and multiplica-
tive ones at the beginning) among elements in a problem situation, shifting
focus from the product to the process, and inducing a translation of the
process itself into a mathematical sentence.
In this way, attention is shifted from the arithmetic objective of solving, to
the algebraic one of representing. At the same time, mediators favour
the achievement of the use of letters, seen as unknown—easier to be
achieved—of indeterminate and of variable.
The structure of the Glossary is outlined so that pre-defined approaches to the
included terms are not sketched. The teacher finds a plurality of routes to be explored
autonomously, depending on the modalities of their approach to early algebra, their
own background, the age of the pupils, the themes they want to deal with, their
curiosity and so on. The exploration of this plurality is an individual adventure,
which originates from the interest in deep concepts and theoretical constructs and
depends on how the teacher decides to interact with it. Its use offers the immediate
contextualization of concepts and the possibility to connect them with other more
familiar ones through a net of internal cross-references. By doing so, the teacher
gradually builds a progressively more articulated conceptual map of the theoretical
frame of the project.
492 A. Cusi et al.
We briefly describe some crucial elements of such a frame concerning the

algebraic-linguistic aspects. We shall therefore leave out the psychological, edu-
cational, social and the more general mathematical aspects involved. To begin with,
we concentrate on the exploration of the concepts that concern so-called algebraic-
babbling.
Algebraic Babbling
Pre-requisite to acquire control over the syntactic aspects of a new language is a slow
and in-depth acquisition of a semantic control. As we know, a child, while learning
natural language, gradually appropriates its meanings and rules, and progressively
develops them through adjustments and imitations, up to the deeper knowledge he
gets to in his school years, when he learns to read and reflect upon grammar and
syntax. In the same way, mental models typical of algebraic thinking should be con-
structed from the early years of primary school, progressively building up in pupils
algebraic thinking as both tool and object of thinking, in a strict intertwining with
arithmetic, starting from the meanings of the latter. For this reason, it is necessary
to build up an environment able to stimulate informally the autonomous elaboration
of formal coding for sentences in natural language, discussing them with the whole
class and gradually producing a playful, experimental and continuously re-defined
appropriation of the new language. The rules of this language are then located in
a didactic contract, which tolerates initial moments of syntactic ‘promiscuousness’.
This process of construction/interpretation/refinement of ‘draft’ formulas is what
we call ‘algebraic babbling’.
In the Glossary, for ‘algebraic babbling’ one finds some excerpts of class tran-
scripts and the related comments. An example follows.
A fourth grade class (9 years-old) is exploring problem situations asking to iden-
tify the relations existing between two quantities. In one of these situations, involv-
ing variable quantities of two different types of biscuits, i.e. sponge biscuits and
chocolate cookies, as usual pupils represent the relations between the two quantities
in natural language, getting to the following correspondence rule, after selecting
different formulations:
‘the number of sponge biscuits is 1 more than twice the number of choco-
late cookies’.
The next step is to translate the sentence in algebraic language. Individual work
leads to the following proposals:
(a) 1×2
(b) a + 1 × 2 (a = n. of sponge biscuits)
(c) a×2+1
(d) sv + 1 × 2 = a
(e) sv = st + 1 × 2
(f) a =b×2+1
(g) a × 2 + 1 = b (a = n. of choc. cookies)
(h) (a − 1) × 2
The comments made on the transcripts show that, beyond the teacher’s wishes
and actions, the choice of the most correct sentence(s), of those that better fit with the
problem situation, of the clearest ones (in the Glossary the pair transparent-opaque
is introduced to indicate this) depends upon the level of algebraic babbling achieved
by the class through previous experiences. In these experiences, the negotiation of
meanings led pupils to share them through a social construction of meanings. The
pupils’ skills determine the quality of their choices of a sentence and its related
justifications, not depending on the teacher’s legitimate expectations.
Suppose we explore the Glossary from a linguistic perspective: it makes refer-
ence to language as well as to its semantic and syntactic aspects.
Algebraic Babbling → Algebra as a Language
A widespread belief about pupils is that solving a problem means identifying the
result. This implies that their attention is focused on the operations. They should
rather learn not to worry too much about the result, and therefore about the search for
the operations that lead to it, and move from the cognitive to the metacognitive level,
where the solver interprets the structure of the problem and represents it through
algebraic language. Algebra thus becomes a language to describe reality, and not
only: it amplifies understanding.
A process of this kind occurs very slowly, and through progressive steps, with
an intertwining of continuities and ruptures between the different levels of knowl-
edge. Traditional arithmetic teaching tends to encourage a mental attitude aiming at
immediately searching for the tools (operations) to identify the answer (the result).
This attitude is also induced by the formulation of the task in some standard word
problems like the following (pupils aged 6):
On a tree branch there are 13 crows. 9 more arrive and 6 fly away. How many crows are
left?
Contrasting with the question above, there is another, extremely different, one:
Represent the situation in mathematical language, in order to find the number of the crows
left.
The first version emphasizes the search for the product (16), the second one the
search for the process (13 + 9 − 6), i.e. of the representation of the relations between
494 A. Cusi et al.
the involved elements. This difference links back to one of the most important as-
pects of the epistemological gap between arithmetic and algebra: while arithmetic
requires an immediate search for the solution, algebra postpones this and starts with
a formal transposition of the problem situation from the domain of natural language
to a specific representation system (think about a problem that can be solved by an
equation).
The perspective of an approach to algebra as a language, in a continuous back and
forth of thinking from arithmetic to algebra and vice versa, fosters more effective
teaching with pupils aged between seven and fourteen, characterized by negotiation
and explicit statement of a didactic contract for the solution of problems, based on
the principle: “first represent, then solve”. This seems extremely promising for deal-
ing with one of the most important key elements in the conceptual field of algebra:
the transposition in terms of representation from natural language, in which prob-
lems are either formulated or described, to the formal-algebraic one in which the
relations and later the solution are translated.
Exploring the items related to ‘algebra as a language’ one meets the pair ‘repre-
senting/solving’, and this leads to another key point in our theoretical framework:
the syntax and semantics of mathematical language. In the approach to these themes,
the Glossary brings the reader to meet a very important character of the ArAl project:
Brioshi.
Algebraic Babbling → Syntax, Semantics → Brioshi
An aspect that strictly links to that of representation is the respect of the rules in the
use of a language, even more necessary when one deals with a formalized language,
given that the symbols used are extremely synthetic.
In everyday life the respect of linguistic rules is gradually learned through their
use, by trial and error; this is favored by the family environment and by the enlarged
social one, as well as by school, through a reflection upon orthography, grammar,
syntax, i.e. upon the structural aspects of a language.
In the learning of mathematics, the rules are generally ‘given’ to pupils, thus
losing their social value as support to understand a language, and hence to share it,
as a tool for communication. Similarly to what happens in linguistics, the syntax
of the mathematical language concerns the structure of the sentence, the elements
that compose it and the formal procedures that express the relations between the
involved quantities—either known or unknown—even in a passage made of several
sentences.
It is thus necessary to lead students to understand that they are appropriating
a new language and that, as all other languages elaborated by the human kind, it
is a system of arbitrary finite symbols, combined according to precise rules. But,
whereas pupils have interiorized the set of rules related to spoken language since
their birth, and understand that respecting them, is functional to communication, it
is rather difficult for them to transfer this peculiarity to mathematical language. To
avoid this key element and highlight the value of written language for communica-
tion, the teacher proposes an exchange of messages in arithmetic-algebraic language
with either real or virtual classes, engaged in the solution of the same problem sit-
uation. Brioshi, a virtual Japanese pupil, variably aged depending on the age of his
interlocutors, knowing only his mother tongue (and therefore not able to communi-
cate using languages that differ to his own), but competent in the use of mathemati-
cal language, is the algebraic pen friend, with whom they need to communicate (for
a wider discussion, see Malara and Navarra 2001).
Pupils get to learn that, like any language, mathematical language also has its
own grammar and a syntax, i.e. a set of conventions, that enable us to construct
sentences correctly. It has a syntax, which provides the conditions—i.e. the rules—
to decide whether a sequence of linguistic elements is ‘well-formed’ (for example,
sentences like ‘9 + +6 = 15’, or the classic chain of operations added one after the
other, like ‘5 + 3 = 8 : 2 = 4 + 16 = 20’) are syntactically wrong. It has a semantics,
which enables one to interpret symbols—within syntactically correct sequences—
and subsequently decide whether the expressions are true or false (for instance, the
sentence ‘1 + 1 = 10’ is either true or false depending on the representation base,
which can be either 2 or 10).
In the perspective we are considering, translating from natural language (or
graphical, or iconic) to mathematical one, and vice versa, is one of the most fertile
territories where reflections upon mathematical language can be developed. Trans-
lating, in this case, means interpreting and representing a problem situation through
a formalized language or, on the contrary, recognize in a symbolic expression, the
situation it describes.
In the learning of mathematics, where the exchange between verbal language and
mathematical language is continuous, it is necessary to activate in pupils, on the one
hand, a control over expressive registers and, on the other hand, the meta cognitive
skill to understand how syntactic transformations of formal expressions condense
thinking processes that can be hardly realized through natural language.
Meeting Brioshi is very important to help very young pupils approach the idea of
a language of mathematics that allows them to represent and communicate proce-
dures and relations. All the teachers who either collaborated with the ArAl project
or adopted the project’s teaching sequences, were able to test the power of Brioshi’s
metaphor, independently on the pupils’ age.
Further important pre-algebraic key-points are faced with Brioshi. One of these
is the form of representation of a number and, related to this, the equality sign.
Brioshi → Canonical/Non Canonical form of a Number → ‘=’
Facing the ontological question: ‘Is [3 × (11 + 7) : 9]2 a number?’ usually both
students and trainee teachers answer in the same way: “No, they are operations”, “It
is an expression”, “They are calculations”. At times, someone dares to say “Can we
say it is the representation of a number?”. In order to promote a reflective attitude on
496 A. Cusi et al.
the answer, we make use of analogy, by examining how people are usually denoted.
Each person has their first name but can also be denoted by the relationship that
links them to other individuals or to the environment (Daughter of . . . , Sister of
. . . , Owner of the dog . . . , Living in So-and-so Street, etcetera). Each of these
expressions adds information not included in the first name and permits a deeper
knowledge of the person.
The situation is similar with numbers: each number can be represented in many
different ways, through any odd equivalent expression. Among these representa-
tions, one (for instance 12) is its name, called a canonical form of its representation,
all the others (3 × 4, (2 + 2) × 3, 36/3, 10 + 2, . . .) are its non canonical forms, and
each of them will make sense in relation to the context and the underlying process.
This experience enables older pupils to answer the initial question we left unan-
swered: [3 × (11 + 7) : 9]2 is one of the many non canonical forms of the number 36.
Being able to recognize and interpret these forms builds up the semantic basis for
the understanding of algebraic expressions like p − 4q, ab, x 2 y, k/3. Moreover, the
concept of canonical/non canonical form has crucial implications when it is about
harmonizing the meanings of the equality sign as directional operator and indicator
of equivalence. The process through which these skills are constructed is very long
and is to be developed along the whole course of the first school years.
In this long paragraph we examined some fundamental concepts of our approach
to early algebra, going through one of the possible routes for exploring the Glossary.
We will now move on to the side of practice, with reference to both (i) the class-
room based action, and (ii) the teacher’s reflection upon the latter. We will go on
to analyze a second—and similarly fundamental—instrument: the methodology of
multi-commented transcripts.
The Multi-Commented Transcripts Methodology (MCTM)

Earlier in the text, we underlined how noticing and critically studying and reflecting
upon classroom-based processes allow the teacher to become aware of the dynamic
processes characterizing the teaching activity, as well as of the variables that deter-
mine them.
MCTM has been conceived to promote this awareness. Its aim is to lead teachers
to acquire an increasing capacity of interpreting the complexity of class processes
through the analysis of the micro-situations that constitute them, to reflect upon
the effectiveness of one’s own role and become aware of the effects of one’s own
micro-decisions.
We implement the MCTM in in-service teacher training courses.4 An initial pe-
riod of study (8–10 weekly 3-hour-long meetings in which the theoretical frame-
work, general literature and the Glossary are presented and discussed), is followed
4 These courses are generally two-three-years-long and organized by Universities jointly with local
training agencies. The teachers involved come from both primary and secondary schools, depend-
ing on the specific course.
by a second period characterized by laboratories involving small groups of teachers

(two to four units). For each group, a mentor5 coordinates the work, providing the
link with other researchers involved in the project: a little community of enquiry is
thus constituted. The small groups mutually interact in specific moments devoted to
collective exchange.
In the laboratories, participants plan lengthy teaching experiments (at least five
teaching sessions lasting 90–120 minutes) designed drawing on the ArAl Units6 and,
most importantly, carry out a complex critical analysis of the processes enacted by
the use of MCTM. The latter activity is the core of the educational process.
MCTM is structured in a sequence of phases. Teacher-experimenters make audio-
recordings of lessons on topics they previously chose in agreement with researchers
and, after transcribing them in digital text version (the ‘transcripts’) and filling them
with comments and reflections, they send them to their mentors, for further com-
ments. Then, the latter send the transcripts to the authors, to other teachers engaged
in similar activities and to other researchers. Often the authors intervene back in the
cycle, making comments about the comments or rather inserting new ones.
Therefore, lesson after lesson, the multi-commented transcript (MCT)7 of the
enacted didactical process is objectified.
An Example of MCT
We report here a short excerpt from an MCT referred to an activity on the non-
canonical representations of a number carried out in grade 3. It is centered around
an iconic representation linked to a problem situation that asks pupils to represent
the numerousness of marbles in boxes set out as follows: one line of eight boxes,
each containing two marbles and another line of eight boxes, each containing five
marbles. Pupils are supposed to make explicit the way of counting the total num-
ber of marbles through a numerical expression. Many proposals are written on the
blackboard, including: 16 + 40 (made by Andreina) and (2 × 8) + (5 × 8) (made by
Giovanni), to which the following excerpts refer.
1. Melania Andreina’s translation is opaque for me.
5 The mentor is usually a teacher-researcher, a typical figure of the Italian research for innovation
(Malara and Zan 2002).

6 The ArAl Units are monographic booklets about experiences in early algebra, which can be seen
as models of processes of teaching arithmetic in an algebraic perspective, to be carried out in the
long term. They are the result of the progressive refinement of numerous experimentations and
are fine-tuned on the basis of cross-analyses of records of class activities, and of comparisons of
reflections between teachers, mentors and researchers. They are not tools for immediate use in the
classroom, but require a theoretical study, before being put into practice.
7 This phase requires a hardly quantifiable and lengthy time both for teachers and for the other
actors involved. An average of three hours for commenting upon each lesson may be estimated.
In addition, time devoted to the sessions of joint analysis, involving teachers of the group and
researchers working on a specific teaching sequence, should be considered. This time is generally
acknowledged by schools and certified by the University.
498 A. Cusi et al.
2. Teacher What do you mean?

3. Bruno It is opaque because Andreina has already found the number of marbles.
4. Chiara It was not our job to find it, we had to write a translation for Brioshi.
She has almost solved the problem.
5. Bruno It’s true! She found the product and not the process! Giovanni’s trans-
lation is the right one, since it is more transparent.
Melania (line 1) refers to Andreina’s representation as opaque because it hides the
counting process and does not respect the given task, as Bruno (line 3) and Chiara
(line 4) clarify after the teacher’s prompt (line 2). Bruno (line 5) feels the need to
refer to other terms from the Glossary to point out further the problematic aspects
of Andreina’s expression as opposed to that produced by Giovanni.
The excerpt clearly shows how the pupils have appropriated the terms of the ArAl
Glossary (translation for Brioshi, opaque/transparent, process/product) and how ap-
propriately and consciously they use them. These expressions at meta level indi-
rectly show that the teacher has stabilized and conceptualized—and now masters—
these constructs: not only this, she is also competent in adopting them in her class-
room practice.
Multi-commented transcripts are important instruments from four points of view:
• Diagnostic: transcripts provide the mentor, and therefore the whole team, with an
overview of the teacher’s teaching action and enable a check on the coherence
between teaching practice and reference to the theory at stake (both mathematics
and mathematics education).
• Formative: through comments, they enable the teacher to develop competences
and sensitiveness and hence to improve the overall quality of his teaching action.
• Evaluative: by making clear both coherence and inconsistencies of the teaching
action, the transcripts provide both the teacher and the researchers with elements
that can empower the effectiveness of the interventions in the respective areas,
and make it possible to detect teachers’ attitudes and cultural backgrounds.
• Social: the transcripts promote a sharing of knowledge, since they are sent out
to the other components of the group and a periodic reflection upon the most
significant excerpts is undertaken. Moreover, each teacher, comparing their own
progress on the realization of a certain part of the teaching sequence to that of
other colleagues, can identify important distinctive elements and reflect upon both
effectiveness and limits of their own work.
Taking part in the MCTM the teacher questions their action at different levels:
• transcription promotes a posterior reflection upon the activity and how it has been
carried out and guided;
• writing down the comments fosters a critical reconstruction of the activity through
an interpretive effort which has a high formative value;
• their analysis by mentor and by other researchers—focusing on both mathemat-
ical and methodological aspects—leads to a re-elaboration of the activity with a
significant impact on both teaching practice and teacher training.
MCTM enables the teacher critically to connect three key points: their own con-
ception of the mathematical knowledge at stake (and of mathematics itself); the
conflict induced by the meeting—clash with the teaching modalities enacted by col-
leagues and the results they achieved; the mediation between these two key points
produced by the collective exchange and the dialogic relationship with researchers.
The comments (almost line by line) make clear the overall analysis which deals
with aspects related to content (the approach to the exploration of the problem,
the mathematical aspects developed, the objectives achieved or missed out, . . .),
aspects related to communication and language used (formulation of the questions
posed, interlocutory expressions used, operative directions suggested, . . .), issues re-
lated to the control over pupils’ participation (number and type of interventions), as
well as to the didactical contract established (attitudes induced in pupils and socio-
mathematical norms in the classroom). Due to the fine and multi-faceted analysis
the teacher’s action is fully explored and, consequently, they become aware of their
own oversights, missed chances and faults. Being constantly observed, the teacher
is induced to have a better control over their actions and attitudes and gradually gets
to change their way of being in the classroom.
From the Comments to a Classification of Attitudes
As we repeatedly underlined, the aim of the project is to train metacognitive teach-

ers. In this view, the high number of comments in the transcripts (by the year 2009
nearly 4000 comments) offers a very meaningful overview of the teachers’ attitudes
towards their own activity in the classroom and their own capacity of reflecting upon
it afterwards. The analysis of comments, in turn, brings about powerful feedback on
training interventions.
Let us assume two categories, both related to the teacher and their ‘being
metacognitive’:
(a) in their action in the classroom;
(b) in the posterior reflection upon their action in the classroom.
The two categories enable us to identify four different types of attitudes, summa-
rized in the table (Fig. 1).
The four deriving profiles may be defined as follows:
MA MC The teacher effectively drives the class discussion towards the mathe-
matical objective, encouraging pupils to explore the problem situation;
they try to take into account pupils’ interesting and unforeseen cues and
to find an equilibrium between their own objectives and the stream of
the pupils’ thoughts.
When the session is transcribed, the teacher keeps detached from
the events and this leads his/her to write either theoretical or practical
reflections both with relation to mistakes—stiffness, wrong interpreta-
tions—and to fruitful interventions.
500 A. Cusi et al.
Teachers
MC ¬MC
metacognitive in non metacognitive in
comments comments
MA They stimulate meta They stimulate meta

cognitive attitudes; cognitive attitudes;
Metacognitive in
they comment upon they insert few
action in the classroom
transcripts in depth. meaningful comments
Teachers
in transcripts.
¬MA They do not stimulate They do not stimulate

meta cognitive meta cognitive
Non meta cognitive in
attitudes; transcripts attitudes; they insert in
action in the classroom
are filled in with transcripts few
posterior reflections meaningful comments.
upon missed out
chances.
Fig. 1 Types of attitudes at metacognitive level
MA ¬MC The teacher manages the activity with a global self-confidence, thus fa-
voring a reflection upon the processes, the coherence in pursuing the
objective, the exchange among peers. In the transcription phase, the
teacher does not detect stimulating moments for reflection and views his
task as essentially completed in the classroom activity. In other words,
the phase of in-depth analysis and generalization of behaviors with re-
lation to theoretical issues seems to be weak.
¬MA MC The teacher–due to little expertise, low control over mathematical con-
tents, difficulties in managing the discussion—is not able to guide the
activity in a productive way, stimulating the attention of the class to-
wards a meaningful reflection. At the moment of transcription, they re-
alize how weak his guidance was (a feeling often emerged at the very
moment of its appearance) and the comments point out this awareness,
often joint to a request of help addressing the mentor.
¬MA ¬MC The teacher keeps the class working on a little stimulating activity, con-
stantly playing a central role and driving the pupils towards a funda-
mentally a-critical acquisition of ‘compulsorily reachable’ contents. In
the phase of transcription, the comments refer to marginal aspects, con-
cise remarks on pupils’ personality, superficial clarifications.
In general terms, the type MA MC represents a desirable model and possible target
for any formative process. In fact the teacher is often not used to an exchange in
which they play the role of the student, with hardly predictable consequences.
This is confirmed by a teacher who, after a two-year long collaboration writes
down in his reflections:
At the beginning the Glossary and the ArAl Units were my points of reference. I started to
produce my first transcripts rather timidly. Through these, and helped by the comments of
my mentor and another researcher, I was able to analyze my behaviours and those of my
pupils in teaching and learning situations, focusing on both positive and negative aspects,
and to reflect upon some crucial points, mainly related to the management of the mathemat-
ical aspects and to communication. And that was the beginning of an itinerary.
Often, an MA ¬MC -type attitude is not a sort of ‘limited disposition’ to reflec-

tion, but rather fruit of the lack of meaningful stimuli towards the direction Mason
talks about, of a constant search within a process and towards oneself, which may
leave deep and long-lasting traces at the professional level, contributing to the con-
struction of a way of being that will become the foundation of a continuous path of
change”.
The definition of attitudes ¬MA is more complex. They are also linked to the age,
and hence expertise, of the teacher, but strongly reflect his personality (low self-
confidence, fear of losing control over the class, tendency to keep to a reassuring
professional stability, tiredness and so on) and his personal history (education, lim-
ited attention to refresher courses, skepticism towards theories viewed—sometimes
correctly—as too abstract etc.). In these cases, the formative intervention steered to-
wards early algebra becomes powerful because it does not aim to add some knowl-
edge but rather attempts to induce a reflection which might prepare the ground for
a restructuring of knowledge. Most of all, it might promote the construction of a
mental attitude open to new perspectives concerning both theory and practice.
The four profiles are not pictures of teachers, in the sense that they are not defi-
nitions to be taken as absolute. They rather illustrate temporary stages at which the
teachers who adopt the multi-commented transcripts methodology come to be and
which can be modified over time. The improvement in the capacity of expressing
metacognitive attitudes can thus be seen as the outcome of a formative process to
which researchers contributed—we hope significantly—but it will inevitably be up
to the teacher to make it grow autonomously during their professional activity, if
they decide to.
The following example is taken from a transcript of a teacher in the phase
¬MA MC . This example highlights the negative incidence of the teacher’s language
on the discussion: this language is not appropriate and pays little attention to the
aims of the work that is being done. The comments in this case are made by the
teacher himself and by three researchers, but after some time, the teacher pre-
sented excerpts from his transcripts at a conference, enriching them with personal
remarks—often in the form of meta-comments—which highlight how the method-
ology had positively influenced his attitude, leading it towards a model MA MC .
The example was chosen because it shows the value of the methodology we
adopted, to promote teachers’ awareness of the limits of their own action. We realize
that this example alone cannot fully transmit the richness and variety of the emergent
issues, as well as what the teachers come to achieve; for this reason we send the
reader back to Malara (2008).
502 A. Cusi et al.
Example
The example refers to a moment of a teaching sequence about the study of sequences
that can be modeled algebraically. The main objective of the sequence is to make
pupils achieve a functional view of sequences and lead them to construct algebraic
representations for the latter, modeling the relation between the ranking (or place)
number and the correspondent term of the sequence. The teacher had proposed the
exploration of the sequence after giving the first three terms: 4; 11; 18; . . . . The
class had grasped that the sequence was generated, starting from 4, by the opera-
tor “+7”. Then the teacher had raised the question of searching for a formula that
could represent the correspondence place-term. In the classroom, attention was
then focused on the problem of generalization and the whole class had worked on
the meaning of the term ‘n-th’. To make the pupils’ exploratory activity easier the
teacher had summarized in a table all that they knew at that moment. In the study
of the case at place 30th there was a mistake: the number before 30 is swapped with
the number after it, during the generalization phase.
Place Number Operations Rule 1 Rule 2
1◦ 4
2◦ 11 4+7 4 + (7 × 1) 7 × 3 − 10
3◦ 18 4+7+7 4 + (7 × 2) 7 × 4 − 10
4◦ 25 4+7+7+7 4 + (7 × 3) 7 × 5 − 10
30◦ 4 + (7 × 31)
n
Teacher: I want to know: if I am at place n, that we said—do you remember?—it

was a place at a certain point, without knowing what point it was. Eh,
I want to know what is the rule that allows me to find this number at
place n (1) [I point to the n-th term on the blackboard] are you with me?
Teacher: Good, so let’s find the rule (2). Benedetta?
Benedetta: Eh, because I believed that n-th was the last, so I wrote “ there isn’t
because the sequence is infinite (3)”.
Teacher: All right, this is a true remark and perhaps it will be useful later, we will
keep it. So, how can we find the formula we need? Don’t look at me, look
at your sheet and the blackboard! How can you find it? (4) Andrea?
Andrea: Now, if we know. . . last time we said that n-th stood for any place (5)
Teacher: Question: number n means a number at any place (6) without saying
what number it is, this is the hard part! What formula do I write for the
number at the n-th place? (7)
Sergio: Er. . . I think you can’t find it because n-th is a number you don’t know
Andrea: As you said, n-th stands for a number at any place, therefore I say with
Sergio, if the place is indefinite, we will never get to know what number
it is! (8)
Teacher: Exactly, I agree with you too! If I don’t tell you, at the 3rd , at the 4th , at
the 100th , at the 7003rd place, you don’t know. But if I tell you that this
number . . . about this number, instead of telling you the place number I
tell you it is at place n, can I calculate . . . can I write a formula to write
this number? (9)
Comments
(1) T8 I now realize I have used the wrong terms, thus inducing students to give
the answers that will follow, and that I desperately fought against. By saying
“the rule to find this number at place n” the students understood that I wanted
to know that value of an . Perhaps I should have said “the rule to find a number
of the sequence, given its position”.
(2) R1 I suggest that the class be led to discover and highlight with arrows rela-
tions, repetitions of numbers, ‘local’ regularities. Many of these might be not
productive, but they help pupils get used to global explorations. For example,
the same sequence, proposed in another class, led some pupils to identify a re-
lation between the numbers of the first two columns and to represent it with
11 = 2 × 7 − 3, 18 = 3 × 7 − 3, 25 = 4 × 7 − 3, and so on. The arrows might
link the various four’s with the first term of the sequence, numbers 1, 2, 3 of the
fourth column with the place numbers of the first one, shifted one line down,
etc. These two last arrows might show that 31 is wrong and that it should be
substituted for 29.
(3) R2 Benedetta contradicts herself, if the sequence is infinite, also the places are
infinite and n cannot be the last one. Perhaps she means n as ‘very large’ num-
ber. With this contradiction she expresses her belief that a number at a non-
defined place cannot be represented. Anyway, she does not know the meaning
of n as indicator of a number we do not want to state explicitly.
(4) M T is worried by the idea that she should get to the formula written in algebraic
language. I keep believing that, in this phase of the work, the objective is to
lead pupils to grasp the relation between place and correspondent number, and
to express that relation clearly.
R1 Why not encourage expression in natural language, describing the forms of
columns 3 and 4: “I get the number by adding to the initial number as many
7 as . . .” or in any other way. The paraphrases proposed by pupils can then
be compared and the most suitable to be translated in algebraic language for
Brioshi can be chosen.
(5) M Well done Andrea, that “any place” is like gold!
onwards, we will use the following codes: T = teacher; M = mentor ; R1 = a researcher;

8 Here
R2 = another researcher.
504 A. Cusi et al.
(6) R2 More than ‘any’, term which is linked to the idea of variable, it would have
been appropriate to underline that it is a number we do not want to specify,
‘indeterminate’ (term which, focusing on the element, fixes it somehow).
(7) M I wondered many times: why not putting, in the column with the mathemat-
ical sentence, the (either mental or not) operation made to identify the factor
that multiplies 7, starting from the given number? Pupils would have grasped
the regularity, the reiteration of a procedure, getting closer to the construction
of the formula smoothly.
R2 The formula is gradually determined by identifying invariant parts (4 + 7 ×
. . .) and variant parts (number of place—1) in the studied cases.
(8) R1 the approach to the letter is rather complex, requires lengthy times, differ-
ent strategies, comparisons, explorations, entails continuous and unpredictable
evaporations. The joint presence of intuitions of different meanings in the in-
terventions of Sergio and Andrea is absolutely inevitable, almost physiological.
Probably the need (real or presumed) to conclude and get to the rule, imposes
to the teacher rhythms that can hardly coexist with that complexity. We are
fully immersed in algebraic babbling, and the learning of a new language, of its
meanings and rules, must respect the needs of a required settling.
(9) T Now I see why they could not answer! We don’t understand each other! As I
said at the beginning, the verb “to find” puts them on the wrong track! Perhaps
I should have said “find a representation of the number at place n which makes
us understand that this number is in the sequence”. Too complicated! I don’t
know. . .
M I agree on the damage caused by the term “to find”.
R1 I also agree on representing, even more if this term (Glossary) becomes one
of the keywords of the class’ cultural background, and hence acquires a shared
and negotiated meaning (Glossary again).
R2 Finally, well done T! Representing, yes, representing is the keyword.
The analysis of the comments clearly shows the epistemology of the researcher who
produced them, due to the prevalence of some types of comments. Both agreements
and disagreements in these comments turn out to be fruitful for the teacher, the
former by reinforcing the comment, the latter as enriching complements.
Concluding Remarks
An analysis of the discussion carried out so far is needed here. We said that, for
most students, the big obstacle in the study of algebra is represented by the diffi-
culty in having control over the meanings of formal expressions. They are led to
the manipulation of the latter through the application of rules that are semantically
opaque. This is the main reason why in the curricula of the K-8 stages, increasing
space is given to early algebra, in association with a socio-constructive teaching
practice. The aim is to propose a kind of teaching that may revisit arithmetic in
a pre-algebraic perspective introducing in primary school activities that foster the
development of pre-concepts useful for the learning of algebra.
Our investigations (in the ArAl project) are carried out within this framework.
The underlying conviction is that the main cognitive obstacles in the learning of al-
gebra arise in arithmetical contexts and might condition the development of math-
ematical thinking in students, due to a weak conceptual control over the meanings
of Algebraic objects and processes. Some of its most important principles are: an-
ticipation of pre-algebraic activities of a generational type, social construction of
knowledge, central role of natural language as didactical mediator, identification
and explicit expression of algebraic thinking, often ‘hidden’ in arithmetical con-
cepts and representations.
As a consequence, the issue of teacher training comes to be crucial for our aims.
In this respect, we underlined the value of critical reflections upon classroom-based
processes, also through participation in ‘communities of enquiry’ made of teachers
and researchers, like those involved in the ArAl project.9 Starting up a continuous
reflection upon oneself as a professional in education, implies that one understands
the directions he should go to support transformation through an inter-exchange
between theory and practice.
We drew attention to how, from our point of view, early algebra defines its area of
interest starting from both disciplines (either arithmetic or algebra) and acquires a
different, and mainly original, identity. We have defined early algebra as a meta dis-
cipline, dealing not only with entities, processes and properties of the two subjects,
but rather with the genesis of a unifying language and, therefore, of a meta-language.
A Glossary supports the construction of this meta-disciplinary knowledge. We
mentioned some key constructs: algebraic babbling, the pair solving and represent-
ing, canonical and non canonical form of the representation of a natural number, the
equality sign, the respect of the rules in the approach to the algebraic code, syntax
and semantics, and Brioshi.
Finally, we illustrated the Methodology of Multi-commented Transcripts
(MCTM) and its central role in the teacher’s formative process, to empower his
capacity of reflection, as well as to construct a constant attitude of noticing his own
behavior in the classroom-based action and to have control over the impact that his
way of acting may have on pupils’ attitudes and conquests.
As we said earlier, our main goal, of a ‘meta’ type, is to form meta-cognitive
students. But to do this, it is necessary that teachers learn to be meta-cognitive
teachers in turn. We examined instruments and methods we outlined to promote
meta-cognition in teachers, in a strict intertwining of reflections upon knowledge at
stake (theory) and action in the classroom (practice). We also showed the value of
an educational process which, in the long term, is able to give teachers a new pro-
fessional identity, more consistent with the role they need to play, and not only with
reference to early algebra. This is a condition for inducing in pupils, since the early
school years, a view of algebra as a language and as an instrument for thinking, a
constructive and reflexive attitude and, more in general, a conception of mathemat-
ics as carrier of meanings.
9 Since 2000 nearly 1000 teachers and more than 10.000 pupils from 12 Italian regions participated
in the project.
506 A. Cusi et al.
Finally we wish to add some remarks about an evaluation of the results attained
with the teachers who worked with MCTM.
We cannot, and do not want to, give quantitative data about pupils’ learning: we
only want to stress the kind of attitude spotted in very young pupils involved in our
teaching experiments. Some simple examples.
• Anna (final year at kindergarten) recognizes that two trains that continue who-
knows-where beyond the door of the room—one made of wagons with two yellow
and one red blocks and the other made of wagons with two nuts and a seed of
sunflower—“are almost equal”. Anna plays with structural analogy.
• Federica (grade 2, primary school) finds on her book the expression ‘3 × =
27’ and writes down ‘3 × 9 = 27’. Federica solves a linear equation with one
unknown.
• Piero (grade 3, primary school) notices that “It is correct to say that 5 plus 6 is
11, but you cannot say that 11 ‘is’ 5 plus 6, and then it is better saying that 5 plus
6 ‘equals’ 11, because in this case the contrary is true as well”. Piero is arguing
on the relational meaning of equal.
A linguistic and constructive approach to algebra, like the one we propose, seems
to be productive in terms of leading pupils to view activities such as translating, ar-
guing, interpreting, predicting, communicating as mathematical activities. Making
calculations is still there, but subordinated to ‘higher’ purposes, it helps prepare
reasoning, argumentations, refutations, corrections. As the complexity of the alge-
bra they will deal with increases, pupils will be led to understand that manipulation
of symbols (polynomials, equations, functions etc.) is not self-referential, but helps
them mathematize, explore, reason, deduce, in other words, produce thinking and
achieve new knowledge.
Due to the objective situation of the school environment in which our activities
are set, only a qualitative evaluation of the teachers’ growth is possible, by observing
their behaviors, through their own self-observation and by means of the instruments
we illustrated.
Inducing long-term processes of change means promoting individual develop-
ment throughout time, beyond the period of collaboration with the project.
Teachers need the same structural steps highlighted in the Glossary for their own
construction of knowledge. Reflection on one’s knowledge, habits, behaviors and
stereotypes, occurs through a process of negotiation with the often conflicting per-
spectives offered by early algebra. Achievements in this area are to be mediated,
re-elaborated, metabolized—negotiated as such—in a continuous sharing of new
cultural values. Devolution should be constantly at stake, even in the relationship
between researchers and teachers: the latter should accept the responsibility of the
learning situation they are engaged in. Institutionalization of both mathematical and
methodological knowledge will occur in time, as the teacher will become more and
more aware of the sense that the perspective of early algebra can assume with re-
spect to the changes of the questioned didactical processes.
A meaningful conclusion seems to be offered by the content of an e-mail written
by a teacher after a four-year-long collaboration:
In my twenty-years-long experience I’ve always asked my many teacher trainers: “Well,

after this nice premise, tell me HOW I should act with my students! What am I supposed
to do for them not to damage them?” Many of them answered that they were dealing with
the contents of the discipline and not with its teaching, that it was up to me to elaborate the
right strategies after all I had heard from them etc. etc. . . . AND SO!
Well, it was different with you. What you gave me made a better teacher of me. To use one
of your terms, I might say that four years ago I started my ‘epistemological babbling’.
References
Ainley, J. (coord.) (2001). Research forum on early algebra. In M. van den Heuvel-Panhuizen (Ed.),
Proceedings of the 25th Annual Conference of the International Group for the Psychology of
Mathematics Education, Utrecht (The Netherlands) (Vol. 1, pp. 128–159).
Anghileri, J. (2006). Scaffolding practices that enhance mathematics learning. Journal of Mathe-
matics Teacher Education, 9, 33–52.
Arcavi, A. (1994). Symbol sense: informal sense-making in formal mathematics. For the Learning
Arzarello, F. (1991). Procedural and relational aspects of algebraic thinking. In F. Furinghetti (Ed.),
Proceedings of the 15th Annual Conference of the International Group for the Psychology of
Mathematics Education, Assisi (Italy) (Vol. 1, pp. 80–87).
Arzarello, F., Bazzini, L., & Chiappini, G. (1993). Cognitive processes in algebraic thinking: To-
wards a theoretical framework. In I. Hirabayashi, N. Nohda, K. Shigematsu, & F. L. Lin (Eds.),
Proceedings of 17th International Conference on the Psychology of Mathematics Education,
Tokio (Japan) (Vol. 1, pp. 138–145).
Bell, A. (1976). A study of pupils’ proof explanations. Educational Studies in Mathematics, 7,
23–40.
Bell, A., Swan, M., Onslow, B., Pratt, K., & Purdy, D. (1985). Diagnostic teaching for long term
learning (Report of ESRC Project HR8491/1). Shell Centre for Mathematical Education, Uni-
versity of Nottingham.
Blanton, M., & Kaput, J. (2001). Algebrafying the elementary mathematics experience: Transform-
ing task structures: Transforming practice on a district-wide scale. In E. Chick et al. (Eds.),
Proceedings of the 12th ICMI Study ‘The Future of the Teaching and Learning of Algebra’,
Melbourne (Australia) (Vol. 1, pp. 87–95).
Blanton, M., & Kaput, J. (2002). Design principles for tasks that support algebraic thinking in
elementary school classroom. In A. D. Cockburn & E. Nardi (Eds.), Proceedings of 26th Inter-
national Conference on the Psychology of Mathematics Education, Norwich (UK) (Vol. 2, pp.
105–112).
Blanton, M. L., Stylianou, D. A., & David, M. M. (2003). The nature of scaffolding in undergrad-
uate students’ transition to mathematical proof. In N. Pateman, B. J. Dougherty, & J. Zilliox
(Eds.), Proceedings of 27th International Conference on the Psychology of Mathematics Edu-
cation, Honolulu (Hawaii) (Vol. 2, pp. 113–120).
Borasi, R., Fonzi, J., Smith, C., & Rose, B. (1999). Beginning the process of rethinking math-
ematics instruction: A professional development program. Journal of Mathematics Teacher
education, 2(1), 49–78.
Booth, L. R. (1984). Algebra: Children’s Strategies and Errors. A Report of the Strategies and
Errors in Secondary Mathematics Project. Windsor, Berks: NFER-Nelson.
Cai, J., Lew, H. C., Morris, A., Moyer, J. C., Ng, S. F., & Schmittau, J. (2005). The development of
students’ algebraic thinking in earlier grades: A cross-cultural comparative perspective. ZDM,
37(1), 5–15.
Carpenter, T., & Franke, M. L. (2001). Developing algebraic reasoning in the elementary school:
Generalization and proof. In E. Chick, K. Stacey, Jl. Vincent, & Jn. Vincent (Eds.), Proceedings
of the 12th ICMI Study ‘The Future of the Teaching and Learning of Algebra’, Melbourne
(Australia) (Vol. 1, pp. 155–162).
508 A. Cusi et al.
Carpenter, T. P., Franke, M. L., & Levi, L. (2003). Thinking Mathematically. Integrating Arithmetic
and Algebra in the Elementary School. Portsmouth, NH: Heinemann.
Carraher, D., Brizuela, B., & Schliemann, A. (2000). Bringing out the algebraic character of arith-
metic: Instantiating variables in addition and subtraction. In T. Nakahara & M. Koyama (Eds.),
Proceedings of the 24th Conference of the International Group for the Psychology of Mathe-
matics Education, Hiroshima (Japan) (Vol. 2, pp. 145–152).
Chick, E., Stacey, K., Vincent, Jl., & Vincent, Jn. (Eds.) (2001). Proceedings of the 12th ICMI
Study ‘The Future of the Teaching and Learning of Algebra’. Melbourne, Australia: University
of Melbourne.
Cusi, A. (2008). An approach to proof in elementary number theory focused on representation and
interpretation aspects: The teacher’s role. In B. Czarnocha (Ed.), Handbook of Mathematics
Teaching Research (pp. 107–122). Rzeszów: Rzeszów University Press.
Cusi, A., & Malara, N. A. (2008). Approaching early algebra: Teachers’ educational processes and
classroom experiences. Quadrante, 16(1), 57–80.
Da Rocha Falcão, J. T. (1995). A case study of algebraic scaffolding: From balance to algebraic
notation. In L. Meira & D. Carraher (Eds.), Proceedings of the 19th Annual Conference of the
International Group for the Psychology of Mathematics Education, Recife (Brazil) (Vol. 2, pp.
66–73).
Dougherty, B. (2001). Access to algebra: A process approach. In E. Chick, K. Stacey, Jl. Vincent,
& Jn. Vincent (Eds.), Proceedings of the 12th ICMI Study ‘The Future of the Teaching and
Learning of Algebra’, Melbourne (Australia) (Vol. 1, pp. 207–212).
DFE (Department for Education) (1995). Mathematics in the National Curriculum. London:
HMSO.
Ferreri, S. (2006). Parole tra quantità e qualità. In I. Tempesta & M. Maggio (Eds.), Linguaggio,
mente, parole, Dall’infanzia all’adolescenza, Collana GISCEL (pp. 19–25). Milano: Franco
Angeli Editore.
Filloy, E. (1990). PME algebra research. A working perspective. In G. Booker, P. Cobb, & T. N.
Mendicuti (Eds.), Proceedings of the 14th Annual Conference of the International Group for
the Psychology of Mathematics Education, Oaxtepex (Mexico) (Vol. 1, pp. 1–33).
Filloy, E. (1991). Cognitive tendencies and abstraction processes in algebra learning. In F. Fur-
inghetti (Ed.), Proceedings of the 15th Annual Conference of the International Group for the
Psychology of Mathematics Education, Assisi (Italy) (Vol. 2, pp. 48–55).
Gray, E., & Tall, D. (1993). Success and failure in mathematics: The flexible meaning of symbols
as process and concept. Mathematics Teaching, 142, 6–10.
Harper, E. (Ed.) (1987–88). NMP Mathematics for Secondary School. Essex, UK: Longman.
Kaput, J. (1991). Notations and representations as mediators of constructive processes. In E. von
Glasersfeld (Ed.), Radical Constructivism in Mathematics Education (pp. 53–74). The Nether-
lands: Kluwer Academic Publishers.
Kaput, J. (1995). A research base supporting long term algebra reform. In D. T. Owens, M. K. Reed,
& G. M. Millsaps (Eds.), Proceedings of the 17th Annual Meeting of PME-NA (Vol. 1, pp.
71–94). Columbus (OH): ERIC Clearinghouse for Science, Mathematics and Environmental
Education.
Kaput, J., & Blanton, M. (2001). Algebrafying the elementary mathematics experience: Transform-
ing task structures. In E. Chick, K. Stacey, Jl. Vincent, & Jn. Vincent (Eds.), Proceedings of the
12th ICMI Study ‘The Future of the Teaching and Learning of Algebra’, Melbourne (Australia)
(Vol. 1, pp. 344–353).
Kaput, J., Carraher, D. W., & Blanton, M. L. (Eds.) (2007). Algebra in Early Grades. Mahwah
(NJ): Lawrence Erlbaum Associates.
Kieran, C. (1989). The early learning of algebra: A structural perspective. In S. Wagner & K.
Kieran (Eds.), Research Issues in the Learning and Teaching of Algebra (pp. 33–56). Reston
(Virginia): LEA.
Kieran, C. (1998). The changing face of school algebra. In C. Alsina, J. Alvarez, B. Hodgson, C.
Laborde, & A. Perez (Eds.), 8th International Congress on Mathematics Education: Selected
Lectures (pp. 271–290). Seville, Spain: S.A.E.M. Thales.
(Singapore), 8(1), 139–151.
Kuchemann, D. E. (1981). Algebra. In K. Hart (Ed.), Children Understanding Mathematics: 11–16
(pp. 102–119). London: Murray.
Jaworski, B. (1998). Mathematics teacher research: Process, practice and the development of teach-
ing. Journal of Mathematics Teacher Education, 1, 3–31.
Jaworski, B. (2003). Research practice into/influencing mathematics teaching and learning de-
velopment: Towards a theoretical framework based on co-learning partnerships. Educational
Studies in Mathematics, 54, 249–282.
Jaworski, B. (2004). Grappling with complexity: Co-learning in inquiry communities in mathemat-
ics teaching development. In M. J. Hoines & A. B. Fuglestad (Eds.), Proceedings of the 28th
Annual Conference of the International Group for the Psychology of Mathematics Education,
Bergen (Norway) (vol. 1, pp. 17–36).
Lee, L., & Wheeler, D. (1989). The arithmetic connection. Educational Studies in Mathematics,
20(1), 41–54.
Linchevski, L. (1995). Algebra with numbers and arithmetic with letters: A definition of pre-
algebra. Journal of Mathematical Behaviour, 14, 113–120.
Lins, R. C. (1990). A framework for understanding what algebraic thinking is. In G. Booker, P.
Cobb, & T. N. Mendicuti (Eds.), Proceedings of the 14th Annual Conference of the Interna-
tional Group for the Psychology of Mathematics Education, Oaxtepex (Mexico) (Vol. 2, pp.
93–100).
Malara, N. A. (2003). Dialectics between theory and practice: Theoretical issues and aspects of
practice from an early algebra project. In N. A. Pateman, B. J. Dougherty, & J. Zilliox (Eds.),
Proceedings of the Joint Meeting of PME and PME-NA, Honolulu (Hawaii) (Vol. 1, pp. 33–48).
Malara, N. A. (2005). Leading in-service teachers to approach early algebra. In L. Santos et al.
(Eds.), Mathematics Education: Paths and Crossroads (pp. 285–304). Lisbon: Etigrafe.
Malara, N. A. (2008). Methods and tools to promote a socio-constructive approach to mathematics
teaching in teachers. In B. Czarnocha (Ed.), Handbook of Mathematics Teaching Research (pp.
89–102). Rzeszów: Rzeszów University Press.
Malara, N. A., & Iaderosa, R. (1999). Theory and practice: A case of fruitful relationship for the
renewal of the teaching and learning of algebra. In F. Jaquet (Ed.), Proceedings of CIEAEM 50
(pp. 38–54). Neuchatel: Neuchatel University Press.
Malara, N. A., & Navarra, G. (2001). “Brioshi” and other mediation tools employed in a teaching
of arithmetic with the aim of approaching algebra as a language. In E. Chick, K. Stacey, Jl.
Vincent, & Jn. Vincent (Eds.), Proceedings of the 12th ICMI Study ‘The Future of the Teaching
and Learning of Algebra’, Melbourne (Australia) (pp. 412–419).
Malara, N. A., & Navarra, G. (2003). ArAl Project: Arithmetic Pathways Towards Favouring Pre-
Algebraic Thinking. Bologna: Pitagora.
Malara, N. A., & Navarra, G. (2009). The analysis of classroom-based processes as a key task in
teacher training for the approach to early algebra. In B. Clarke, B. Grevholm, & R. Millman
(Eds.), Tasks in Primary Mathematics Teacher Education (pp. 235–262). Berlin: Springer.
Malara, N. A., & Zan, R. (2002). The Problematic relationship between theory and practice. In L.
English (Ed.), Handbook of International Research in Mathematics Education (pp. 553–580).
Mahwah (NJ): LEA.
Malara, N. A., & Zan, R. (2008). The complex interplay between theory and practice: Reflec-
tions and examples. In L. English (Ed.), Handbook of International Research in Mathematics
Education (2nd ed.) (pp. 539–564). New York: Routledge.
Mason, J. (1998). Enabling teachers to be real teachers: Necessary levels of awareness and structure
of attention. Journal of Mathematics Teacher Education, 1, 243–267.
Mason, J. (2002). Researching Your Own Practice: The Discipline of Noticing. London: The
Falmer Press.
510 A. Cusi et al.
Meira, L. (1990). Developing knowledge of functions through manipulation of a physical device.

In G. Booker, P. Cobb, & T. N. Mendicuti (Eds.), Proceedings of the 14th Annual Conference
of the International Group for the Psychology of Mathematics Education, Oaxtepex (Mexico)
(Vol. 2, pp. 101–108).
Menzel, B. (2001). Language conceptions of algebra are idiosyncratic. In E. Chick, K. Stacey, Jl.
Vincent, & Jn. Vincent (Eds.), Proceedings of the 12th ICMI Study ‘The Future of the Teaching
and Learning of Algebra’, Melbourne (Australia) (Vol. 2, pp. 446–453).
NCTM (National Council of Teacher of Mathematics) (1998). The Nature and Role of Algebra in
the K-14 Curriculum. Washington, DC: National Academy Press.
NCTM (National Council of Teacher of Mathematics) (2000). Principles and Standards for School
Mathematics. Reston (VA): NTCM. http://standars.NTCM.org.
Ponte, J. P. (2004). Investigar a nossa própria prática: una stratégia de formaçáo e de construçáo de
conhecimento professional. In E. Castro & E. De la Torre (Eds.), Investigación en educación
matemática, University of Coruña (Spain) (pp. 61–84).
Potari, D., & Jaworski, B. (2002). Tackling complexity in mathematics teaching development:
Using the teaching triad as a tool for reflection analysis. Journal of Mathematics Teacher Edu-
cation, 5(4), 351–380.
Schoenfeld, A. (1998). Toward a theory of teaching in context. Issues in Education, 4(1), 1–94.
Schwarz, B., Dreyfus, T., Hadas, N., & Hershkowitz, R. (2004). Teacher guidance of knowledge
construction. In M. J. Høines & A. B. Fuglestad (Eds.), Proceedings of the 28th Annual Confer-
ence of the International Group for Psychology of Mathematics Education, Bergen (Norway)
(Vol. 4, pp. 169–176).
Sherin, M. G. (2002). A balancing act: Developing a discourse community in a mathematics class-
room. Journal of Mathematics Teacher Education, 5, 205–233.
Wood, T. (1999). Approaching teacher development: Practice into theory. In B. Jaworski, T. Wood,
& S. Dawson (Eds.), Mathematics Teacher Education: Critical International Perspectives (pp.
163–179). London: The Falmer Press.
Yackel, E., & Cobb, P. (1996). Sociomathematical norms, argumentation, and autonomy in math-
ematics. Journal for Research in Mathematics Education, 27(4), 458–477.
A Procedural Focus and a Relationship Focus
to Algebra: How U.S. Teachers and Japanese
Teachers Treat Systems of Equations
Margaret Smith
Abstract This chapter explores two contrasting ways of presenting algebra by

looking at key differences across the presentation of simultaneous equations to stu-
dents in eighth-grade. The examples are from a qualitative analysis of the 1995
TIMSS Video Study data including eighth-grade mathematics instruction in Japan
and the United States covering topics on simultaneous equations. The U.S. lesson
example shows a procedural approach to this topic, where students focus on getting
answers through a series of routine steps. In contrast, the Japanese lesson highlights
a strong focus on building generalized solution methods and understanding relation-
ships represented in systems of equations. A discussion of key differences as they
relate to important ideas in understanding algebra compared to how it was treated in
the classrooms follows the examples.
The purpose of this chapter is to highlight the differences across two contrasting
ways of teaching algebra during classroom instruction. In particular, this chap-
ter uses data from the TIMSS 1995 Video Study to compare an example of a
Japanese teacher’s relationship focused method of teaching algebra to an exam-
ple of a United States teacher’s procedurally focused method of teaching alge-
bra. There are many goals in algebra, but one of the most common and fun-
damental goals is helping students move beyond an arithmetic approach to a
more generalized approach to understanding relationships (Carpenter et al. 2003;
This data used in this article were collected using video data from TIMSS Laboratory at the
University of California, Los Angeles. I wish to thank the TIMSS Video Laboratory for access to
a unique and resourceful data set. This research was supported by a grant of the American
Educational Research Association which receives funds for its “AERA Grants Program” from the
National Center for Education Statistics and the Office of Educational Research and Improvement
(U.S. Department of Education) and the National Science Foundation under NSF Grant #
RED-9452861. Opinions reflect those of the author and do not necessarily reflect those of the
granting agency.
M. Smith ()
Department of Mathematics, Iona College, New Rochelle, USA

512 M. Smith
Carraher and Schliemann 2007). Far too often students focus on the more procedural
aspects of calculating solutions rather than exploring the relationships represented
in algebraic expressions; the lessons presented here help highlight at some of the
key areas teachers can emphasize to help students better understand the generalities
and functional relationships represented with algebraic expressions.
Background
Algebraic Reasoning
Algebra is a very broad topic covering a variety of important components. Both

Kieran (2007) and Carraher and Schliemann (2007) describe a variety of different
complex cognitive components contributing to the understanding of algebra. Even
if we try to limit our understanding to the algebra important for students to learn, we
can see that there are many complex ideas that span a variety of concepts (Arcavi
2008; Chazan 2008). This chapter is limited to three of these main ideas: equality,
generalizing, and variables. These were chosen to help highlight some of the key ar-
eas important in the transition from arithmetic to algebra (Carraher and Schliemann
2007; Lins and Kaput 2004).
Equality Equality refers to students recognition that the equal sign represents
equivalence between sets, showing a specific relationship; for example 5 = 5 =
2 + 3 = 7 − 2. Equality carries with it some specific mathematical properties that
are important for understanding some more advanced concepts in algebra; for ex-
ample a = b is the same as saying b = a, if a = b and b = c then a = c, and others.
Therefore, the understanding of equality can be described in very simple terms but
understanding it carries with it some very complex ideas. This is probably one rea-
son why children’s understanding of equality has been researched as much as it has
(Kieran 1981, 1992, 2007; Knuth et al. 2006; Falkner et al. 1999).
One of the consistent findings of research is that students tend to view the equal
sign as a command, or prompt, rather than as a means of expressing a relationship
(Knuth et al. 2006; Falkner et al. 1999). For example, Falkner et al. (1999) describe
how all sixth-grade students, when given the expression 8 + 4 = + 5 thought that
12 or 17 would go in the box. These solutions indicate that these students were
looking at the command as “do this,” instead of a ways of expressing a relationship.
It is not surprising that students develop this notion of equality since many textbooks
often imply that nothing after the equal sign implies a command, such as 5 + 8 = .
So the question arises, how necessary is it to treat the equal sign as a relation-
ship rather than a command? The work by Falkner et al. (1999) and Knuth et al.
(2006) indicate that this transition is extremely important. The idea of equality as
a relationship can be built with young students, but they have difficulty extending
this relationship to mathematical symbolism (Falkner et al. 1999). Moreover, Knuth
(2006) and his colleagues found that the use of equality as a relational understand-
ing was important for both solving equations and using algebraic solutions to solve
Comparing Approaches to Algebra Instruction 513
tasks. These ideas indicate that helping students move beyond the idea of the equal
sign as an operator to one in which the equal sign represents a mathematical rela-
tionship is extremely important as students begin to study algebra.
Variables Variables are a way of representing quantities; in early algebra vari-

ables are usually included in equations and in late studies of algebra will include
inequalities. For example, 6 + y = x + 4 is the set of all pairs, x and y, that make
the equation true, such as x = 5 and y = 3 or x = 0 and y = −2. In the equa-
tion 5x = 10, x can also be considered a variable, even though only one value
of x, namely x = 2, makes the equation true because x is used to represent the
quantity that makes this true. However, many students who study algebra focus pri-
marily on the arithmetic calculations and the specific solutions to equations, leav-
ing the variable to the role of a place-holder or unknown (Lins and Kaput 2004;
Stacey and MacGregor 1999). This notion is problematic, because in order to un-
derstand that y and x have a specific relationship to each other in the equation
6 + y = x + 4, namely any pair of values where x is two more than y, one has to first
understand that any pair of quantities with that relationship will satisfy the equa-
tion; moreover understanding how these equations give rise to coordinate graphs
will likely be difficult to understand.
The notion of looking at a variable as characterizing a set of quantities, rather
than a single quantity is not beyond the scope of introductions to algebra. Carpen-
ter and Levi (1999) show how first and second grade students can begin to make
the transition from specific cases to a class of solutions. In their study, Carpenter
and Levi look at how some students reason about whether 78 − 49 + 49 = 78 is
true; some students were able to rationalize about how subtracting forty-nine and
then adding forty-nine is true for all numbers, indicating that they recognized that
78 − x + x would equal 78, even if x varied, showing an early understanding of vari-
able. Fujii and Stephens (2001, 2008) and Lins and Kaput (2004) consider this shift
from specific arithmetic solutions to understanding a class of solutions as quasi-
variable thinking; that is students’ understand that solutions to equations can be
represented by a class of solutions but they cannot yet explain how a generalized
expression can be used to represent this relationship. It is clear that to help students
move from the specificity of solutions to arithmetic calculations (5 + 4 = ) to un-
derstanding the general relationship expressed in equations we must help students
build on their informal understandings to ways of expressing more formal relation-
ships, with variables in expressions.
Generalizing Generalizing refers to connecting solutions to specific tasks to un-

derstanding how the relationships in these tasks represent ideas that have a larger
class or relationships. Blanton and Kaput (2003) refer to this as “algebrafy,” gen-
eralizing mathematical thinking and justifying generalizations. Recognizing both
equality and variables, as discussed above, are two examples of learning to gener-
alize; but students also learn to generalize when they learn to examine solutions to
problems (Stacey and MacGregor 1999). Kaput (2007) identifies this as a key com-
ponent of introducing algebra, particularly to young children; in particular he argues
514 M. Smith
that this generalizing process enables students to move beyond a specific to a more
mathematical way of reasoning about situations.
The study by Carpenter and Levi (1999) show how students can begin to rea-
son about a more general idea from a specific problem. Swafford and Langrall
(2000) show how sixth-grade students are able to reason from specific cases to
more generalized solutions; however the students had a difficult time using a gen-
eralized solution to find a specific solution. In particular, Swafford and Langrall
gave students a series of problems, the first was to find a particular solution to
a real-life context with one specific answer, next students were asked to gener-
ate a way to represent a set of problems with the same context, and finally stu-
dents were asked to use these general forms to solve specific problems. Students
in the study were successful in solving the specific cases and many could build an
equation to represent a general solution, but they had difficulty using this general-
ized form to solve a specific case. Research by others (Blanton and Kaput 2003;
Levin 2008; Radford 1996) also shows how children can reason through problem
situations to find both specific problems and begin to discuss the generalized rela-
tionships represented in the problems. These studies indicate that students bring im-
portant understandings about relationships in real-world contexts that can be built
upon to help them understand generalizing mathematical relationships. However,
Swafford and Langrall (2000) study also shows that students may have a difficult
time understanding what it means to have a generalization; indicating a need for
classroom instruction to help students connect between the general and the specific.
TIMSS Video Studies
There have been two TIMSS Video Studies completed, to date, the 1995 and 1999
TIMSS Video Studies (Stigler et al. 1999; Gonzales et al. 2005). Each of the stud-
ies collected videotaped samples of eighth-grade mathematics instruction; the 1995
TIMSS Video Study collected data from Germany, Japan, and the United States,
and the 1999 TIMSS Video Study collected data from Australia, the Czech Repub-
lic, Hong King SAR, the Netherlands, Switzerland, and the United States (again).
This data set was significant in that it provided a large-scale representation of class-
room instruction in several different countries and it provided a data set that could
be analyzed by others for a variety of different purposes.
Results from both studies have shown that “Teaching is a cultural activity”
(Stigler and Hiebert 1998, p. 11), built upon traditions and practices of the soci-
ety. In addition, both the 1999 and 1995 TIMSS Video Studies show similarities
within a country, as well as across countries, and also shows how mathematics in-
struction varies across countries in some very important ways (Stigler et al. 1999;
Gonzales et al. 2005). For example, in the 1999 TIMSS Video Study (Gonzales et al.
2005) countries varied on the level of procedural complexity of problems presented
during a lesson, with Japan presenting significantly more high procedural complex-
ity problems than the other countries. As another example, countries showed sig-
nificant differences on the amount of problems presented that were connected to

real-world applications.
The results from these studies provide significant contributions to understand-
ing mathematics instruction from a variety of different approaches. In addition, the
data set provided opportunities for secondary analysis of the data so that others
could research mathematics instruction for a variety of purposes (Jacobs et al. 2006;
Kieran 2004; Smith 2000). This secondary analysis has allowed researchers to ex-
amine other aspects of mathematics instruction that were not included in the original
analysis. For example, Kieran (2004) describes how students’ ideas of equality can
be used to help make connections to understand algebraic relationships of equality.
These secondary studies provide researchers opportunities to provide more in-depth
analysis of the classroom instruction in the data set.
Data
The data used for this study were from the 1995 TIMSS Video Study (Stigler et
al. 1999). This was the first large-scale video study of classroom instruction which
allowed for secondary analysis studying classroom teaching. This data set, rather
than the data from the 1999 TIMSS Video Study (Gonzales et al. 2005), was used
here because the 1995 study showed that Japanese classroom teaching in the data
set reflected the ideals of the NCTM Standards (1989, 2000) while instruction in
the United States reflected a more traditional approach (Stigler and Hiebert 1999;
Jacobs et al. 2006); however Japan did not collect new data for the 1999 TIMSS
Video Study. Because the videotapes of classroom instruction in Japan reflects the
instruction different from that of the United States, it can provide ways to better
understand how classroom instruction could meet the ideals presented in U.S. re-
form documents (NCTM, 1989, 2000). It should be noted that achievement by U.S.
eighth-grade students has improved since the 1995 TIMSS Study, both overall and
in algebra, indicating a possible change in some instruction methods in the United
States (Gonzales et al. 2005, 2009). However, it is still possible that some classroom
instruction may look very similar to the U.S. data collected in 1995 and therefore
it is worthwhile to consider the contrast of the two ways of approaching algebra
instruction to highlight the need to re-examine algebra instruction.
Analysis
The data presented here were analyzed using qualitative methods; examining pat-
terns and trends that reflect the classroom instruction in each of the two countries.
This analysis was completed after quantitative methods were used to identify key
differences in classroom instruction in each of the countries (Gonzales et al. 2005;
Smith 2000; Stigler et al. 1999). The examples presented here, therefore, contain
516 M. Smith
the patterns and trends as illustrated in several different classroom instruction seg-
ments, but did not occur verbatim as presented here. The usefulness of this method
is to bring to light some of the key qualitative differences in classroom instruction
(Jacobs et al. 2007; Jacobs and Morita 2002).
Two Teachers’ Lessons
The following examples illustrate two different approaches to teaching algebra, par-
ticularly solving simultaneous equations. Because of privacy issues (Arafeh and
McLaughlin 2002; Jacobs et al. 2007), neither lesson happened as presented here;
however each represents a set of lessons within each respective country which cov-
ered similar topics. Moreover, the dialogue presented mimics patterns observed in
the respective countries, for example the types of ideas provided by students in the
Japanese lesson represent the types of ideas they provided in observed lessons; simi-
larly the nature of teacher presented ideas in the United States are also representative
of the ways material was presented in the videotapes of lessons in the U.S. (Stigler
and Hiebert 1998).
Mr Kirkyle’s Lesson In the first example (Fig. 1), Mr. Kirkyle focuses on the pro-
cedures to help students solve simultaneous equations and connect these to linear
graphs on the coordinate plane. Although Mr. Kirkyle attempts to connect equation
solutions with the relationships between the corresponding lines, it is not clear that
students appreciate this connection; rather they likely take away the procedures and
algorithms demonstrated in the lesson (Thompson and Thompson 1994). A discus-
sion of these ideas as seen in the examples is discussed below in an examination of
key differences.
Mr. Nakamura’s Lesson In the second lesson (Fig. 2), Mr. Nakamura focuses
on relationships and generalizing solutions of simultaneous equations. In contrast
to Mr. Kirkyle’s lesson, Mr. Nakamura helps focus students’ attentions on the rela-
tionships the system of equations represents. This lesson shows how problems that
appear procedural can still be completed with conversations that provide rich math-
ematical connections, allowing students to begin to connect to the relations and gen-
eralizations which characterize algebra (Smith 2000). A discussion of these ideas as
seen in the examples is discussed below in an examination of key differences.
Discussion of Key Differences
There are many differences across these two lessons, but I would like to focus specif-
ically on how these two lessons introduce some algebraic concepts to the students
in their classes.
Setting and Discourse On Chalkboard
As students enter into the room they are asked to copy the following Two lines have the following relationships
information into their notebook. After completing introductory Intersecting Parallel One line
routines, Mr. Kirkyle begins his lesson.
Mr K.: So as you copied down here, there are three ways lines can be drawn
together, they can intersect at a point, they can be parallel, or they can be
one line
Mr. K.: So let’s look at the following examples 1. −x + 4y = −21
3x − 12y = −21
Mr. K:. So What do we have here? I see the first equation has two 2. −4x + y = 2
unknowns, so we must have two equations. If I had three variables 12x − 3y = −6
how many equations do I need?
Comparing Approaches to Algebra Instruction
S: Three.
Mr. K.: Three, good. So let’s solve these. For the first one we need 1. −x + 4y = −21 (times 3)
to multiply both sides of the top by three. 3x − 12y = −21
Mr K.: So how can we solve these? Solve by combinations
S: With combinations. −3x + 12y = −63
Mr. K.: Good so we can combine these two. 3x − 12y = −21
Solve by combinations
Mr. K.: Let’s multiply the top by three, hat will give us. . . negative
−3x + 12y = −63
three, plus twelve, equals negative sixty-three. . .so we get zero and
3x − 12y = −21
negative eighty-four so there are no solutions.
0 = −84
Mr. K.: Okay, let’s look at the next one. We’ll multiply by three 2. −4x + y = 2
again. 12x − 3y = −6
517
Fig. 1 Mr. Kirkyle’s lesson

518
Mr. K: What do I get when I multiply it by three? −4x + y = 2 (times 3)

S: Negative twelve x plus three y equals six. 12x − 3y = −6
Mr. K.: Good now let’s add these two together and we get zero, −12x + 3y = 6
zero, and zero, so we get zero equals zero. What does this tell us 12x − 3y = −6
about the lines? 0=0
Mr. K.: They are the same line; that means there will be many solutions.
Let’s try it out let’s put in one and. . .what will work with that. . .six, let’s
put one and six into the first equation and see what we get
Mr. K.: Let’s try it out let’s put in one and. . .what will work with −4(1) + 6 = 2
that. . .six, let’s put one and six into the first equation and see what 12x − 3y = −6
we get
Mr. K.: Now let’s put it into the second one because they should work for 4(1) + 6 = 2
both. So we get twelve minus eighteen and that’s negative six, good so it 12(1) − 3(6) = −6
works. What ever you find that works for one will work for the other.
Mr. K.: So when I graph these lines number one is parallel because there −3x + 12y = −63 −12x + 3y = 6
are no solutions and the second is the same line because there are many 3x − 12y = −21 12x − 3y = −6
solutions. 0 = −84 0=0
No solutions Many solutions
Mr. K.: Okay one last one −2x + 3y = 6
2x + y = 10
S: You can just combine them. −2x + 3y = 6
Mr. K.: Okay so we get 2x + y = 10
4y
S: Four y equals 16 divided by four and we have y equals 4 = 16
4 4
y=4
Fig. 1 (Continued)
M. Smith
Mr. K: So that is y. Can someone else help us get x? −2x + 3(4) = 6
S: We just plug that into the equation. −2x + 12 = 6
Mr. K.: Which one? −12 −12
−2x −6
S: It doesn’t matter; I put it in the first one and got minus two x plus −2 = −2
three times four is six. . .so I subtracted twelve and got minus two x is x=3
negative six and divided by negative two is three.
Mr. K.: So the answer is (3, 4) and because it is a point you know the lines (3, 4)
intersect.
Mr. K.: Okay for the rest of the class I would like you to work on numbers
8 through 12 but you do not have to draw the graphs, just write down if the
lines are parallel, intersecting, or the same.
Fig. 1 (Continued)
519
520
Setting and Discourse On Chalkboard

As students walk in the simultaneous equations (1) 2x + y = 9 x 0 1 2 3 4 5 6 x 0 1 2 3 4 5 6
shown here are on the chalkboard. After greeting x +y =5 y y
the students, Mr. Nakamura asks students to (2) x − y = 2 x −5 −4 −3 −2 −1 0 1 x −5 −4 −3 −2 −1 0 1
complete the problems. x − 2y = 6 y y
Mr. N.: Here are equations with two unknowns, we
call these simultaneous equations; try to answer this
in two minutes. Ready. Go.
Mr. N.: Okay how many people were able to finish
both? Oh, no one. How many people finished just
number one. Oh, a few. What makes this so hard.
S: It takes a lot of tries.
Mr. N.: Then I wonder how we can solve it so we can
do them fast.
Mr. N.: What about these problems you learned in 9x − 5 = 2x + 16
seventh grade?
Mr. N.: Can you solve these fast?
Students: Yes.
Mr. N.: Okay thirty seconds, ready go.
Mr. N.: Okay tell me how you solved this. 9x − 5 = 2x + 16
S.: I used properties of equality to find out the x 9x − 2x = 16 + 5
that makes this true. 7x = 21
x=3
Fig. 2 Mr. Nakamura’s lesson

M. Smith
Mr. N.: Good. Before you knew how to follow the
steps how did you solve this? Did you put in num-
bers?
S: Yes we would try different numbers until we
made it equal, so like nine time three minus five is
the same as 2 times three plus sixteen.
Mr. N.: Very good. So how can we use that to do (1) 2x + y = 9
the first problem very quickly? Let’s show it with x+y =5
out diagram too. Do these say the same thing?
Students: Yes
Mr. Nakamura and his students both seem familiar
with using the shapes to represent the letter sym-

bols with the common language that the sticks are
counted as individual ones and all the other shapes
“next to” is the same as multiplication.
Mr. N.: Okay now work with your groups to see how
you can solve these.
After students have worked for a considerable Group 1
amount of time with one another, Mr. Nakamura 2x + y = 9
gathers the students to discuss their strategies. Transpose 2x x 1 2 3 4 5
y = 9 − 2x y 7 5 3 1 −1
Mr. N.: Let’s see the different ways the groups used
to solve this.
Fig. 2 (Continued)
521
522
S.: Well we used it like the ones we did up there x +y =5

(refers to 9x − 5 = 2x + 16) and found that y transpose x x 1 2 3 4 5
equals nine minus two x and five minus x; so y =5−x y 4 3 2 1 0
then we found what numbers were the same for
both equations which were four and one.
Mr. N.: Very good, so you knew that the numbers
had to be true for both equations so you set up a ta-
ble to find out which two had the same pair of num-
bers. Class how fast do you think this will work?
S.: Very slow.
Nr. N.: Yew, while it gives us the correct numbers it
will not help us work very quickly.
Mr. N.: Let’s look at another group, they used the
diagrams.
S.: Well we knew that the circle and the square were
the same as five, so we made the two circles circle
plus circle, then we put in the five and found that the
circle was four, then four plus one is five.
Mr. N.: Very good, so you noticed that this was the
same as this (rewrites pictorial model with symbols)
and then you said you know that this is the same as
five so you put that in and then it was easy to find
that x equals four, so then y equals one if x plus y
must be five.
Fig. 2 (Continued)
M. Smith
Mr. N.: Okay our last group has only equations.
S: Yes we started like the first group but didn’t use
the table, we knew they were the same so they must
be equal.
Mr. N.: Why do you think that?
S.: Well it is like the balance we used that the two
things are the same thing so they must be balanced
with each other. . .so we made them equal to each
other and found x equals four and then put that into
x plus y equals 5 and found that y is one.
Mr. N.: Very good you knew that these must be the
same thing so that they would have to be equal.
What do you think about this method class?
S.: It is much faster.
Teacher re-visits the solution strategies again de-
scribing the how each gives a correct answer but
how the last is more efficient.
Mr. N.: Good now let’s try to do this problem. I will Takako and Kumiko were shopping. Takako bought three notebooks and four pen-
give you a few minutes cils for 750 yen. Kumiko bought three notebooks and two pencils for 450 yen.
How much does one pencil cost?
Fig. 2 (Continued)
523
524
After students work on this problem, the teacher 3x + 4y = 750

allows one student to show how she solved it. 3x + 2y = 450
S.: Well know that each paid the same for the 3x = 450 − 2y
pencils and notebooks so I know that if three x is 3x = 750 − 4y
how much Takako spent on notebooks, then
Kumiko also spent 3x on notebooks. Takako 750 − 4y = 450 − 2y
bought four pencils so she spent 4y on pencils, 750 − 450 = −2y + 4y
but Takako only bought two, so she has 2y for 300 = 2y
pencils. Since both of these need to have the 150 = y
same price for notebooks and pencils I need to
find out which one. Each of these has three x so I So pencils cost 150 yen each
used those to make them equal to each other
since three is the same thing in both equations.
So then since they are equal I calculated that the
pencils are 150 yen each.
Mr. N.: A very good solution.
Mr. Nakamura rephrases the students explanation
and solution and summarizes the main strategies
learned during the lesson.
Fig. 2 (Continued)
M. Smith
• Japanese teachers used expressions to generalize; U.S. teachers focused on an-

swers and the steps to get the answer.
In the Japanese lesson, Mr. Nakamura places a lot of attention on generalizing
both ideas and solution processes. This can be seen as he helps students examine
the usefulness of solution methods to solve the problem and the connections across
these methods. In contrast, Mr. Kirkyle treats each problem with its own set of steps
to solve the problem. The need to move beyond procedural approaches of solving
algebra to a better understanding of the general relationships has been well docu-
mented (Carraher and Schliemann 2007; Kieran 1992, 2007). However, this transi-
tion is difficult for students, as they tend to focus on specific solutions to specific
problems (Khng and Khng 2009). By helping students connect the different solution
methods, as well as connecting the procedures to the problem solving context, Mr.
Nakamura was helping students understand the relationships being represented by
the operations as well as across models for showing these relationships.
• Japanese teachers addressed how variables were used to express variation; U.S.
teachers focused on variables as unknowns or place holders.
In the Japanese lesson, Mr. Nakamura helps focus students’ attentions on the
role of the variables in the equations. He highlights this most clearly when he looks
at Group One’s solution, pointing out how the set of numbers that solved the first
equation solved the second equation. It would be easy for Mr. Nakamura to have
dismissed this solution strategy as being too rudimentary to present, but by present-
ing this solution strategy he was able to help students see how the pairs of number
changed (varied) in each equation. Mr. Kirkyle began to address this very important
connection in the first set of infinite solution simultaneous equations, but did not
help articulate how he knew what numbers to use in these equations or how more
than the one pair he found could also have satisfied these equations. This limited
exposure appeared to be more focused on getting the calculations correct than in
understanding the role of variables in expressions.
Early introduction to algebra often focuses on solving for an unknown (Kieran
1992); however, the mathematical use of a variable is much more powerful than that
as it allows for the study of relationships. This focus on the variable as a thing that
changes is an important component of understanding the relationships expressed in
equations, which later are studied as functions (Kieran 2007; Malisani and Spagnolo
2009). It is easy to see how the variable can become yet another symbol to be oper-
ated upon, so careful consideration needs to be given to the connections required to
draw out the relationships they represent.
• Japanese teachers looked at the relationship expressed by equality; U.S. teachers
treated the equal sign as an operator.
It is clear that Mr. Nakamura is building a foundation for the relationship of
equality; this is most evident as Groups Two and Three use the equality to gener-
ate solutions to the simultaneous equations. Group Three’s solution indicates that
the students are able to think flexibly about how the equivalence must be main-
tained across both equations. Group Two’s use of equality indicates a fairly deep
526 M. Smith
understanding of equality and simultaneous equations (even if the solution method

is viewed as less symbolically sophisticated) because of it focus on an expression,
not just a variable, being equivalent. In contrast, Mr. Kirkyle pays little attention to
the relationship; this is most notable when he solves simultaneous equations from
parallel lines. In particular, Mr. Kirkyle comes up with the unusual statement of
0 = −84 and tells students this means the lines are parallel, without indentifying
the fallacy of the statement. This lack of attention may seem trivial to some, but by
leaving this statement as equivalent it makes it difficult for students to move beyond
using the equal sign as a command to compute because in this case it is simply a
means to an end.
The importance of equivalence as a relationship appears in many places (Kieran
1981, 1992, 2007). In particular, the research indicates that even though students can
reason about equivalence as a mathematical operator, or relationship, it is most com-
monly viewed as a command (Stacey and MacGregor 1999). However, the research
also indicates that the need for the teacher to help students make this connection
is critical (Molina et al. 2009; Stacey and MacGregor 1999). Molina et al. (2009)
found that students who received instruction focused on building this meaning were
able to reason about the relationships of this operator.
Conclusions
In this chapter two contrasting ways of introducing students to algebraic expression

were presented. The purpose of these examples was to help highlight some of the
key differences in presenting algebra to students, one which focused on studying re-
lationships expressed by simultaneous equations and one which focused on building
efficient procedures of solving simultaneous equations. Some of the key differences
across these lessons indicate that building connections into the instructional process
can help illuminate some of these important components in the study of algebra.
References
Arafeh, S., & McLaughlin, M. (2002) Legal and Ethical Issues in the Use of Video in Education
(NCES2002-01). Washington, DC: National Center for Educational Statistics.
Arcavi, A. (2008). Algebra: Purpose and empowerment. In C. E. Greenes & R. Rubenstein (Eds.),
Algebra and Algebraic Thinking in the Schools (pp. 37–49). Reston, VA: National Council of
Teachers of Mathematics.
Blanton, M. L., & Kaput, J. J. (2003, October). Developing elementary teacher’s “algebra eyes and
ears.” Teaching Children Mathematics, 70–77.
Carpenter, T. P., & Levi, L. (1999, April). Developing conceptions of algebraic reasoning in the
primary grades. Paper presented at the annual meeting of the American Educational Research
Association, Montreal, Quebec.
Carpenter, T. P., Franke, M. L., & Levi, L. (2003). Thinking Mathematically. Portsmouth, NH:
Hienemann.
Carraher, & Schliemann (2007). Early algebra and algebraic reasoning. In F. K. Lester (Ed.), Sec-
ond Handbook of Research on Mathematics Teaching and Learning (pp. 669–705). Reston,
VA: National Council of Teachers of Mathematics.
Chazan, D. (2008). The shifting landscape of school algebra in the United States. In C. E. Greenes
& R. Rubenstein (Eds.), Algebra and Algebraic Thinking in the Schools (pp. 19–33). Reston,
VA: National Council of Teachers of Mathematics.
Falkner, K. P., Levi, L., & Carpenter, T. P. (1999). Children’s understanding of equality. Teaching
Children Mathematics, 6, 232–236.
Fujii, T., & Stephens, M. (2001). Fostering and understanding of algebraic generalization through
numerical expressions; The role of “quasi-variables. In K. Stacey, H. Chick, & M. Kendal
(Eds.), The Future of the Teaching and Learning of Algebra: The 12th ICMI Study (pp. 259–
264). Boston: Kluwer Academic Publishers.
Fujii, T., & Stephens, M. (2008). Using number sentences to introduce the idea of the variable.
In C. E. Greenes & R. Rubenstein (Eds.), Algebra and Algebraic Thinking in the Schools (pp.
127–140). Reston, VA: National Council of Teachers of Mathematics.
Gonzales, P., Guzmán, J. C., Partelow, L., Pahlke, E., Jocelyn, L., Kastberg, D., & Williams, T.
(2005). Highlights from the Trends in International Mathematics and Science Study (TIMSS)
2003 (NCES 2005005). Washington, DC: National Center for Educational Statistics.
Gonzales, P., Williams, T., Jocelyn, L., Roey, S., Kastberg, D., & Brenwald, S. (2009). Highlights
from TIMSS 2007: Mathematics and Science Achievement of U.S. Fourth- and Eighth-Grade
Students in an International Context (NCES 2009001). Washington, DC: National Center for
Educational Statistics.
Jacobs, J. K., & Morita, E. (2002). Japanese and American teachers’ evaluations of videotaped
mathematics lessons. Journal for Research in Mathematics Education, 3(33), 154–175.
Jacobs, J. K., Hiebert, J., Givvin, K. B., Hollingsworth, H., Garnier, H., & Wearne, D. (2006).
Does eighth-grade mathematics teaching in the United States align with the NCTM Standards?
Results from the TIMSS 1995 and 1999 Video Studies. Journal for Research in Mathematics
Education, 37(1), 5–32.
Jacobs, J. K., Holligsworht, H., & Givvin, K. B. (2007). Video-based research made “easy”:
Methodological lessons learned from the TIMSS Video Studies. Field Methods, 19(3), 284–
299.
Kaput (2007). What is algebra? What is algebraic reasoning? In D. Carraher & M. Blanton (Eds.),
Algebra in the Early Grades. London: Routledge.
matics, 12, 317–326.
Kieran, C. (1992). The learning and teaching of school algebra. In D. A. Grouws (Ed.), Hand-
book of Research on Mathematics Teaching and Learning (pp. 390–419). New York: MacGraw
Publishing.
Kieran, C. (2004). The equation/inequality connection in constructing meaning for inequality sit-
uations. In M. J. Hoines & A. B. Fuglestad (Eds.), Proceedings of the 28th Conference of the
International Group for the Psychology of Mathematics Education, Bergen, Norway (pp. 143–
147).
Kieran, C. (2007). Learning and teaching of algebra at the middle school through college levels:
Building meaning for symbols and their manipulations. In F. K. Lester (Ed.), Second Hand-
book of Research on Mathematics Teaching and Learning (pp. 707–762). Reston, VA: National
Khng, K. H., & Khng, K. (2009). Inhibiting interference from prior knowledge: Arithmetic intru-
sions in algebra word problem solving. Learning and Individual Differences, 19, 262–268.
Knuth, E. J., Stephens, A. C., McNeil, N. M., & Alibali, M. W. (2006). Does understanding the
equal sign matter? Evidence from solving equations. Journal for Research in Mathematics
Education, 37(4), 297–312.
Levin, M. (2008). The potential for developing algebraic thinking from purposeful guessing and
checking. In G. Kanselaar, J. van Merriënboer, P. Kirschner, & T. de Jong (Eds.), Proceedings
of the International Conference of the Learning Sciences. Utrecht, The Netherlands: ICLS.
528 M. Smith
Lins, R., & Kaput, J. J. (2004). The early development of algebraic reasoning: The current state of
the field. In K. Stacey, H. Chick, & M. Kendal (Eds.), The Future of the Teaching and Learning
of Algebra: The 12th ICMI Study (pp. 47–70). Boston: Kluwer Academic Publishers.
Malisani, E., & Spagnolo, F. (2009). From arithmetical thought to algebraic thought: The role of
the variable. Educational Studies in Mathematics, 71(1), 19–41.
Molina, M., Castro, E., & Castro, E. (2009). Elementary students understanding of the equal sign in
number sentences. Electronic Journal of Research in Educational Psychology, 7(17), 341–368.
Radford, L. (1996). Some reflections on teahing algebra trough generalization. In Bednarz et al.
(Eds.), Approaches to Algebra (pp. 107–113). Dordrecht: Kluwer.
Smith, M. (2000). A comparison of the types of mathematics tasks and how they were completed
during eighth-grade Mathematics instruction in Germany, Japan, and the United States. Un-
published doctoral dissertation, University of Delaware.
Stacey, K., & MacGregor, M. (1999). Learning the algebraic method of solving problems. Journal
of Mathematical Behaviour, 18(20), 149–167.
Stigler, J. S., & Hiebert, J. (1997). Understanding and improving classroom instruction: An
overview of TIMSS Video Study. Phi Delta Kappan, 79(1), 14–21.
Stigler, J. W., & Hiebert, J. (1998). Teaching is a cultural activity. American Educator, 22(4), 4–11.
Stigler, J. W., & Hiebert, J. (1999). The Teaching Gap. New York: Free Press.
Stigler, J. W., Gonzales, P. A., Kawanka, T., Knoll, S., & Serrano, A. (1999). The TIMSS Videotape
Classroom Study: Methods and Findings from an Exploratory Research Project on Eighth-
Grade Mathematics Instruction in Germany, Japan, and the United States (NCES 1999074).
Washington, DC: National Center for Educational Statistics.
Swafford, J. O., & Langrall, C. W. (2000). Grade 6 students’ preinstructional use of equations to
describe and represent problem situations. Journal for Research in Mathematics Education,
31(1), 89–112.
Thompson, P. W., & Thompson, A. G. (1994). Talking about rates conceptually, Part 1: A teacher’s
struggle. Journal for Research in Mathematics Education, 25(3), 279–303.
Teaching Algebraic Equations with Variation
in Chinese Classroom
Jing Li, Aihui Peng, and Naiqing Song
Abstract This chapter gives a detailed analysis of how teaching with variation
is helpful for students’ learning of algebraic equations by using typical teaching
episodes in grade seven in China. Also, it provides a demonstration showing how
variation is used as an effective way of teaching through the discussion after the
analysis.
Introduction
International studies of mathematics achievement show that in the past decades,

East Asian students have consistently outperformed their counterparts in Western
countries (Stevenson and Lee 1990; Lapointe et al. 1989; Husen 1967). Findings
from the increased interest in the study of mathematics classrooms in East Asian
countries suggest that teaching with variation, a common characteristic of mathe-
matics education in East Asian countries, could be one of the powerful explana-
tions for the gap (Sun 2009; Park 2006; Huang and Leung 2004). Furthermore,
The research reported here is supported by a granted project from the 2010 National Education
Science Planng Fund in China ( ) (Grant No. GOA107019), entitled Research
into Teaching with Variation Based on Multi-representations about Junior Algebraic—A Reform
Experiment on “Teaching from the Perspective of Learning”.
J. Li · N. Song ()
School of Mathematics and Statistics, Southwest University, Chongqing, 40071, China
J. Li
School of Mathematics and Information, Langfang Teachers’ College, Langfang, 065000, China
A. Peng
Institute of Higher Education, Southwest University, Chongqing, 400715, China

530 J. Li et al.
studies in mathematics education and cross-cultural psychology aiming to reveal

the reasons for the contradictory situation called “The Confucian Heritage Culture
(CHC) Learner Paradox”1 show that teaching with variation may be helpful to un-
cover the paradox where, even in large classes, students still can actively involve
themselves in the process of learning and achieve excellent results (Gu et al. 2004;
Biggs and Watkins 1996). In China, “teaching with variation” has been applied ei-
ther consciously or intuitively for a long time, and it has almost become the teaching
routine for Chinese mathematics teachers (Marton et al. 2004).
Algebra has been identified as the most important “gatekeeper” in mathematics
(Cai et al. 2005). While algebraic equation plays a key part in the learning of algebra,
it is a challenge to students because it is probably the first time they really understand
the meaning of the equal sign from their previous arithmetic experiences (Pirie and
Martin 1997). Meanwhile, the studies suggest that many students also have difficulty
in grasping the syntax or structure of algebraic expressions and do not understand
either the procedures for transforming equations or the reasons why transformations
are done the way they are (National Mathematics Advisory Panel 2008; Pirie and
Martin 1997). These and other difficulties are compounded as equations become
more complex when students attempt to solve word problems (Vlassis 2002). In
a summary on research about the learning and teaching of school algebra, Kieran
(1992, p. 390) poses these questions, “What makes the comprehension of school
algebra a difficult task for the majority? Is the content of algebra the source of the
problem? Or, is it the way it is taught that causes students to be unable to make sense
of the subject?”
However, studies also show positive findings of teaching algebraic equations with
variation in China. According to Chen and Song (1996), in spite of the situation of
whole-class instruction, individual-supplemented counseling, large classes with an
average of more than 50 junior middle students, natural classes (classified randomly)
with various ability levels of students, the performance of the algebra teaching pro-
cess in China is successful in that the students’ interests in learning algebra are in-
spired through variants provided by the teacher, and they have ultimately developed
algebraic thinking. Naturally, one might wonder how algebraic equations are taught
effectively in China. Using typical teaching episodes in grade seven, this study will
give a detailed analysis showing how teaching with variation helps develop students’
algebraic thinking. What must be mentioned here is that—although there is a grow-
ing agreement that teaching with variation could be regarded as a kind of Chinese
wisdom of mathematics teaching (Sun 2009) and although teaching with variation
is gaining more and more attention in mathematics education (Watson and Mason
2006)—there is little empirical data available to confirm the promise of teaching
with variation (Cai and Nie 2007). This empirical study aims to contribute to the
development and refinement of this practice.
1 The CHC Learner Paradox: CHC students are perceived as using low-level, rote- based strategies
in a classroom environment, which should not be conducive to high achievement, yet CHC students
report a preference for high-level, meaning-based learning strategies, and they achieve significantly
better in international assessments.
Teaching Algebraic Equations with Variation in Chinese Classroom 531
In the following sections, the study will start with the source of the data, followed
by the theoretical framework. Then the methods of research will be presented, fol-
lowed by analyses of the data. Finally, there will be a discussion and conclusion.
The Source of the Data
The data in this study were collected as a result of a broad background of teach-
ing with variation in China. During the past several years, mathematics teaching in
China has undergone significant changes. Due to historic and realistic reasons, some
typical features in mathematics education in China are characterized by teaching
difficult contents, emphasizing the rigor, abstraction, and application of mathemat-
ics. However, as a result of the mathematics education reform around the world,
teaching contents are gradually changing to adapt to students’ interest in mathe-
matics learning and to relate to real life. At the same time, the traditional teaching
method emphasizing practice is gradually moving towards the method of combining
learning and discovery, teacher-student interaction, group cooperative learning, etc.
Thus, teaching with variation has developed with the times.
The data for teaching algebraic equations with variation in grade seven were col-
lected from video tapes of classroom teaching. Researchers selected the data from
typical teaching episodes continuous in content covering “equation introduction,
equation understanding, equation solving, and equation application” carried out in
two classes in different middle schools in China in 2009. The teaching episodes in-
clude 7 periods of class, each lasting 45 minutes. The students in those classes are
13 or 14 years old, and their mathematics teachers, Zhang Shang and Zhao Bing,2
are experienced in their schools. Although the teachers’ exposition and whole-class
discussion were videotaped and audio recorded, the researchers took notes when
the students were working in groups. For an in-depth analysis of how the teachers
conducted their teaching and the learning sequences, the researchers interviewed
the two teachers and their student representatives after the classroom observations.
Although the classes were originally conducted in Chinese, the version offered in
this paper is the English translation. To better understand the lessons, the teaching
plan for “Teaching Equation Based on Problem Variation” along with reflections
after the lesson is attached.
Theoretical Framework
This study is set within the perspective of the theory of teaching with variations de-
veloped by Gu (1994). Gu (1999) stated that teaching with variation is an important
2 All the names are pseudonym.

532 J. Li et al.
method through which students can definitely master relevant concepts. And, it il-
lustrates the essential features by using different forms of visual materials and some-
times highlighting the essence of a concept by changing the nonessential features.
Its aim is to understanding the essence of an object and to form a scientific concept
by ignoring nonessential features. Based on a series of longitudinal mathematics
teaching experiments in China, Gu (1994) systematically analyzed and synthesized
the concepts of teaching with variation. He identified and illustrated the two forms
of variations, namely “conceptual variation” and “procedural variation,” referring
to understanding concepts from multiple perspectives and then gradually unfolding
mathematics.
Teaching with variation was developed in China, but it is strongly supported
by several well-known Western theories of learning and teaching. Marton’s theory
offers an epistemological foundation and conceptual support for the theory of teach-
ing with variation. According to Marton et al. (2004), learning is a process in which
learners develop a certain capability or a certain way of seeing or experiencing.
In order to see something in a certain way, the learner must discern certain fea-
tures of the object. Experiencing variation is essential for discernment and is thus
significant for learning content. Marton et al. (2004) argue that it is important to
attend to what varies and what is invariant in a learning situation. What’s more, as
put forward by Ausubel (1968), only by establishing a reasonable and substantial
connection between learners’ new and old knowledge can meaningful learning take
place, and such a connection is reasonable and essential between new knowledge
and some specific aspects of prior knowledge (e.g. a symbol, a concept or an ex-
ample in the representations of learners’ cognitive structures). For learning algebra
with variation, students, on the one hand, need to understand the concept from mul-
tiple perspectives (from concrete to abstract, from specific to general) to acquire the
nature and connections of concepts by eliminating background interferences, high-
lighting the essence of mathematic concepts, and clarifying the connotation of the
concept of algebra. On the other hand, when students understand the origin and use
of mathematic knowledge, gain experiences in concept formation and problem solv-
ing, apprehend different components of mathematics, they improve their knowledge
structure of mathematics; therefore, they establish proper connections between new
and old knowledge of mathematics. As pointed out by Cai and Nie (2007), teaching
with variation, by presenting a series of interconnected problems, can help students
understand concepts and master the problem-solving method, thereby developing
students’ knowledge of mathematics. Thus, it is clear that this method of teaching
can promote students’ meaningful learning of mathematics.
As was pointed out by Gu et al. (2004), teaching with variation helps students
develop a profound understanding of a concept from multiple perspectives. They do
this by discerning certain features of the denotation or by the denotation itself of the
concept, both of which are easily confused (i.e. conceptual variation). For example,
the concept of equation can be introduced to students by using visual or specific vari-
ations; the essence of the concept can be highlighted by contrasting standard forms
with non-standard ones; and misunderstandings of the concept can be corrected by
using counter-examples. Moreover, using activities of teaching with variation, stu-
dents can understand how the concepts are generated and developed, acquire the
representations and strategies of problem solving, and then build a hierarchical re-
lationship among different concepts (i.e. procedural variation). For instance, in the
“procedural variation” process, the teacher designs the operational model of the
concept of equation to promote the formation of the concept of equation, scaffolds
stratified problems to form problem-solving strategies of equation, and adopts mul-
tiple solutions or variations for one problem or one solution for multiple problems
so students acquire particular experiences in problem solving with equations.
It is believed that teaching algebraic equations based on appropriate varia-
tions (represented as multi-variant stratified problem space consisting of system-
atically and deliberately varying problems) can help students understand the con-
cept of equation and facilitate their development in representations and strategies
of problem-solving, thereby making the students’ learning of algebraic equations
meaningful. Thus, it not only helps students grasp the “Two Basics” (basic knowl-
edge and basic skills) of algebraic equation training, but it also enhances their
problem-solving abilities. Eventually, students’ ability in algebraic reasoning (i.e.
the ability of imagining, representing and thinking of connections in algebraic
knowledge and problem-solving strategies following certain procedures and struc-
tures) is cultivated and developed.
The Method of Research

This study—giving priority to the observation of episodes of equation teaching
and supplemented by after-class interview with teachers and students—analyzed
the specific operation of teaching algebraic equations with variation and the cor-
responding learning by students. This was done using qualitative research of the
continuous teaching episodes of algebraic equations with variation conducted in the
natural environment of middle schools. Using this process, researchers analyzed
and summarized the operational rationality of teaching algebraic with variation in
enacting students’ effective learning.
Analysis of Data
In the following four sub-sections, teaching algebraic equations with variation is
demonstrated in detail with typical teaching episodes, in the order of knowledge
development of equations, including equation introduction, equation understanding,
equation solving, and equation application. The data will be analyzed to account for
the successful growth of understanding in students’ learning.
The Introduction of the Concept of Equation
There are many studies showing that the concept of equation is challenging for stu-
dents because of their inability to spontaneously operate with or on the unknown
534 J. Li et al.
(Herscovics and Linchevski 1994; Sadovsky and Sessa 2005) and also due to their
misunderstanding of the equal sign as an operator, that is, as a signal for “doing
something” rather than a relational symbol of equivalence or quantity sameness
(Behr et al. 1980). This session will show how the concept of equation is introduced
with different variations to eliminate the potential cognitive obstacles by analyzing
observed teaching episodes.
Teaching episode: Problem variation to enhance the formation of concept of
equation
Teacher: Here is a problem: Xiaoming bought 3 pieces of chocolate with 3 yuan,3 and got
change of 6 jiao. How much was one piece of chocolate?
Teacher: This problem may be shown as an expression: 3 yuan − 3( ) = 6 jiao. The number
in the parentheses is unknown and it can be replaced by x, namely 3 yuan − 3x =
6 jiao, according to 30 − 3x = 6.
Teacher: How can we express this problem directly with an equation?
Students are required to look into the problem variations. With the guidance of the teacher,
students are aware of the transition from “concrete representation” to “symbolic represen-
tation,” and summarize the process as the following:
Suppose that each piece of chocolate costs x jiao, the following equation is obtained through
unifying the units: 30 − 3x = 6.
Teacher: Please follow the above example and formulate expressions for the following
problems:
Problem 1: Xiaohong had 9 yuan. She bought 4 notebooks with this money and
got 8 jiao in change. How much did one notebook cost?
Problem 2: Xiaoli helped her mother buy fruit. She bought oranges with 10 yuan,
and bought 5 jin4 of grapes. She spent 12.5 yuan in total. How much did one jin
of grapes cost?
Question 3: Xiaoqing bought 4 batteries with 8.5 yuan and got change of 0.1 yuan.
How much did one battery cost?
Question 4: Xiaogang went to buy stamps with Daming. Xiaogang bought 8 and
Daming bought 6. Daming spent 6 yuan less than Xiaogang. How much did one
stamp cost?
Question 5: In 50 basketballs, there are 30 red ones and 20 yellow ones, how many
basketballs are there?
Question 6: Xiaolan has 50 apples, 30 oranges, which one is the larger number?
(Classroom observation: April 14, 2009)
From classroom observation, we found that, after practice with the varying prob-
lems, students could write out the expressions, showing that they successfully ex-
perienced the transition from verbal representation to symbolic representation with
regard to simple equations. By using different problem variations, understanding
of equal sign which is a sophisticated and essential concept for understanding of
algebraic equations—was enhanced. We also observed that, based on these experi-
ences, students were also asked to classify expressions that differentiated essential
and nonessential features of equation using problem variation, which is shown in
the following example.
3 One yuan in the Chinese monetary system equals ten jiao.

4 One jin is equal to 500 grams.
Teaching episode: Problem variation to differentiate essential and nonessential

features of equation
Teacher: Look at the following equations and answer this question: How many types can
they be divided into?
30−3x = 6, 90−4x = 8, 10+ 5x = 12.5, 8.5−4x = 0.1, 8x −5x = 6, 20+ 30 =
50, 50 > 30.
While students are discussing the evolving process of the question, the teacher helps those,
who have difficulties in understanding, by filling in the blanks. These expressions can
be classified into “equations” and “non-equations,” with the former further classified into
“equations with unknowns” and “equations without unknowns.”
Teacher: With regard to the equal expression, 30 − 3x = 6, 90 − 4x = 8, 10 + 5x = 12.5,
8.5 − 4x = 0.1 and 8x − 5x = 6 are defined as equations. An equation is defined as the
expression which includes unknowns and the equal sign.
According to our observation, the essential features of equation were differen-

tiated by two types of variations. One is based on whether there is an equal sign
in the expression. This includes two types of expressions, 30 − 3x = 6, 90 − 4x =
8, 10 + 5x = 12.5, 8.5 − 4x = 0.1, 8x − 5x = 6, 20 + 30 = 50, and 50 > 30. The
other is based on the number of unknowns. This includes 30 − 3x = 6, 90 − 4x = 8,
10 + 5x = 12.5, 8.5 − 4x = 0.1, 8x − 5x = 6, and 20 + 30 = 50, 50 > 30. With the
variation, students had deeper understanding of the concepts and lay the groundwork
for better understanding of equation. The following interview transcript provides
more evidence:
Interviewer: Do you students have any ideas about the concept of equation?
Student 1: It is very easy. The key point is to recognize whether there are an equal sign and
unknowns.
Student 2: I think so.
Interviewer: Do you like the way that your teacher taught in this lesson?
Student 1: Yeah. I think those different expressions really make sense to me.
Student 2: When I read the definition of equation in the book, I feel it is hard to understand.
But when our teacher presents those examples, I have a feeling of “I got it.”
(Interview transcript: April 14, 2009)
From the analysis of teaching episodes and interviews, it is clear that the loca-
tion of variation is rather critical. Through changing nonessential features of things,
students understand the essential features of things, and they gradually make use
of the cognitive models to reduce their memory burden, rather than “changing for
change’s sake” to increase their cognitive burden.
The Improvement of Understanding of Equation
Connotation and denotation (extension) of a concept are two opposite yet comple-
mentary aspects. If the connotation is clarified, the extension is defined and vice
536 J. Li et al.
versa. Understanding the concept of equation includes its connotations and deno-
tations. This session of observed lessons will show how the essential nature of the
equation is consolidated by designing problem variation, putting emphasis on clar-
ifying the connotation, and differentiating the boundary of the set of objects in the
extension.
Teaching episode: Problem variation to solidify essential feature of equation
Teacher: Point out which of the following are equations and which are not.
2x = 1; 3x + 4 = 7; 4y − 3 = 5; 3x + 4y = 12; x 2 − 1 = 0, x 2 + y 2 = 1,
2x − 3 > 2; 2 + 6 = 8; x = 8; 1 + 8.
After discussion in groups, representatives of the groups are required to give

answers.
During the lesson, we observed that, to solidify essential feature of equation
through problem variation above, the teacher helped those students—who have dif-
ficulties in understanding—learn to recognize and emphasize two key points for
the judgment: equal expression and unknown numbers. The teacher provided stu-
dents with varying exercises to help them understand equations from multiple per-
spectives, such as highlighting the essence of equation concept by contrasting non-
standard form x = 8 and using counter-examples. These activities were continued
according to students’ classroom performance until they were skilled in judging
equations. Thus, by identifying the different types of equation variations, the stu-
dents clearly recognized the characteristics of the conceptual representation of equa-
tions. Based on this, the teacher further used concept variation to help them recog-
nize the essential feature of equation with one unknown.
Teaching episode: Concept variation to discern the essential feature of linear
equation with one unknown
Teacher: Fill out the parentheses after identifying all the equations from the above equal
expressions:
There are ( ) equations with more than one unknown, ( ) with unknowns of higher
than the power of one and ( ) with one unknown of the power of one.
Our observation shows that concept variation was designed to discern the two
essential features of linear equation with one unknown; namely, the power of one
and one unknown. And the equations with the power of one were first identified,
including 2x = 1, 3x + 4 = 7, 4y − 3 = 5, 3x + 4y = 12, x = 8; then, they was
further identified according to one unknown, including 2x = 1, 3x + 4 = 7, x = 8.
Using a series of problems in which the essential features of equations were kept
unchanged but the nonessential features of equations were changed students easily
caught the meaning of linear equation with one unknown.
The teacher helped students learn linear equation with one unknown, step by
step. This strategy maximized the advantages of teaching concepts with variations.
After the teacher and students summed up the concept of linear equation with one
unknown, students did relevant exercises to further understand the features of linear
equation with one unknown. It was noted that learning linear equation with one
unknown helped students to deepen their understanding of the equation concept as
a whole.
Teaching episode: Problem variation to percept transition from arithmetic and

algebra
Teacher: After learning the equation concept, how can we find the value of the unknown?
Please answer the following question using as many methods as possible.
Xiaoming bought 3 pieces of chocolate with 3 yuan, and got change of 6 jiao.
Problem: How much was one piece of chocolate?
Method 1: Use arithmetic.
(30 − 6) ÷ 3 = 8 (jiao)
Method 2: Use equation or algorithm to change the equation from 30 − 3x = 6 to
x = 8.
30 − 3x = 6
Divided by 3 from both sides, then, 10 − x = 2
Minus 2 from both sides, then x = 10 − 2, so
x=8
After discussion in groups, the teacher gave the instructional explanation, “From
the answer above, we can see that, in the arithmetic approach, the unknown is repre-
sented by a formula including only the given; however, in the algebraic phase, first
the unknown is supposed, then the relationship between the given and the unknown
is written in a equation, and finally we find value of the unknown in the equation.
Now, do some word problems with two methods of arithmetic and algebra, and ex-
perience their solving.”
After getting the answers using the two methods, the teacher encouraged the stu-
dents to compare the two methods, i.e., synthesis and analysis, to experience the
advantages of analysis in problem solving. With more practice of “multiple solu-
tions for one problem,” the students understood the transition from the “arithmetic
approach” to the “algebraic approach.”
Teaching episode: Problem variation to experience the process of equation solv-
ing
Teacher: Since the second method has advantages over the first in problem solving, we will
use equation or algorithm to answer the questions listed above in the first session.
Please answer question 1 above by using equation.
90 − 4x = 8
Divided by 2 from both sides, then, 45 − 2x = 4
Minus 45 from both sides, then, 2x = 45 − 4,
2x = 41
Divided by 2 from both sides, then, x = 20.5.
Emphasizing “equality prosperities and algorithm rules” when solving equations

helps students understand the idea of transformation thinking first, as they clarify
the reasons and purposes for each step. Finally, students master solving equation
through doing a number of variation exercises.
Teacher: How can we check whether the solution we got is correct?
Please put the x = 8, x = 20.5, x = 0.5 into equations of 30 − 3x = 6, 90 − 4x =
8, 10 + 5x = 12.5 and find out whether equations are tenable.
After doing this, students found out: (1) that every number matches one spe-
cific equation, (2) and that the steps of equation solving are reasonable. Through
538 J. Li et al.
practice, the students have the potential to appreciate the idea of the function and
corresponding relationship.
Teacher: It can be seen that x = 8, x = 20.5, x = 0.5 match the equations of 30 − 3x =
6, 90 − 4x = 8, 10 + 5x = 12.5 respectively. When a value of a unknown is substituted into
an equation with one unknown, the equation is tenable, then, the value of the unknown is
called as “a solution of the equation” or “a root of the equation” when the equation has only
one unknown number.
The process to get the solution is called “equation solving.”
By looking back on the process of equation solving, students understand the

meaning of “solution of equation” and “equation solving” and form an understand-
ing of the process representation.
Teacher: (1) Which is the solution of 3x = 4 + 2x, x = 3 or x = 4?
(2) Write out a simple equation with the solution of x = 6.
(3) What is the solution of 3x − 8 = 10?
Problems with variation set by the teacher encourage students’ regular thinking
and reverse thinking; therefore, these problems help students understand the equa-
tion solution and its representations from multiple perspectives (conceptual varia-
tion). Through teacher’s and students’ analysis of the difference between the equa-
tion and non-equation, the students get a deeper understanding of the meaning of
equation. As a result, cognition of the equation representation tends to be abstract.
For a different evolution of equation solution, the students accumulate the dynamics
representation of equation solving, and this sets up the potential to lay the ground-
work for further learning of solving equation.
It is necessary to help students transfer different representations of the concept of
equations: symbols and words. When understanding equation solution, it is impor-
tant to make clear its meaning statically and dynamically so students can experience
its creation through procedural variation. In addition, the meaning of key words
and expressions is also taken into consideration when training students for accuracy
in algebraic language because this helps develop their algebraic thinking. Setting
concept variation helps students understand equations, and it is suitable to apply
procedural variation to the understanding of equation solution.
Therefore, to understand the topic broadly, such conceptual variations as linguis-
tic and visual representations (physical objects, diagrams, etc.) should be adopted;
to understand the topic in depth, procedural variations from concrete to abstract
representations should be adopted to build students’ symbolic representations.
This teaching episode makes it absolutely clear that types of variation are very
important. The understanding of concepts cannot be done without identifying, gen-
eralizing, and abstracting various types of examples related to concepts, nor can it do
without the judging and thinking about these examples. The types of variation used
in teaching with variation should focus on understanding concepts, and emerge step
by step according to students’ cognitive level, and thereby ensuring a substantive
improvement of students’ learning quality.
Equation Solving
This session will show how varying equations from simple to complex are designed
to help students grasp the unchanging “routines” and “methods” when solving equa-
tions through procedural variation.
Teaching episode: Procedural variation to experience the generalization of equa-
tion solving
Teacher: Do such simple equations with one unknown so different in formats as 5x − 2 = 8,
4 x = − 2 x + 3, 4(x + 0.5) + x = 17, 7 (x + 14) = 4 (x + 20), 5 (x + 15) = 2 −
1 1 1 1 1 1
3 (x − 7)
1
share a common solving process?

Teacher: Taking the equations listed above as examples to interpret the idea of equation
solving: “simplification”.
transpose convert coefficient of x to 1
5x − 2 = 8 −−−→ 5x = 8 + 2 −−−−→ x = 2
(property of equality 1) (property of equality 2)
transpose combine of similar terms change coefficient of x to 1

1 1 1 1 3
= x = − x + 3 −−−−→ x + x = 3 −−−−→ x = 3 −−−−→ x = 4

4 2 4 2 4
(property of equality 1) (property of equality 2)
remove brackets transpose combine similar terms
4(x + 0.5) + x = 17 −−−→ 4x + 2 + x = 17 −

→ 4x + x = 17 − 2 −−−−→
change coefficient of x to 1
5x = 15 −−−→ x = 3
remove denominator remove brackets
❃ 17 (x + 14) = 14 (x + 20) −−−−→ 4(x + 14) = 7(x + 20) −−−−→
(property of equality 2)
transpose and combine similar terms change coefficient of x to 1
4x + 56 = 7x + 140 −−−−→ −3x = 84 −−−−→ x = −28
(Note: that brackets in this exercise ❃ can be removed in the first step.)
(Classroom observation: May 13, 2009)
The purpose for students’ progressive and hierarchical practice of variation is to

help them to be skilled in equation solving.
Note here that as to property of equality 1, if a = b, then a + c = b + c; if
a = b, then a − c = b − c. As to property of equality 2, if a = b, then ac = bc; if
a = b and c = 0, then a/c = b/c. Through the training in different levels, students’
understanding of the ideas and methods of equation solving is enhanced, and they
develop richer and more abstract representations.
540 J. Li et al.
Through designing these equations from simple to complex, students gradually

improve their skills in equation solving and experience the ideas of transformation
and equivalence.
Teaching episode: Procedural variation to grasp the fluency of algorithm of
equation solving
1 1 1
Solve the equation:(x + 15) = − (x − 7)
5 2 3
Complex equation solutions:remove denominator, so
6(x + 15) = 15 − 10(x − 7)
−−−−−−→
remove brackets, so
change 6x + 90 = 15 − 10x + 70
transpose, so
6x + 10x = 15 + 70 − 90
simple equation, combine similar terms, so
16x = −5
divided by 16 from both sides, so
5
x=−
16
(Note that brackets in this exercise can be removed in the first step.)
Building on the fluency of algorithm, students are expected to understand why

these algorithms work and what mathematical thinking methods are embedded in
them. The experience of simple equation solving lays the groundwork for solving
complex equations.
Teacher: (1) Narrate the process of solving simple equations with one unknown.
(2) How do you check the unknown you have got is the solution to the equation?
For example, is x = 2 the solution to 3x + 4 = 5x + 7 or 3x + 4 = 5x?
This encourages students to synthesize what they have learned, to understand the
process of equation solving, and to form effective methods of learning. Strengthen-
ing the verbal representation of equation improves students’ mutual exchanges and
develops their mathematics thinking.
Teacher: We can now sum up the process of equation solving as follows: (1) remove the
denominator, (2) remove the bracket, (3) transpose terms, (4) combine the similar
terms, (5) change an equation into its simplest form with coefficient of x being 1.
Only by practice do we master the ideas of equation solving.
Next, the teacher gives students similar assignments and asks them to solve the equations
and to check the results independently or cooperatively so that they master the equation
solving process. Appropriate exercises help students consolidate the knowledge of equation.
Students gradually learn to master the equation solving process through explor-
ing varying problems. The variant training under the guidance of the equation solv-
ing helps students grasp the unchanging “routines” and “methods” for solving equa-
tions (i.e. one solution for multiple problems), obtains the dynamic and abstract rep-
resentations of the equation solving, and trains skills for the equation application.
To strengthen familiarity in equation solving, it can be deduced that students first

learn the general steps by solving various types of equations and then understand
the reasons and basics for equation solving. In addition, they are able to narrate the
steps of equation solving and then do sufficient variation practices until they can
do it automatically. To grasp the procedural knowledge, students not only practice
but also understand the ideas of algorithm, such as applying the equivalent equation
principle, understanding changing an equation from complex to simple, and so on.
It is known that students need to have a considerable number of guided variation
practices and show high proficiency in various types of equation solving in order to
consolidate the knowledge of equation solving.
This teaching episode makes it very clear that levels of variation are important.
The development of mathematical thinking is inseparable from formalizing complex
materials and the model generalization of factual materials. The levels of variation of
teaching algebra with variations should be set within students’ “recent development
zone,” in consideration of students’ levels, to motivate them make progress with
challenges, and to ensure that they can “jump for peaches” to develop rapidly and
orderly.
The Application of Equations
Constructing equations from word problems—as well as interpreting, rewriting,

and simplifying algebraic expressions—are named as key difficulties in the learn-
ing of algebra (Herscovics and Linchevski 1994; Linchevski and Herscovics 1996;
Sfard 1995). This session, it will show how to use problem variations, including
one problem multiple solutions and one problem multiple changes to help construct
equations from word problems.
Teaching episode: Problem variation (one problem multiple solutions) to learn
different presentations and strategies of constructing equations
Teacher: Here is a problem: “Can Xiaoming be caught up with?” (Ma 2002)
Xiaoming has to arrive at school which is 1000 m away from his home before 7:50 every
morning. One day, he went to school at the speed of 80 m/min. Five minutes later, his
father found he had left his Chinese textbook home, so he chased Xiaoming at the speed of
180 m/min, and caught up with him on the way to school.
(a) How much time did Xiaoming’s father spend to catch up with Xiaoming?
(b) What is the distance to the school from the place where Xiaoming was caught up with?
Teacher: (1) What is the relationship among time, speed, and distance? The journey prob-
lems include “meeting problems” and “what problems.” Please illustrate their in-
dividual “equivalent quantities in the process of journey” in these two types of
journey problems.
(2) On the solution of journey problems, we generally analyze the quantitative
relations between different unknowns such as time, speed, and distance as well as
the relation of the same parameter with the help of line segments, tables, and so
on, and then set the equations according to the quantitative relations of the same
unknowns. Is this correct? Are there any other ideas?
542 J. Li et al.
Our observations show that, the teacher helped students form correct procedures
of “problem solving with equation” until they could use them easily. The teacher
helped the students recall background knowledge, apply learned representations,
and design problem-solving strategies to lay the foundation for analyzing and solv-
ing questions, and reduce their difficulties in learning.
Teacher: We will try to analyze the quantitative relations in this problem with the following
methods. It is necessary to understand the problems clearly and analyze the quantitative
relationship appropriately in order to solve any kind of problem. As for this problem, we
can make use of line segments and tables to represent the thinking process.
Fig. 1 The relations between different distances
Segment AB is the distance from Xiaoming’s home to school, i.e., 1000 meters.
The quantitative relations (of the same unknown) in this problem:
(1) Time: (2) Distance:
Supposing the father catches up with Xiaoming after x minutes.
Table 1 The quantitative relation by using time x

Speed Time Distance Quantitative Equation
Relation
Xiaoming’s Section 1 80 5 80 × 5 Distance Xiaoming 80 × 5 + 80x = 180x

journey Section 2 80 x 80 × x covered = Distance
his father covered
His father’s journey 180 x ?
Supposing Xiaoming is y meters away from school when caught up with by his father.
Table 2 The quantitative relation by using distance y

Speed Time Distance Quantitative Equation
Relation
1000−y
Xiaoming’s Section 1 80 5 80 × 5 Time Xiaoming 80 −5
Journey Section 2 80 1000−y
−5 ? used in Section 2 = 1000−y
= 180
80
1000−y Time Father used
His father’s journey 180 180 1000 − y
From Tables 1 and 2, we know that, supposing time is unknown, the distance can be found
as the equal quantity and vice versa. Please finish Table 3.
Table 3 The quantitative relation by using distance x

Speed Time Distance Quantitative Relation Equation
Xiaoming’s Section 1 80 5 80 × 5 ? ?
Journey Section 2 80 ? ?
His father’s journey 180 ? x
The analysis of the procedural variation has allowed students to experience the
exploration process of “problem solving with equation.” In the process of model-
ing equations, the representations help students gradually deepen their understand-
ing of questions: the representation of language is helpful for understanding the
scenario; the pictorial representation is helpful for visualizing the relationship be-
tween different quantities; and the tabular representation is helpful for establishing
the quantitative relations. Such a comprehensive consideration and construction of
quantitative relationships help students completely grasp procedural operations of
“problem solving with equation”; teaching with variation through “exemplary prob-
lems,” applying all kinds of representations, help students learn how find out effec-
tive strategies for “quantitative relations” and to form various solving methods. The
“multiple solutions for one problem” helps the students understand the meaning of
the problem and the relationship among quantities from different angles, which, in
fact, enriches the multi-representations of equation problem-solving and develops
students’ algebraic thinking.
Appropriately expanding teaching topics helps students understand the quantita-
tive relations deeply, think about the steps and strategies of problem solving, learn
flexible representations of the problem space, experience operational representa-
tions of the problem-solving procedure, and master the structural modeling repre-
sentations. It helps the development of students’ algebraic thinking.
What’s more, to grasp the methods of solving application problems by solving
equations, experiencing the thoughts of algebraic modeling structure, and cultivat-
ing students’ ability of divergent thinking, the problems are changed; changing ei-
ther the conditions or the intended results. Students are encouraged either to ask
questions and then answer questions after removing the intended results or to create
different types of practical problems just as journey questions and efficiency ques-
tions based on an equation. Students also are asked to try mathematics experiments
to verify whether the theoretical analysis conforms to the practice. All the ideas
mentioned above help students to experience the invariant features within variant
process, namely algebraic structure model, and thus forms the students’ algebraic
thinking mode.
Teaching episode: Problem variation (one problem multiple changes) to cultivate
students’ explorative ability during constructing equations
[Variation 1]
Teacher: Now we can have a variation, please give your solution:
544 J. Li et al.
Xiaoming must arrive at school which is 1000 m away from home before 7:50 every morn-
ing. One day, he went to school at the speed of 80 m/min. 5 minutes later, his father found
he had left his Chinese textbook home, so he chased Xiaoming at a certain constant speed,
and it took him 4 minutes to catch up with Xiaoming.
(1) What was the speed of Xiaoming’s father while chasing Xiaoming?
(2) What was the distance to school when Xiaoming was caught up with?
According to our observation, based on the learning in the previous problems,
students solve the variation in a reversed way and understand the methods and ideas
of “problem solving with equation” from multiple perspectives. They especially
learn from checking the previous answer through the variation.
[Variation 2]
Teacher: Now we can have another variation, please give your solution:
Xiaoming must arrive at school from home before 7:50 every morning. One day, he went
to school at the speed of 80 m/min. 5 minutes later, his father found he had left his Chinese
textbook home, so he chased Xiaoming at the speed of 180 m/min, and caught up with him
at the place that is 280 m from the school.
Question: What is the distance from Xiaoming’s home to the school?
[Variation 3]
Teacher: Now you can make some variations of the example by yourselves or discuss within
your group and find out to what variations there are one solution and to what variations there
is no solution.
The original problem is as follows: Xiaoming has to get to the school which is 1000 m away
from his home before 7:50 every morning. One day, he went to school at the speed of 80
m/mins. 5 minutes later, his father found that Xiaoming had left his Chinese book home, so
he chased Xiaoming at the speed of 180 m/mins and caught up with Xiaoming on the way
to school.
Students pose other variations based on the problem with “whether there are so-
lutions to the variations of the problem” in mind, to analyze, to solve, and to check
the variations. The cyclic thinking of posing variations of the problem, solving the
variations, and checking the solutions to the variations, along with extensive sum-
mary and discussion, helps students understand the essence of the thinking about
“problem solving with equation” and acquire the abstract representation of “prob-
lem solving with equation” and develop their mathematical thinking ability.
Through the problem structural variants training (i.e. multiple variations for one
problem) to the multi-representations of problem solving with equation as a means
and purpose, the students appreciate the constant thinking mode about problem solv-
ing with equation. Based on the premise of a variety of the newly learned equation
skills, the students’ high-level algebraic thinking is being developed naturally.
Students’ problem-representing abilities and problem-solving strategies can be
trained by comprehensively using the ideas of equation and the skills of equation
solving to solve conventional or non-conventional problems and by using modeling
to solve contextual problems. They can experience the interaction between con-
structing and modeling algebraic concepts so as to improve their cognitive under-
standing of algebraic equations.
In teaching with variation to construct equations from word problems discussed
above, several steps are followed in teaching students how to solve contextual prob-
lems by equation solving. First, students are provided with background knowledge
of the problems, including common sense and relationship knowledge (formulas).

Second, students are taught how to analyze the relationship between variables, i.e.,
to illustrate the relationship between variables through language, diagrams, tables
and symbols. Third, students are required to write down an equation with the un-
knowns and the givens. They learn to analyze problems under guidance, to put the
unknowns and the givens together into the equation, and to formulate the equation
step by step. They are trained to develop an awareness of an equation for the pur-
pose of modeling. It is known that teaching and learning of “solving application
questions by equation method” needs to develop basic skills through practice with
conceptual variations and to develop algebraic thinking ability through practice with
procedural variations.
Thus, when applying this newly learned knowledge, students can use their men-
tal representations of the learned contents, i.e., the algebraic relation structure con-
structed, to specific problems. When solving problems, by using various represen-
tations from visual representations to abstract ones, students create appropriate rep-
resentations and strategies of solving the problems, and finally they express their
solutions to the problems by using symbolic representations.
This teaching episode makes it noticeably clear that the rational development
of teaching with variations can help set questions for students to inquire coopera-
tively, and this has significant positive effects. Understanding algebraic structures
and their application in the question variants or the construction of question variants
can help students understand the algebraic way of thinking on the whole and train
their mathematics thinking ability.
Discussion and Conclusion

As the analysis in this study shows variation, as a means, can be a powerful way
to help students develop mathematical thinking no matter how the content and the
problem are “changed” by the teacher. Furthermore, it also shows that, to better
conduct teaching algebra with variation, the lessons should be well structured and
the variation should be carefully chosen. Some observations are made based on the
analysis in the previous session, which will be discussed in the following sections.
Process of Teaching Algebra with Variation
Students can be helped when learning algebra, if teachers appropriately implement

teaching with variation by adopting different variation types and levels of represen-
tations according to algebraic learning goals, build on students’ existing knowledge
and ability, and use different teaching contents at different teaching phases.
It is believed that algebraic concepts are characterized by concrete operation,
abstract structure, and wide application and vice versa, thus, learning algebraic con-
cepts should focus on operational, structural, and applicable aspects. Therefore, in
546 J. Li et al.
the teaching of algebraic equations with variation, these steps should be followed.
First, the teaching is designed to promote students’ concept formation and assimi-
lation through variation, to get the initial formation of concept of equation, then to
help students understand the abstract nature of the equation concept, and to connect
the equation concept into the algebraic concept system, to clarify its position and
function in this system. Second, students are asked to apply the equation concept
to solve systemic varying problems, to make the equation concept an operational
object, to experience the structural function, and to enrich the equation concept
thoughts through the equation concept operation and the equation concept formation
reflection. Therefore, when teaching equation concept with variation, it is necessary
to construct the structure of equation concepts and develop the application ability of
system of equation concept so as to develop students’ algebraic thinking ability.
Operation of Teaching Algebra with Variation
It can be seen from the above analyses of classroom teaching episodes and inter-
views that the scientific and reasonable operation of teaching algebra with variations
lies in the proper grasp of the aspects of “orientation of variation, types of variation,
levels of variation, and variation exploration.”
Variation Orientation
In conducting teaching with variations, the teacher is required to recognize at what

level of thinking the students are and to know what representation features the stu-
dents have. This is the starting point for teaching. Second, the teacher is required
to analyze the “unchangeable” essential contents from what he or she will teach,
namely the principles of concept, the ways of thinking, and so on. Third, the teacher
is required to understand “unchangeable” ways of thinking of students, namely the
set mapping, variable relationship, program analysis and so on, in learning the “un-
changeable” essential contents. Finally, the teacher and the students are both re-
quired to use a reasonable setting of “changing of the unchangeable” in the shortest
possible time, to motivate the development of the students from “concrete represen-
tations” to “abstract representations” to efficiently achieve their learning goals, that
is, the grasp and reflection of the “unchangeable” essential contents, any negligence
of which will lead into a condition of “changing only for change.”
Types of Variation
To facilitate the introduction and understanding of the algebraic concepts, it is nec-

essary to provide students with various examples of different types (homogeneous
or heterogeneous) and verbal explanations, even with graphics and symbols to in-
crease students’ algebraic representations. The number of each type is subject to the
change of teaching objects, contents, and environment. These examples are set to
help students with their gradual classification, induction, abstraction, deep under-
standing, and application of variations. For example, the introduction to the concept
of equation needs some examples with slight differences in form, making it easy for
students to identify, induce, and summarize, rather than consuming too much of the
students’ energy in studying the examples. Indeed, the teacher should be flexible
in deciding the quantity and difficulty of the variation examples, depending on the
students’ actual level of understanding so as to encourage them to achieve relatively
abstract algebraic representations in the near future.
Levels of Variation
In order for students to form a three-dimensional concept network with a relatively

high level of abstraction and to foster a flexible application capacity, it is necessary
to set question strings of variation at different levels to enable them to gradually
experience the invariable modes of thinking in the development of algebraic knowl-
edge, that is, the formation of the abstract internal representations. Here the levels
should be set with gradient, inspiration, challenge, and control, that is, the arrange-
ment of questions is a matter of degree around “invariable mode of thinking.” For
example, in the teaching of equation solving with variations, the role of the un-
changeable “idea of equivalence” is shown in the changing of equations through
the gradual increase of equation complexity. For example, in equation application
teaching, question strings should be set intentionally in order to train the students in
equation lining and to motivate students to see the intrinsic link between different
types of knowledge, to experience the methods of equation lining, and to be skilled
in representation strategies of equation lining. The setting of questions of variation
should follow as far as possible the principle of “no matter how the conditions and
conclusions change, the searching for the equivalents will never change,” the main
purpose of which is to help students learn equation lining. Surely, in order to cul-
tivate students’ convergent thinking and divergent thinking, it is necessary to raise
their thinking on “how to change to solve the problem.” It will be more worthwhile
when students are trained into advanced and critical algebraic thinking.
Variation Exploration
The variant training of the traditional algebra “two basics” can be transformed into
inquiry-based learning under guidance, which requires teachers to specifically re-
search the approach of how to present algebra knowledge in an evolving way and
how to represent the problems related to them. When students have a certain ba-
sis of the “two basics,” teachers could try to “melt the exploration in the variant.”
For example, under the guidance of teachers, students may explore the direction
of knowledge transferring, the knowledge points after question transformation, the
method after information transformed, the problem after the replacement of the orig-
inal conditions or intended results, the different kinds of quantitative relationships
548 J. Li et al.
in a problem, and the mutual links between the different problems at different lev-
els, and so on. Therefore, when the types and levels of variant are designed, the
problems should be instructive, thinking should be targeted, span should have conti-
nuity, content should be a spiral, etc. All of these are conducive to student learning in
inquiry-based learning. For example, the above equations to learn, through the vari-
ant inquiry, with multi-angle and multi-level, displays the learning process of the
introduction and understanding, consolidating and applying of the equation, gets
the objective of “student learn to think” as the key point. Using both the breadth and
depth of knowledge, the teacher and students explore the knowledge of the equa-
tion and the related problem-solving and representations. Such a direct guiding of
inquiry-based learning should proceed in the variant and focus on the algebraic rep-
resentations, which would promote students to develop the mathematical thinking
and to stimulate student enthusiasm. The benefit for this mathematics learning at
this age is obvious.
This teaching method, based on variation of the teacher’s external knowledge
and questions to achieve students’ internal multi-representations, can integrate the
teacher’s external guidance and the students’ internal subjectivity so that teaching
and learning go in harmony.
Final Comments
In China, teachers usually design some conceptual variations or procedural varia-

tions while teaching algebraic equations. On the one hand, these variations can in-
spire students’ motivations to learn algebraic equations, understand algebraic equa-
tions from multiple perspectives and achieve the fluency of solving algebraic equa-
tions. On the other hand, students can obtain the experiences in algebraic equations
activity, master the thinking methods of algebraic equations, and enhance their prob-
lem solving abilities. Therefore, through appropriate teaching with variation, it is
ensured that the training for the two basics in mathematics education in China is not
drill training; on the contrary, teaching with variation can promote the development
of students’ algebraic thinking (Zhang and Song 2004)
As it is shown in this study, there is great potential for teaching with variation;
however, there are some unanswered questions. For example, as suggested by a
recent study (Mok et al. 2008), if the type of engagement the teacher created in the
lesson using problem variations leads to missed opportunities for fostering students’
higher-order thinking skills, how do teachers balance the development of basic skills
and higher-order thinking skills when conducting teaching with variation? More
empirical studies are expected to confirm and support the promise of teaching with
variation.
Acknowledgements The authors would like to thank Dr. Jinfa Cai for his guidance and help on
this study.
The team for this paper consists of members: Song Naiqing, Yu Bo, Li Jing, Peng Aihui, Yang
Bingjun, Zhou Jiazu, Li Ling, Zhang Kun, Xu Zhaoyang, Shen Lin, Zhang Qinqiong, Chen Yan,
Wang Xue, Xu Keqiang, Shen Nanshan, Li Guoqiang, Fu Daoling, Tang Hengjun.
Appendix: English Translation of a Teaching Design and Its

Reflection after the Lesson: Teaching Equation Based on
Problem Variation
Analysis of the Teaching Material “Can Xiaoming Be Caught Up With” is se-
lected from the Mathematics Textbook for Grade Seven published by Beijing Nor-
mal University in 2002.
Materials with a practical example of “Can Xiaoming Be Caught Up With” build-
ing a problem situation should inspire students to explore problem-solving strate-
gies, and to experience the process of “translating the practical problems into the
mathematical problems” and the mutual transformation between the verbal repre-
sentation, the symbolic representation, and the graphical representation. This teach-
ing course is designed from the scenarios in which the students should be interested
and give the students access to information through the representations, such as
drawing line segments and diagrams. Using the problem structure variant (that is,
multiple variations for one problem, by replacing the conditions and conclusions of
the original problem, forming its tributaries), the students look for different equal-
ity relations from different angles, thus, students should initially find “mathematics
modeling” as a means, and better develop students mathematical representing and
thinking ability.
Key points: to enable students to find the known and the unknown quantities and
relationship between them.
Difficult points: to analyze the quantitative relations between complex problems
using line segments and other methods.
Analysis of the Students Students have learned to solve simple application ques-
tions using line segments in the elementary school. In previous lessons, they have
learned many kinds of application questions using a linear equation with one un-
known. However, exercise results show that they do not understand the essence of a
linear equation with one unknown to solve application questions. Application ques-
tions are some of the most difficult questions, and students may become tired of
them. This is not because of lack of interesting scenarios for practical questions, but
because of the fact that students are not led into a deep, orderly and multi-angled
thinking of “the discovery of the quantitative relationship.” Creating a scenario for
the question can help students think originally, but if too much emphasis is put on
the scenario and the problems, then the theme of mathematics teaching activities
(that is, to help students learn to think in mathematics) will be weakened. The ar-
rangements of teaching, as a result, should foreground the target of “learning to
think” in solving application questions by linear equation with one unknown.
Principles for Designing

1. Proceeding from the practical scenario of the question, students will gradually
enter into the problem-solving process in a relaxed environment. Simultaneously,
variation questions and activities should be designed according to students’ cur-
rent level of competence. From easy to difficult, from simple to complicated,
550 J. Li et al.
students can understand different approaches to the questions and their pros and
cons, and know the importance of problem-solving strategies. Students study in
the free environment, each adopting different ways of thinking in the class. Thus,
their ability to think is cultivated in this process.
2. After acquiring the approaches of solving application questions by linear equa-
tions with one unknown in classroom activities, students begin to ask questions in
the discussion section and try to answer them under the guidance of the teacher.
This process can help students develop their thinking, find a chance to innovate
and perform freely, and fully enjoy the sense of success after exploration.
Teaching Objectives
1. Through learning to solve application questions by linear equations, students can
perceive the role of mathematics in daily life.
2. Through using line segments and other methods to search the quantitative re-
lationships among complicated questions, students can enhance their algebraic
thinking ability.
3. Through analyzing and studying different equal relationships, students can expe-
rience the diversities of problem-solving strategies and develop their innovative
ability.
4. Students will reflect themselves and learn from others during cooperation and
communication.
Teaching Process
1. Creating the scenario of questions
Teacher: Class, have you ever seen someone “being caught up with”? Who
can give an example?
Students may give many examples around them, about human beings or ob-
jects.
Teacher: In the case of chasing, is the faster chasing the slower? Or vice versa?
Students will know it is the faster that chases the slower, a simple reasoning
in their life.
Teacher: Who have ever seen a cat or a dog chasing a rat?
Students may laugh and say, “A dog tries to catch rats. That is to poke one’s
nose into other’s business.” The class atmosphere is thus activated.
Teacher: We do not care about “dog chasing rats.” Today, we will discuss
Xiaoming’s father chasing Xiaoming (Show the story).
Xiaoming must arrive at school at 7:50 every morning. One day, he went to
school at the speed of 80 m/min. 5 minutes later, his father found he had left his
Chinese textbook home, so he chased Xiaoming at the speed of 180 m/min, and
caught up with him on the way to school.
Please answer the questions.
2. Organizing students’ activities
Question 1: If Xiaoming’s father caught up with Xiaoming within 4 minutes,
how many meters had Xiaoming and his father covered respectively?
With the formula “distance = time × speed” learned in the primary school,
students can give the answer that they both covered 720 meters. The teacher can
guide students to draw a line segment and fill out the following table:
V (min/m) T (min) D (m)
Xiaoming Section 1 80 5 400

Section 2 80 4 320
His father 180 4 720
Teacher: Let’s look at the variables in the figure and the table, and find how
many of Xiaoming’s variables (speed, time and distant) correspond with his fa-
ther’s.
After observation, students may find out: (1) the distance Xiaoming covered
= the distance his father covered (when Xiaoming was caught up with); (2) the
time Xiaoming spent in his second section = the time his father spent.
Teacher: what should the above equations be when Xiaoming’s father walked
3 minutes?
After discussion in groups, representatives from the groups are asked to dis-
cuss the following: When the father walked 3 minutes, he had covered 540 me-
ters, and Xiaoming covered 640 meters. Variables in equation (1) are not equal,
which means Xiaoming’s father didn’t catch up with him, but equation (2) is still
tenable. The teacher then attaches a condition “When Xiaoming was caught up
with” to equation (1).
Teacher: Why equation (1) is tenable when the father walked 4 minutes? How
can it be checked with the equation lining methods we have learnt? So we have
question 2.
Question 2: If the father caught up with Xiaoming on the way, how much time
did he spend?
Students may discuss in groups and answer the question with the help of the
figure and the table in question (1).
(Analysis: The original question was changed in the textbook so the students
could answer with relevant knowledge acquired in primary school. In this way,
the students can review the “distance = time × speed” equation and understand
the relations among all variables. This specialization helps students understand
the essence of problem-solving and enables them to experience the process of ob-
servation, supposition, and verification, which can inspire their interest in ques-
tion exploration. This benefits the students’ study in mathematics to a great extent
and opens a large space for the students and teachers to innovate. The configura-
552 J. Li et al.
tion of thinking based on the obtained knowledge and experiences is the thinking
process of “mathematicization.”)
Following the example of question 1, the students have little difficulty in
finding out the unknown variables with figures and tables, hypothesizing the
unknown, and lining the equation according to the equivalent relation in equa-
tion (1):
Supposing the father catches up with Xiaoming with X minutes.

Section 2 80 x 80x
Father 180 x 180x
400 + 80x = 180x

x=4
Teacher: If we have verified that the father caught up with Xiaoming in 4

minutes, then how many meters had the father covered when he caught up with
Xiaoming?
The teacher asks a lagged-behind student to answer the question, and the an-
swer is: 4 × 180 = 720 meters.
Teacher: Can we hypothesize the father’s distance as x and directly build the
equation? Let’s check whether it is 720 meters. We will take this as question 3.
Question 3: If the father caught up with Xiaoming on the way, then how many
meters had the father covered?
Students are required to discuss in groups, and the teacher helps those lagged-
behind students analyze the problem with the help of figures and tables. The
teacher helps them to fill out the table, find out the equivalent relation of equation
(2) and build the equation.
Supposing the father had walked X meters when he caught up with Xiaoming.
Xiao Section 1 80 5 400

ming Section 2 80 ? x − 400
Father 180 ? x
X − 400 X
=
80 180
X = 720
Teacher: We can immediately calculate Xiaoming’s second section distance

= 720 − 400 = 320 m or 80 × 4 = 320 m. Then, can we hypothesize his second
section distance as X m and build the equation? This is question 4.
Question 4: If the father caught up with Xiaoming, what is Xiaoming’s second
section distance to school?
Students are asked to discuss in groups, and the teacher guides lagged-behind
students and asks representatives in each group to answer the questions.
Supposing Xiaoming covered X meters in his second section when his father
caught up with him.

Section 2 80 ? x
His father 180 ? 400 + x
X 400 + X
=
80 180
Teacher: If you are told that the distance from Xiaoming’s home to the school
is 1000 meters, it will be very easy to figure out 1000 − (400 + 320) = 280 m.
But can we answer it directly by building equation? This is question 5.
Question 5: If the distance from Xiaoming’s home to the school is 1000 me-
ters, what is the distance from the place where the father caught up with Xiaom-
ing to the school?
Students build the equations alone and work out the answer. It is 280 meters.
554 J. Li et al.
The teacher asks the lagged-behind students to write the answer on the black-
board and emphasizes the writing forms.
(Analysis: The problem string is helpful for the students to carry out a series
of conceptual activities such as observation, experimentation, and verification; to
understand the essence of the application question to be solved by equation; and
to experience the superiority of such representations as line segments and charts.
This not only stimulates the students’ desire to explore, but also expands the
students’ thinking span to a reasonable extent. Therefore, students are thinking
of mathematics in an exciting class atmosphere, embodying the fundamental and
developmental idea of Mathematics Curriculum Standards issued by Ministry of
Education of China in 2001.)
3. Broadening thinking in the lesson
Question 6: Supposing Xiaoming’s father caught up with him right at the
school gate. How much time did he take? What was his speed? In what cir-
cumstances couldn’t he catch up with Xiaoming?
(Question 6 echoes question 1, but it is more flexible. Here, questions and
students’ interests reach a climax. Students feel questions emerge endlessly, and
they can experience the mysteries of mathematics. They further understand the
close relationship between mathematics and life.)
They quickly answer the first two questions. As for the last one, the teacher
can ask representatives of different groups to demonstrate their research results,
giving them appropriate guidance, offering appropriate recognition and praise.
Finally the teacher makes a summarized statement.
Question 7: See the “Discussion section” in the textbook.
Students of grade seven from Yuhong School traveled to the town’s outskirts.
Students of class one formed the first team and walked at the speed of 4 km/h,
and those of class two formed the second team and walked at the speed of 6 km/h.
The second team set off one hour after the first team had started and at the same
time sent a liaison riding a bicycle at the speed of 12 km/h back and forth between
the two teams.
Please pose questions based on the above facts and try to answer them.
The teacher can ask the students two or three questions according to their per-
formance and the time left in the class. As for those lagged-behind students, the
teacher can help them with some tips to imitate the previously answered ques-
tions and tell them how to analyze the hidden meaning of the questions. Finally,
the teacher can ask the representatives to write their answers on the blackboard
and may also ask them to finish the rest questions after class.
Finally, the students are encouraged to change the conditions or the intended
results of the above question, and to solve it.
(Analysis: For the discussion of complex questions, the teacher should take
students’ current knowledge level into consideration but should not give too dif-
ficult or too many questions, except for some interesting and challenging ones. It
is always the teacher’s responsibility to ensure students to constantly experience
success in their mathematics study according to the idea of “public mathemat-
ics.”)
Reflection after the Lesson Comparatively speaking, for the first-year students
in junior high schools, this lesson contains relatively too many “thinking contents.”
When designing the teaching process, the teacher should set out from the two dimen-
sions of the depth and breadth, and take in full consideration the organization style
of this lesson and the students’ current level of competence. The teacher should set
question variations in connection with practice, which is good for students’ think-
ing development and is in accordance with the goals and requirements of the Math-
ematics Curriculum Standards. Teachers can disassemble the thinking contents in
procedural variation so that students can march upwards gradually along the steps
of thinking, experiencing the inner beauty and application value of mathematics.
Efforts should also be made to encourage students to make essential preparations
for thinking prior to students’ cooperating, exchanging, and discussing. The issues
for discussion should be based on the purpose of understanding the main objectives
of the lesson proper. Otherwise, purposeless discussions will result in wasting of
students’ time, and they make to progress. This lesson aims to start out with the
basis of students’ knowledge, focus on their future development, with emphasis on
practice on students’ mathematical thinking and on the training strategies for prob-
lem solving. It shows the effective use of a variety of methods to understand the
meaning of the questions and experience the fun and significance of mathematics
learning. This lesson broadens students’ thinking and well reflects the connotation
of all-around education.
References
Ausubel, D. P. (1968). Educational Psychology: A Cognitive View. New York: Holt, Rinehart &
Winston.
Behr, M., Erlwanger, S., & Nichols, E. (1980). How children view the equal sign. Mathematics
Biggs, J. B., & Watkins, D. A. (1996). The Chinese learner in retrospect. In D. A. Watkins & J. G.
Biggs (Eds.), The Chinese Learner: Cultural, Psychological, and Contextual Influences (pp.
269–285). Hong Kong/Melbourne: Comparative Education Research Centre, the University of
Hong Kong/Australian Council for Education Research.
Cai, J., & Nie, B. (2007). Problem solving in Chinese mathematics education: Research and prac-
tice. ZDM, 39, 459–475.
Cai, J. et al. (2005). The developments of students’ algebraic thinking in earlier grades: A cross-
cultural comparative perspective. ZDM, 37(1), 5–15.
Chen, Z., & Song, N. (1996). Mathematics instruction experiment for improving efficiency in class.
Journal of Mathematics Bulletin, 8, 21–23. In Chinese.
Gu, L. (1994). Theory of Teaching Experiment: The Methodology and Teaching Principle of Qinpu.
Beijing: Educational Science Press. In Chinese.
Gu, M. (1999). The Grand Educational Dictionary. Shanghai: Shanghai Education Publishing
House.
Gu, L., Huang, R., & Marton, F. (2004). Teaching with variation: A Chinese way of promoting
effective mathematics learning. In L. Fan, N. Wong, J. Cai, & S. Li (Eds.), How Chinese Learn
Mathematics: Perspectives from Insiders. New Jersey: World Scientific.
tional Studies in Mathematics, 27(1), 59–78.
556 J. Li et al.
Huang, R., & Leung, F. K. S. (2004). Cracking the paradox of the Chinese learners Looking into
the mathematics classrooms in Hong Kong and Shanghai. In L. Fan, N. Wong, J. Cai, & S.
Li (Eds.), How Chinese Learn Mathematics: Perspectives from Insiders. New Jersey: World
Scientific.
Husen, T. (1967). International Study of Achievement in Mathematics: A Comparison of Twelve
Countries, Vols. I, II. New York: Wiley.
of Research on Mathematics Teaching and Learning. Reston, VA: National Council of Teachers
of Mathematics.
Lapointe, A. E., Mead, N. A., & Phillips, G. W. (1989). A World of Differences: An International
Assessment of Mathematics and Science. Princeton, NJ: Educational Testing Service.
Linchevski, L., & Herscovics, N. (1996). Crossing the cognitive gap between arithmetic and alge-
bra: Operating on the unknown in the context of equations. Educational Studies in Mathemat-
ics, 30, 39–65.
Ma, F. (2002). Curriculum Standard Experimental Textbook for Compulsory Education: Mathe-
matics (for the Second Semester of Grade Seven). Beijing: Beijing Normal University Press.
Marton, F., Runesson, U., & Tsui, A. B. M. (2004). The space of learning. In F. Marton, A. B. M.
Tsui, Chik, P. Y. Ko, M. L. Lo, & I. A. C. Mok (Eds.), Classroom Discourse and the Space of
Learning (pp. 3–40). New Jersey: Lawrence Erlbaum.
Mok, I. A. C., Cai, J., & Fung, A. F. (2008). Missing learning opportunities in classroom in-
struction: Evidence from an analysis of a well-structured lesson on comparing fractions. The
Mathematics Educator, 11, 111–126.
Park, K. (2006). Mathematics lessons in Korea: Teaching with systematic variation. Tsukuba Jour-
nal of Educational Study in Mathematics, 25, 151–167.
Pirie, S., & Martin, L. (1997). The equation, the whole equation and nothing but the equation! One
approach to the teaching of linear equations. Educational Studies in Mathematics, 34, 159–181.
Sadovsky, P., & Sessa, C. (2005). The didactic interaction with the procedures of peers in the
transition from arithmetic to algebra. Educational Studies in Mathematics, 59, 85–112.
tives. Journal of Mathematical Behavior, 14, 15–39.
Stevenson, H. W., & Lee, S. (1990). Contexts of Achievement: A Study of American, Chinese, and
Japanese Children. Chicago: University of Chicago Press.
Sun, X. (2009). “Problem variations” in the Chinese text: Comparing the variation in problems in
examples of the division of fractions in American and Chinese mathematics textbook. In M.
Tzekaki, M. Kaldrimidou, & H. Sakonidis (Eds.), Proceedings of the 34th Conference of the
International Group for the Psychology of Mathematics Education.
Vlassis, J. (2002). The balance model: Hindrance or support for the solving of linear equations
with one unknown. Educational Studies in Mathematics, 49, 341–359.
Watson, A., & Mason, J. (2006). Seeing an exercise as a single mathematical object: Using varia-
tion to structure sense-making. Mathematical Thinking and Learning, 8(2), 91–111.
Zhang, D., & Song, N. (2004). Introduction to Mathematics Education. Beijing: Higher Education
Press.
Commentary on Part III
John Mason
Introduction
Most research in mathematics education reports on what is or appears to be the

case currently: obstacles, weaknesses and failings, for and by learners, teachers, and
curricula. Some intervention studies look locally at what is the case under excep-
tional circumstances and hence indicate what perhaps could be the case generally,
but inevitably they build on a legacy of socio-cultural-historical practices and expe-
riences inherent in the specific situation. Following the lead presented by Maslow
(1971), whose hierarchy of human needs provides a context, I find myself attracted
most by what could be the case globally. What could the learning of algebra be like
if teachers understood profoundly and appreciated deeply both school algebra and
its pedagogy and didactics,1 and if teaching were carried out consistently over a
sustained school experience of several years, following an enlightened curriculum?
What sort of progress have we made in this direction in the last 25 years, and where
might we be headed?
Stimulated by the chapters in this part, I make use of a framework developed
by Bennett (1966, 1970) which provides a structure for discussing aspects of the
activity of teaching a mathematical topic, or in the case of algebra, of a way of
thinking, being and acting mathematically; what school algebra is and could be;
and what is and could be researched. I use the framework three times, to structure
discussion of the current state, the possible state, and the activity of going from
where we are now to where most of would like to be.
1I use pedagogy to refer to acts of teaching that apply to many different topics, and didactics to
refer to acts of teaching specific to a particular topic.
J. Mason ()
Open University, Milton Keynes, UK
J. Mason
University of Oxford, Oxford, UK

558 J. Mason
Systematics: Structure of Activity

In order to identify what some activity might achieve, it is of course vital to know
where one is currently. As Bennett (1966, 1970) pointed out, activity is usefully
conceived as a four-termed system.2 In addition to the current and goal states, there
are both tasks or actions to undertake and resources upon which to call. For activity
to succeed there has to be a balance between all four terms: resources need to be ad-
equate for the tasks required, the tasks required need to be suitable for achieving or
approaching the goal from the given state, the tasks and resources need to be acces-
sible to the individuals or group concerned. In the language of Gibson (see Greeno
1994) the affordances and constraints of both resources and tasks undertaken, and
the attunements of the current state all need to be compatible with and appropriate to
the desired goal(s). This structure for activity applies to learners working on tasks,
as well as to teachers teaching mathematics and to researcher-educators trying to
influence policy, curriculum, pedagogy and didactics. Since I propose to use this
structure three times, it is worth elaborating a little.
Current state (what is the current situation): Descriptions of what is currently the
case not only establish a baseline, but reveal what the observer is sensitised or
attuned to notice and so what is valued, ignored or eschewed in relation to the goal
state(s).
Goal State (the aims, purposes, intentions, desires): As is well known from the
Vygotskian-Davydovian ‘Activity Theory’ perspective, goals may or may not be
explicit, or even appreciated, even by those engaged in the activity. What observers
interpret as the assumed goal(s) may not always be compatible with the lived ex-
perience of those observed.
Motivation (axis between current state and goal state(s)): The relation between cur-
rent and goal states provides direction and motive in relation both to cognition and
affect.
Tasks or Actions: Tasks are what initiate or subdivide the sequence of acts or actions
that themselves constitute the flow and development of activity. They provoke ac-
tions carried out which constitute the observable activity. Tasks may be self-set,
imposed or made available.
Resources: Resources are what can be called upon in carrying out tasks, such as
your own powers, dispositions and experience, the presence of colleagues, and
established and institutionalised ways of working.
2 Current Vygotskian-based activity theory is based on a triangle of three terms, elaborated into 6
or more. In this section I make covert use of Bennett’s five-fold structure of potential, and later, the
six-fold structure of the present moment.
Commentary on Part III 559
Operational Means (axis between resources and tasks): Tasks are merely tasks; re-
sources have to be called upon in order to be useful. The combination of tasks and
resources form an operational axis, but just as with the vertical axis, it is necessary
that the resources be adequate for the tasks and that the tasks make effective use of
resources.
What Is Algebra?
Algebra is seen by many as ‘arithmetic with letters’, and there is a long histori-
cal precedent in textbooks stretching back to the 14th century. As such it depends
upon experience and facility with arithmetic calculations, and it provides students
with skills to carry out algebraic manipulations, many of which parallel arithmetic
computations. At the very least, school algebra is a collection of mathematical
practices and procedures to be internalised and integrated into learners’ function-
ing. At the very most in its traditional form it affords a glimpse of a powerful
tool for modelling and thus resolving problems. Usually however the ‘problems’
it is used to solve are obviously artificial (Gerofsky 1996; Verschaffel et al. 2000;
Mason 2001). Thus the essence of traditional school algebra is of a hurdle to be
overcome in order to gain qualifications to do other things. Some learners succeed
by subordinating their own desires to mastering the procedures, and some manage
to see through the particular manipulations to how symbols can be used to model
situations and how the manipulations can be used to then resolve problems in those
situations. These few manage to adapt and accommodate aspects of algebraic think-
ing.
Historically, algebra emerged as a device to deal with the as-yet-unknown. Mary
Boole captured it nicely:
. . . we have been dealing logically with all the facts we knew about this problem, except
the most important fact of all, the fact of our own ignorance. Let us include that among the
facts we have to be logical about, and see where we get to then. In this problem, besides
the numbers which we do know, there is one which we want to know. Instead of guessing
whether we are to call it nine, or seven, or a hundred and twenty, or a thousand and fifty, let
us agree to call it x, and let us always remember that x stands for Unknown. Let us write
x in among all our other numbers, and deal logically with it according to exactly the same
laws as we deal with six, or nine, or a hundred or a thousand. (Boole 1231 Tahta pp. 55–56)
. . . the essence of algebra . . . consists in preserving a constant, reverent, and conscientious
awareness of our own ignorance. (Mary Boole quoted in Tahta 1972, p. 56)
This seems to capture the use of a letter as symbol made by Diophantus, al

Khwarizmi, Viéte, Cardano and others of the time. For a long time the unknown
was referred to as shai (Arabic), res and causa (Latin) and then cosa (Italian) and
coss (German) all meaning ‘thing’ and algebra became known as the ‘cossic art’.
Modern algebra as the study of structure arises from abstracting through recog-
nising relationships as instantiations of properties and expressing those particular
relationships as generalities. With the hindsight of experience of axiomatisation in
modern mathematics an alternative interpretation of algebra as the symbolisation of
560 J. Mason
the unknown is available: all uses of letters as symbols in an algebraic context are
expressions of generality: to say “let x be . . . ” can be seen both as an act of hy-
parxis3 and a recognition, acknowledgement or expression of generality. There is a
hidden or ellipsed ‘all’ in letting x be the number of something or others, because
it could be ‘anything’: any or all possible values depending on what is possible due
to the constraints being imposed. All it takes is an ever so slight shift to letting x
be any or all numbers (values etc.), and then imposing constraints on that symbol
through expressions. Thus 3x + 1 expresses ‘any or all numbers one more than a
multiple of 3’, while ‘3x + 1 = some expression’ imposes a constraint arising from
the context under consideration. Note that equality has to be seen as a statement of
relationship not an instruction to calculate.
Newton helped redirect the development of algebra by shifting attention from
using symbols to stand for the unknown in word problems, a process which he con-
sidered to be essentially trivial, to the study of the equations that arise and how they
might be solved (Newton 1707; see Whiteside 1972, pp. 129–157). In other words,
he shifted attention from setting up a model using algebra, to structure and the search
for effective procedures. Other authors around his time were less convinced about
how easy it is to set up the equations (Ward 1706).
An alternative to the traditional perception, referred to here as visionary algebra
teaching is to see school algebra as a manipulable language for expressing rela-
tionships and constraints, on both numbers and actions on numbers (and later other
objects). It depends on and makes use of children’s evident powers to deal with gen-
erality as they learn language (which is inherently general). Thus it could be fed by
explicit practices from the beginning of formal school if not before, in which learn-
ers are prompted to recognise relationships, to perceive these as potential properties,
and to express them as conjectures to be tested and ultimately justified. When done
more systematically, it is what we call modelling. Thus school algebra could con-
tribute to children’s sense of having a way to deal with numerical relationships and
puzzling situations involving quantity, as well as demonstrating one way of express-
ing and testing conjectures about general properties, setting them up to challenge the
over-and sweeping generalisations encountered in the media. The next section uses
the four term structure of activity to analyse these two extremes: traditional and
envisioned teaching of algebra.
What Is and What Could Be: Teaching Algebra as an Activity
Appreciating the current state of play certainly justifies research as descriptions of

what is now the case, but for teaching and learning algebra we have been accumu-
lating such studies for more than 50 years. Merely combining this with intention
and desire to ‘do better than previous generations’ is still not sufficient, otherwise
algebra teaching would have changed long ago. There is something else blocking
3 hyparxis means ‘coming into being’.

development, some obstacles that have so far proved insurmountable, so I use the
four-fold structure to analyse activity in traditional and visionary algebra lessons.
Traditional Algebra Teaching
Current state: Learners come to class with six or more years of immersion in arith-
metic, seen as getting answers to calculations using the four operations on whole
numbers, then rationals,4 then decimals. The research literature is full of reports of
what learners do and do not do, and even what prospective teachers do and do not do.
Unfortunately it is often cast in terms of what the researcher concludes the learners
‘can’ and ‘cannot’ do, despite the adage that absence of evidence is not evidence of
absence. Just because someone does not do something, it does not follow that they
cannot, even in similar circumstances. It depends to a large extent on awareness: on
the actions that come to mind.
Goal State:
As Smith puts it,
one of the most common and fundamental goals [for algebra teaching] is helping students
move beyond an arithmetic approach to a more generalized approach to understanding re-
lationships. (Carpenter et al. 2003; Carraher and Schliemann 2007)
The desired state concerning algebra has recently become wrapped in the language
of problem solving and communication, which over-generalise and so lose contact
with the specifically mathematical feature of manipulability. However it is probably
universally agreed, certainly implicitly in the chapters of this section, that facility
with the use of symbols (letters) to express relationships (to model) and thereby to
resolve problems, is desirable if not essential for full participation in society and
use of the power of mathematics. Unfortunately for most learners this goal is not
achieved. Algebra continues to be a watershed in mathematics for many learners, as
noted by Cusi, Malara and Navarro and even for prospective teachers as noted by
Ellerton and Clements.
Motivation: Neither passing examinations nor claims that algebra is needed ‘later’
or is ‘good for you’ are adequate motivation in the 21st century. Whereas in previous
generations discipline and learner subordination to the institution were sufficient to
immerse learners in activity, that is no longer the case. This opens the vexed problem
of what does motivate learners, which is vigorously disputed (see later section).
Tasks: Ever since the 14th century algebra has most often been presented as arith-
metic with letters, dominated by procedures for manipulating symbols, despite on-
going research evidence of its ineffectiveness, and this is still the case in many text-
books and classrooms today. It could be a symptom of the underlying obstacle, that
4 Although usually called fractions, the arithmetic is on rational numbers; fractions are operators,
and fractions with the same effect (equivalent fractions) are identified with their effect on the unit
to produce rational numbers.
562 J. Mason
learners respond to their powers being made use of, rather than being treated as
clerks whose task is to reproduce otherwise meaningless sequences of symbols.
Resources: Learners’ experience of arithmetic could be a major resource, but often

turns into a hindrance, a theme to be developed in a later section. Yet all learners
who get to school have displayed the requisite powers to make use of algebra as a
language for expressing relationships.
Operational Means: Routine tasks promote dullness and routine as what mathe-
matics is about, rather than mathematics being seen as an expressive, constructive,
creative, exploratory domain of enquiry and justification. Resources called upon in-
clude compliant learners (a declining commodity), traditional texts, an impoverished
vision of what algebra is or could be for learners, and concomitantly an emphasis
on demonstration and practice. It is necessary to draw upon resources in the form of
constructively mathematical ways of working if the activity arising from engaging
in tasks is going to be transformed into constructive actions by and for learners.
The curricular extraction from context of ‘skills’ which has driven the teach-
ing of arithmetic into rehearsal of arithmetic procedures has similarly driven the
teaching of algebra into symbol manipulation procedures, losing contact with what
algebra could be about and what it is for ultimately. For example, Smith contrasts
approaches to the equality sign used in the TIMSS second study videos in Japan
with observations in classrooms in the USA. The former emphasise the relatedness
and expression of generality, the latter the procedures for manipulating objects. She
finds that learners in Japan respond more effectively to probes of their algebraic
facility than learners in the USA. There could be an underlying relationship!
Balance: That the resources called upon and the tasks offered are out of balance and
inadequate is evident from the literature, some of which is referred to in the papers
in this section. Lack of fluency with basic arithmetic may be accompanied by loss
of self confidence (Dweck 2000) and lack of interest (motivation, both affective
and cognitive aspects) in sustaining sufficient concentration and investing sufficient
energy to be successful.
Ellerton and Clements quote and agree with Kieran (2007) who emphasized that
unless students
come to realize that algebra is an arena of sense-making and that they can arrive at rules that
will permit them to obtain the same results as their teacher or classmates, they will never be
able to control their algebraic work (Kieran 2007, p. 732).
It is clear that the stretch from the current state of learners to the desired state is far
too great for the tasks and the resources employed to achieve, so it is no wonder that
algebra remains an unpopular and difficult ‘topic’. Available resources such as the
natural powers learners have displayed in getting to school, learning language etc.
are often under used or even usurped by text and teacher.
Envisioned Algebra Teaching
Algebra need not be seen as a ‘topic’ but rather as a language for communication
with yourself and with others, a means of expression of what is imagined, which is
also readily manipulable.
Current state: If learners experienced arithmetic as the study of actions on ob-
jects (counting, adjoining, removing, scaling, partitioning collections, then adding
to, subtracting from, replicating and scaling and partitioning or quotitioning cor-
responding numbers; fractions as operators), then there would be something on
which algebra could build. If arithmetical operations of adding, subtracting, mul-
tiplying and dividing were studied as actions, so that relationships such as inverse
operations, the role of zero and one, commutativity, associativity and distributivity
emerged and were expressed and commented upon in the particular and gradually
as general properties, then algebra would emerge perfectly naturally.
Goal State: Algebra could become a manipulable language for modelling situations,
perceived both in the material world and in the symbolic world of mathematics.
As one example, Boaler (2002) reported that learners who experienced a way of
working involving exploration that involved practice, augmented by exam-oriented
practice when needed, recognised much more opportunity to think mathematically
or to use their mathematical thinking outside of the classroom than others who tried
to memorise procedures and practised them extensively.
Motivation: By amplifying the natural human desire to generalise, to incorporate
multiple situations under one heading, to characterise and classify, in other words,
to perceive properties as being instantiated rather than each situation as unique,
learners’ could experience the pleasure of using and developing their powers (Ma-
son 2008). There are several contrasting schools of thought about motivating alge-
bra, from a need for specific material world contexts where utility is clear, to the
development of a language for expressing and manipulating generality.
Koellner, Jacobs, Borko, Roberts and Schneider quote Moses and Cobb (2001)
finding that
for students to find algebra valuable and engaging and to participate in discussions, it is
critical that the teacher select tasks that are relevant to their lives and have more than one
solution strategy.
It is quite difficult to demonstrate that it is critical for learners to engage in tasks

relevant to their lives, because there are other effective ways to engage learners in
mathematical thinking other than appealing to their material world experience (for
example, Realistic Mathematics Education: see Gravemeijer 1994). As Vygotsky
(1978) wondered, why go to school if all you are going to do is encounter what
you would encounter outside school? School is about scientific knowledge that can
only be accessed through mediation of a relative expert; it is about developing your
powers and coming into contact with possibilities unavailable elsewhere. Indeed
there are good reasons for not trying to appeal to learners through use of material
world contexts in their own lives, as it is likely to be overly simplified and might
564 J. Mason
be experienced by learners as an intrusion into their world. This needs much more
careful research before such assertions can be justified.
Seeing school algebra as fundamentally about expressing generality suggests that
whereas the customer attends to the particular (what they will get for what cost
in their situation), the entrepreneur has to establish pricing policies (generalities
covering most potential situations) (Mason et al. 2005). So algebra lies at the heart
of taking control of as-yet-unknown situations.
Tasks: Ainley and Pratt (2002) showed how attending to the twin aspects of purpose
(the immediate reason for working on a particular task) and utility (usefulness in
other situations) learners can experience reasons for using algebra.
In their chapter Ellerton and Clements consider what is involved in the task of
solving the equations that arise from acknowledging ignorance and using letters to
stand in for as-yet-unknowns. What seems to work pedagogically is treating equa-
tions as objects on which to act, transforming them into more succinct statements
that nevertheless leave the solution set invariant. For example Watson and Mason
(2002, 2005) report the effectiveness of getting students to construct their own com-
plicated equations from simple beginnings and how this sheds light on what solving
is really about. It is a special case of a principle that could pervade the use of al-
gebra in school and mathematics more generally: get students to make things more
complicated before trying to teach them how to simplify.
Ellerton and Clements also chart major obstacles to appreciating what algebra is
about, mostly arising from failure to invoke learners’ own powers of mathematical
sense-making, but rather trying to induce manipulation of formal symbols with-
out adequate motivation or generative experience. The pedagogic issue is whether
learners are encouraged to develop the multiple perspectives necessary for think-
ing algebraically, where an expression such as (2x + 3) + (3x − 4) can be seen
as an expression of generality; as a calculation to be carried out; as instructions
on how to carry out that calculation; and as the answer to a calculation with an
as-yet-unspecified-unknown value. These subtly different multiple perspectives can
be unified under the mathematical theme of freedom & constraint which pervades
mathematics when viewed as a constructive enterprise (Watson and Mason 2005).
Thus solving any routine exercise or problem can be viewed as seeking all pos-
sible mathematical objects meeting certain constraints. Rather than impose all the
constraints at once, it is sometimes helpful to impose them sequentially, seeking to
express the full general class of solutions at each stage. Freedom is gradually cur-
tailed until a solution set (which may be empty) is located. Theorems can be viewed
similarly, as statements about necessary constraints on full freedom (generality) in
order that some conclusion ‘always’ holds. Learners’ sense of the import and sig-
nificance of ‘always’ and ‘all’ in relation to numbers lies at the heart of successful
algebraic thinking.
Resources: Since all learners who get to school have displayed the requisite pow-
ers to make use of algebra as a language of expression of relationships, these are
available as resources. The social context includes the collective use of powers by
learners working together, as well as institutionalised practices which can be sup-
portive of mathematical thinking or can inhibit it. Pedagogic resources in the form
of prompts to exploration through to standard texts and software packages of various

forms can prompt and facilitate algebraic thinking or can obstruct it.
Operational Means: It is necessary to draw upon resources in the form of con-
structively mathematical ways of working if the activity arising from engaging
in tasks is going to be transformed into constructive actions by and for learners.
Li, Peng and Song offer reflections on and some evidence that using structured
variation in exercises can be effective in drawing upon student resources (pow-
ers) to experience the power of algebra. Variation theory (Marton and Booth 1997;
Marton and Pang 2006) provides a rich resource to inform the construction of tasks
for learners so that learners can actually learn what is intended to be learned, that
is so that their attention is drawn to aspects of situation that can be varied and still
be instances of the same concept or use of some technique. In other words, atten-
tion can be and is drawn to general classes through awareness of properties. In the
next subsection, reference is made to situations in which pattern following and pat-
tern constructing were used in a sympathetically mathematical manner and at least
locally, over time, made a real difference.
Facility in manipulating symbols algebraically, which most text books identify as
the core of algebra, would arise perfectly naturally from the desire to see how it is
that there can be two or more different-looking expressions of the same generality
(Mason et al. 1985, 2005). Explorations in relationships between numbers could
provide all the experience required for symbol manipulation. The core resource is
relational thinking and thinking relationally.
Empson, Levi and Carpenter recast the meaning of learning with understanding
in terms of thinking relationally: to understand arithmetic is to think relationally
about arithmetic. What could feed transition to more formal algebra would be ex-
plicit experience of relational thinking as described by them and by Ellerton and
Clements, and indeed by other authors. Relational thinking involves children’s use
of fundamental properties of [arithmetic] operations and equality to analyze a prob-
lem in the context of a goal structure and then to simplify progress towards this
goal. More phenomenologically, it means having properties of arithmetic operations
come to mind that are or could be instantiated in the current situation (properties of
additive and multiplicative identities, inverse connection between addition and sub-
traction, multiplication and division, and others which go to make up the axioms of
an integral domain and/or field).
Hewitt (1998) also emphasises strongly that it is impossible to learn arithmetic
competently without engaging in algebraic or algebraic-like thinking. In this he mir-
rors Gattegno (1970, 1987). It seems patently clear that no learner is expected to
memorise all possible two and three digit subtractions, much less multiplications
and divisions, so learners must always have been expected to generalise for them-
selves. This involves some sort of algebraic or pre-algebraic thinking at the very
least. Similar sentiments can be found in the work of Varga in Hungary in the 1970s
and in ancient Egyptian papyry and Babylonian tablets, where one finds prompts
such as “thus is it done”; “do thou likewise” (Gillings 1972).
There may be a subtle difference between ‘relational thinking’ and ‘thinking re-
lationally’. The former is more clearly something that can be observed in clinical
566 J. Mason
interviews, while the latter is more experiential, but could be taken to emphasise the
stance being taken, the way attention is structured. Thus a learner can be fully im-
mersed in particularities and ‘acting as if’ (James 1890) they know without actually
being consciously aware of the particular relationships as instantiations of general
properties, what Vergnaud (1981) called ‘theorem in action’. Thus 37 + 45 − 37
can be seen immediately as 45 without any calculation, without being aware more
generally, without bringing to articulation that adding and subtracting the same thing
leaves the rest unchanged and order of operations doesn’t matter here either (Molina
and Mason 2009; Mason et al. 2009). In interview, probing for justifications may
prompt learners to bring to expression some articulation of the property being in-
stantiated, though whether this intervention is sufficient to make the awareness ro-
bust over time is another matter for pedagogic investigation. Thinking relationally
could be indicated by overt expression of awareness that perceived properties are
being instantiated in particular relationships, whereas relational thinking could be
used to indicate at least a recognition of relationships in the particular situation. The
difference between these may be subtle, but may also be at the heart of difficulties
with mathematics. As Empson, Levi and Carpenter suggest, to understand arith-
metic ought to mean to think relationally about arithmetic at least in their sense if
not in the slightly extended sense.
Where learners are stimulated to use algebra to justify the generality of conjec-
tures, they are likely to encounter, among other things, one of the ways in which
symbols liberate attention. You can let go of context and meaning and concentrate
on formal manipulation, confident that values and relationships remain invariant.
Paulo Boero (2001) observed that successful symbol manipulation requires antici-
pation: you don’t just randomly manipulate, but rather you have a goal (simplifica-
tion, factoring, graphing) in mind and you choose actions that further those goals.
Empson, Levi and Carpenter make a similar observation, which they link back to
Piaget et al. (1960).
What Makes ‘Algebra’ Early?
From a traditional perspective, ‘early’ means prior to the institutional decision to

present algebra formally as supposedly familiar procedures applied to as yet unfa-
miliar objects (letters). From a visionary perspective, there is no such thing as ‘early
algebra’, because the roots of algebraic thinking are present from birth if not before.
When a child in the womb starts ‘kicking’ and turning, there are times when it seems
as though a stimulus from outside is producing a response inside, and certainly it
works in reverse for the mother. At some time in the womb, patterns of reaction
are set up. Indeed the behaviourist school of psychology took stimulus-response as
the basic action through which neural pathways are activated and reinforced. Be-
haviourism goes a long way towards explaining or accounting for the majority of
human behaviour where it proceeds in reaction to outer and inner stimuli through
the enacting of habit (for a more detailed history, see Gardner 1985). Patterned
behaviour can be seen as an early form of generalising, whether intentional or so-

matic in origin.
The neonate quickly learns to recognise (strictly speaking ‘to behave as if it
recognises’) mother, food, and then social practices. Intense generalisation is going
on. Granted it is probably not intentional, or reflective, but nevertheless awareness
of repetition makes learned action possible. Here awareness is used in the sense of
Gattegno (1970, 1987) to mean ‘that which enables action’. By the time a child gets
to school, it has made use of its natural powers both to generalise and to specialise
(instantiate generality in particular situations) among others (Mason 2008).
Think of the acquisition of language, the ability to behave in ways that fit with
other people’s expectations and not only to coordinate actions of the senses with oth-
ers, but to coordinate the coordinations of those actions (the definition of language
in Maturana 1972; see also Maturana and Varela 1988). It is immensely complex,
and it has been noted many times that if we had to teach children to speak as well as
to read we might have a largely silent population. Learning to speak is something al-
most all children do for themselves through the stimulus of others. Gattegno (1973,
1975), inspired by such observations, took them a stage further and suggested that
something (he called it the mind) actually teaches the brain.
In these early stages of growth and development, the brain-mind begins to reflect
on its actions and its potentialities. It tries initiating actions: it smiles, it cries, it
produces language-like sounds and sentence-like tonal sequences. It is experienc-
ing relationships between action and reaction. The child who gets great pleasure
from dropping things onto the floor from a highchair may be learning to attract
and retain adult attention. These are all instances of the child developing control
over its powers through developing its awareness (as an ability to act), by general-
ising.
Davydov (1972/1990) proposes that human intelligence instantiates the general
rather than generalising the particular, and that the earliest work on number can be
about expressing relationships symbolically before particular number names are in-
voked. Such an approach has been shown to accelerate the learning of arithmetic and
algebra (Schmittau 2003, 2005; Dougherty and Slovin 2004, see also Dürr 1985, and
Gerhard 2009). Hewitt (1998) strongly suggests that arithmetic cannot be properly
learned without involving algebraic thinking about relationships and generality.
Papic (2007) has demonstrated convincingly that (at least some) children aged
4–6 are perfectly capable of coordinating their actions so as to reproduce, then
create for themselves, complex repeating patterns of objects varying in colour and
spatial position (and doubtless other qualities as well). The experiment arose be-
cause kindergarten teachers asked whether it would be possible to engage children
in mathematically sensible tasks arising from the children’s own play with objects.
This is an example of what Cusi, Malara and Vavarra recommend:
The anticipation of generational pre-algebraic activities at the beginning of primary school,
and even before that, at kindergarten, to favour the genesis of the algebraic language, viewed
as a generalizing language, while the pupil is guided to reflect upon natural language.
Beatty (2010) has then shown that (at least some) 11–12 year olds are perfectly
capable of counting the numbers of objects in complex (linear) growth patterns, of
568 J. Mason
re-presenting the counts graphically as well as arithmetically and pre-algebraically,

and through being encouraged to work and think mathematically, to discover and
use the arithmetic of negative numbers for themselves (see also Moss 2002, 2005;
Moss and Beatty 2006; Beatty 2010). Carrehar, Schlieman and colleagues have a
trail of papers demonstrating how primary school children are perfectly capable of
engaging with algebraic thinking, including symbols (Carraher et al. 2006, 2007;
Schliemann et al. 2007).
Comparisons
Smith studies a contrast between some algebra lessons in the USA and some in
Japan, finding that the former features
a procedural approach . . . where students focus on getting answers through a series of rou-
tine steps, [whereas] in contrast, the Japanese lesson highlights a strong focus on building
generalized solution methods and understanding relationships represented in systems of
equations.
The overt use of learners’ powers may be a core feature of differences in perfor-
mance and appreciation of algebra. Whereas an ‘ideal’ form of teaching algebra
would call upon and develop learners’ powers such as to see generality through par-
ticulars (to generalise, that is, to recognise relationships as instantiations of proper-
ties) and to see particular instances in generalities (to specialise, that is, to recognise
instantiations of perceived properties), traditional algebra teaching tries to carry out
these actions for learners. The result is that learners learn to park their own pow-
ers at the classroom door, and simply try to use worked examples as templates for
rehearsal on exercises.
Transforming Algebra Teaching and Learning as an Activity
Treating the description of traditional algebra teaching as the current state, and the
description of envisioned algebra teaching and learning as the goal, what resources
and what tasks can be called upon to reach this goal?
The kinds of tasks needed to effect a transition from traditional to visionary al-
gebra teaching are many and varied. For example, for Ellerton and Clements
an important aspect of the intervention program was the creation of an environment in
which all students would reflect metacognitively on the strategies that they used when they
attempted to solve equations and inequalities (Clements and Ellerton 2009).
This is manifested in a very different form for Koellner, Jacobs, Borko, Roberts and
Schneider. The range of possible interventions is vast. It may be that there is some
underlying structural factor that is common to many or most successful interven-
tions, but it is much more likely that, since we are dealing with human beings who
often manifest automatised behaviour but who can exercise choice and take initia-
tive, there is no small collection of underlying factors. Rather the situation is and
remains immensely complex.
There are many resources available, the chapters in this section being but the
tip of the iceberg. Many authors have shown how specific interventions in their
own situation can have a marked improvement in learners’ attitudes to, disposition
for, appreciation of and performance in algebra. But are these changes sustained?
What happens when learners are returned to the more traditional approach in which
mastery of procedures dominates mathematical thinking and learners’ powers are
usurped by text and teacher? What sorts of tasks are available to bring about a trans-
formation in teaching and hence learning of algebra for all?
The chapter by Koellner, Jacobs, Borko, Roberts and Schneider reveals much
about the massive inertia that holds back didactic5 innovation through a cycle of
not-learning well and failing to learn how to learn reproducing itself in each gen-
eration. They show another instance of how it is locally possible to break out of
the static cycle in particular cases. However it remains difficult to engineer globally
precisely because trying to change others without attending to the whole person-
psyche is bound to fail eventually, despite appearances of success in the short term.
Engineering solutions through adjusting policies is more likely to suppress than to
redress failing habits.
How Can Locally Successful Teaching Be Engineered for All?
Approaches which engage learners in encountering and expressing generality prior

to mastering rules for manipulating expressions and equations between expressions
have proved successful locally, but how can these be promulgated throughout the
community of mathematics teachers? Although not specific to algebra, this question
has been at the heart of research in algebra and algebraic thinking for centuries.
Robert Recorde (1543) tried to support those who needed arithmetic and algebra
for mercantile and military reasons but could not afford to engage a tutor; Sawyer
(1959) wrote popular books trying to make algebra accessible; the Open University
(1984; see also Mason et al. 1985, 2005) created materials for teachers and fea-
tured them in distance learning courses; mathematics associations have made many
resources available in many different countries. Popularisation of mathematics has
become itself a popular genre, all with little avail to the global state of algebra teach-
ing.
Teaching is a caring profession. It involves both caring for learners as human
beings and caring for mathematics as a powerful discipline. Sometimes it seems
5 I use didactic in the European sense to refer to actions or tactics particular to specific mathematical
topics, concepts, procedures etc., reserving pedagogic for actions or strategies that apply to many
or even any mathematical topic.
570 J. Mason
difficult to combine the two: caring for learners can lead to watering down of math-
ematics (Stein et al. 1996), and caring for mathematics can lead to leaving learn-
ers confused and ill-disposed towards mathematical thinking. As Cusi, Malara and
Navarra say about their project,
Our aim is to make teachers aware and caring about [the obstacles presented by traditional
algebra teaching] and provide them with instruments that enable them to design and imple-
ment powerful interventions to face it.
Ellerton and Clements offer one approach, or contribution to engineering, through

the construction of tasks for teachers that bring them up against pedagogic, didac-
tic and epistemological obstacles. Their cleverly assonant list of Realize, Review,
Reflect, Revisit and Retain has the makings of a useful contribution, at least until it
becomes mechanised, made routine, and treated superficially, as has been the fate of
initiatives to date. Nothing ‘works’ universally, nor for very long locally. The CGI
project (Carpenter and Fennema 1999) worked in a similar way, as do many pro-
grammes of professional development. Perhaps the community could reach some
agreement on such tasks, updating and extending them as necessary, as part of a
unified attack on misunderstandings and blinkered perspectives on algebra and pre-
algebraic thinking.
One of the many tasks involved in engineering a sea-change in algebra teaching
is of course the enhancement of teachers’ professional knowledge about algebra and
the teaching of algebra as Koellner, Jacobs, Borko, Roberts and Schneider remind
us. They propose a problem-solving cycle of workshops in which teachers engage
with algebraic thinking for themselves, then analyze instructional practices, and then
analyze student thinking, both of the latter involving video-taped lessons taught by
the teacher. Although only reporting on one teacher, their approach is consonant
with many other programmes of teacher engagement with mathematics, pedagogy
and student thinking. But could this be accessed by all teachers? What might one do
about teachers content to continue with their present practices and limited vision?
A quotation from Cicero comes to mind:
You will be as much value to others as you have been to yourself.
There is nothing so powerful as becoming aware of your own experience, which

enables you to speak to the conditions and experience of others.
What Is and Could Be Researched?

As indicated at the beginning, there are many descriptions of the state of play under
a limited vision of algebra. Many researchers have tried to show that this or that vari-
ation or alternative treatment can be successful, but rarely is there sufficiently pre-
cise description of the situation didactique, the cultural-historical-contextual back-
ground and the established ways of working so as to enable a reader to attempt to
replicate or try something similar.
The current research climate encourages, even requires well founded and ac-
knowledged theoretical frames, and values empirical studies of what is currently the
case, as the basis for evidence-based action. Studies of what could happen tend to
be, as indicated at the beginning, local rather than global.
It would be really useful to obtain and coordinate empirical evidence supporting
or contradicting visions of what is possible, while much less useful to make yet more
records of the behaviours and dispositions of students, novice teachers and teachers
who are under various illusions about what they are doing and why. What is needed
is research into making the insights arising from the reports of researchers such
as Papic, Beatty and Hewitt, not to say many others, widely available in a setting
that enables teachers to adapt to a broader and more visionary perspective on what
constitutes and what is possible in learning algebra.
What are the conditions which enable teachers to respond to, accommodate and
adapt to a broader vision of algebraic thinking in particular, and mathematics more
generally? Of real value would be larger scale studies of conjectures such as those
promulgated by the authors of these chapters and elsewhere, that expressing gener-
ality is at the heart of algebraic thinking and that it is fostered most effectively by
promoting relational thinking. Is it generally the case that the transition from recog-
nising relationships in particular situations, to perceiving these as instantiations of
properties that hold in many different situations is subtle and delicate but neverthe-
less important? If over the long term fractions, like percentages, are considered as
operators and so always accompanied by an ‘of’ until learners spontaneously treat
them as numbers (the result of acting them on a unit) might this release learners
from the bonds of memorising procedures for rational arithmetic?
Professional development often reduces to tips for teaching. Is the ongoing hy-
parchic ‘becoming’ of teachers being researched? More insight is needed rather than
a catalogue of what teachers do and do not do in the face of particular tasks.
Are tasks being analysed in depth both in terms of the resources currently avail-
able (especially learner propensities and attunements), and ways of working or
socio-mathematical norms but also in terms of what is possible (teacher attunements
and awarenesses)?
Are studies of learner likes and dislikes sufficient to take full account of learners’
motivational axis in relation to their operational-means axis, both as currently man-
ifested, and more valuably perhaps, as what could be the case if algebraic thinking
were recognised and developed throughout the school years?
What Is Really Researched?
On the one hand, every generation has to re-discover and re-interpret the situation
in which they find themselves, and express it in the vocabulary of the times. This
is an ongoing community endeavour. At the individual level, every teacher has to
construct for themselves a teacher-self. Yet it seems wasteful for each generation
to begin anew without drawing on the wisdom of the past. Indeed they must draw
on this wisdom if they are not to be doomed to repeat the inadequacies of the past.
572 J. Mason
Standing on the shoulders of giants6 is not an appropriate image for the ongoing
process of constant hyparchic ‘becoming’ that marks a true teacher, since so much
has to be re-constructed and re-experienced for oneself.
On the other hand, mathematics education has been a domain of enquiry extend-
ing over several millenia: Plato compares Greek and Egyptian teaching, and surely
every generation has had teachers who questioned the orthodoxy and tried to do
better for their students than they felt was done for them. The Sufi mystic Jalaladdin
Rumi (1999) writing in the 13th century observed that
Students of cunning have consumed their hearts and learned only tricks; they’ve thrown
away real riches: patience, self-sacrifice, generosity. Rich thought opens the way.
So mathematics education as a domain of enquiry has a problem: how is it possible

to learn from the past while being sensitive to the present and to the hyparchic
nature of teaching, learning and researching? Perhaps this is really what is being
researched?
Current fashions in research (evidence-based action, large-scale statistical stud-
ies, studies embedded in a single clearly articulated theoretical setting) do not easily
lend themselves to discovering what is possible. Ethnographically-based closely-
watched actions of teacher and learners produces little more than thick descrip-
tions of local current practice. As a recent adage puts it, teaching based on learn-
ing as ‘being told, being shown, and extensive rehearsal’, as ‘show me what you
want me to do, then make me do it over and over’ is the largest-scale educa-
tional experiment ever conducted. And it has been a failure, repeatedly, genera-
tion after generation. The procedural dominates the conceptual, to the extent that
students ask to be shown/told what to do so that they can do it, and compe-
tence declines despite standards supposedly rising. The language of the contrat di-
dactique (Brousseau 1984, 1997) and the associated teaching tension capture this
vividly:
the more clearly and precisely the teacher indicates to students what behaviour is sought,
the easier it is for the students to display that behaviour without generating it for themselves,
without internalising anything.
But the same desire, the same action is visible in teacher education: teachers want
to be told what it is that is required of them (on which they will be judged);
novice teachers want to be told how to teach; masters students want to be told
what they have to read and what they have to do to pass the course, and so on.
Indeed it is not surprising that teachers reflect the dominant culture of ‘tell me
what you want and I will deliver it’, and that this in turn infects their classroom
behaviour, and so amplifies the enculturation process in which students are im-
mersed.
Thus studies which capture only the gross elements of tutor-teacher interaction
(tasks used, discourse employed) at best demonstrate that it is possible to describe
and count types of interactions in one or another vocabulary, usually classified into
6 An expression apparently coined by Bernard of Chartres around 1126 expressing a familiar me-
dieval sentiment, sometimes humbly and sometimes proudly; see Merton (1965).
several phases or stages. But what is missing is the quality of the interactions. What
is missing is the lived experience of teacher and of students. Why is this not more
widely researched?
Interactions are plural because of the complexity of the impulses or agents acting,
among which are
experience of schooling in the past (primary, secondary, tertiary, industry, leisure,
etc.);
expressed and covert desires and demands of various institutions, from national,
state/province/county, school board, school or college, department, and finally,
teachers themselves, not to say parents and guardians, unions and other teacher
associations, and the media;
personal vision, desire, and competencies;
conventions (historical-cultural-social) and practices of mathematicians as mani-
fested in written and digital (animations) presentations;
structural necessities that follow from and can be deduced from and in mathemat-
ics;
heuristics and the use of natural human powers;
social and psychological forces acting on and within individuals and groups of
adolescents particularly.
The list goes on. Furthermore each of these operates through the whole of the hu-
man psyche: behaviour, affect (emotion) and cognition, not to say intention and will
which are manifested as attention. The complexity (Davis et al. 2006) can be over-
whelming. Yet it seems that it is only by maintaining complexity while enquiring,
probing and acting that there is any chance of long term success.
Bennett (1966, p. 48) found it useful to see the present moment as the coalescence
of being and becoming, at the confluence of three dimensions each with two aspects:
the past, as comprehended and the future if things carry on as they are;
what is materially possible and what is desired or imagined as possible
what is available within reach and what is unavailable, out of reach, but influential.
Observation is theory laden, as Hanson (1958) noticed and doubtless many before
him. Goodman asserted that “we want our theories to be as fact laden as our facts are
theory laden”. Maturana (1988) noted that “everything said is said by an observer”.
Research into what is the case is essentially extra-spective, that is, one or more peo-
ple observe one or more other people acting and interacting. What is noticed, what
is discerned, reveals as much about the sensitivities, the attunements, propensities
and dispositions of the observers as it does about the observed. Elsewhere I have
proposed that the ratio of the preciseness of what is revealed about what is observed
to the preciseness of what is revealed about the researcher is roughly constant. In
other words, the more deeply or precisely you probe into actions and activities, the
more you reveal about what it is that catches your attention.
Applied to early algebra, what the different chapters report on reveals something
of what the authors consider to be early and algebraic, among other things, and
what they consider to be of importance for the future, either as a platform on which
to stand, a goal to seek, or a means of achieving that goal. Not all four elements of
the activity are made explicit however.
574 J. Mason
Conclusions
Lack of vision is one way to summarise a plethora of research findings arising from
probing prospective teachers’ understanding of the mathematics they are supposed
to be learning to teach. Perhaps the repair mentality induces a move to ‘at least
getting them to obtain and recognise the correct answers’ rather than deeply under-
standing and appreciating the underlying conceptual relationships.
The obstacles to visionary teaching of algebra lie in the inherent inertia of a com-
plex system that includes human beings. They are augmented by current cultural
foci on ‘being told what is required and then producing that’ at all levels of educa-
tion, though these merely reflect a consequence of the ’customer is right’ stance of
entrepreneurs. More specifically mathematically, a blinkered, procedurally oriented
perspective on what school algebra is and could be inhibits and obstructs the take
up of a richer and broader vision of what school algebra could be, and as far as I am
concerned must be if mathematics education is going to develop.
References
Ainley, J., & Pratt, D. (2002). Purpose and utility in pedagogic task design. In A. Cockburn & E.
Nardi (Eds.), Proceedings of the 26th Annual Conference of the International Group for the
Psychology of Mathematics Education (Vol. 2, pp. 17–24). Norwich, UK: PME.
Beatty, R. (2010). Pattern rules, patterns and graphs: Analyzing grade 6 students’ learning of
linear functions though the processes of webbing, situated abstraction, and convergent change.
Unpublished PhD thesis. Toronto: OISE.
Bennett, J. (1966). The Dramatic Universe: Man and His Nature (Vol. 3). London: Hodder &
Stoughton.
Bennett, J. (1970). Elementary Systematics: A Tool for Understanding Wholes. Sherborne: Coombe
Springs Press.
Boaler, J. (2002). Experiencing School Mathematics (revised and expanded edition). Mahwah:
Erlbaum.
In R. Sutherland (Ed.), Algebraic Processes and Structures (pp. 99–119). Kluwer: Dordrecht.
Brousseau, G. (1984). The crucial role of the didactical contract in the analysis and construction
of situations in teaching and learning mathematics. In H. Steiner (Ed.), Theory of Mathemat-
ics Education, Paper 54 (pp. 110–119). Institut fur Didaktik der Mathematik der Universitat
Bielefeld.
Brousseau, G. (1997). (N. Balacheff, M. Cooper, R. Sutherland, & V. Warfield, Trans.) Theory of
Didactical Situations in Mathematics: Didactiques des Mathématiques, 1970–1990. Kluwer:
Dordrecht.
Carpenter, T., & Fennema, E. (1999). Childrens’ Mathematics: Cognitively Guided Instruction.
Portsmouth: Heineman.
Carpenter, T., Franke, M., & Levi, L. (2003). Thinking Mathematically. Portsmouth, NH: Heine-
mann.
Carraher, D., & Schliemann, A. (2007). Early algebra and algebraic reasoning. In F. Lester (Ed.),
Second Handbook of Research on Mathematics Teaching and Learning (pp. 669–705). Reston:
Carraher, D., Schliemann, A., Brizuela, B., & Earnest, D. (2006). Arithmetic and algebra in early
mathematics education. Journal for Research in Mathematics Education, 37(2), 87–115.
Carraher, D., Schliemann, A., & Schwartz, J. (2007). Early algebra is not the same as algebra
early. In J. Kaput, D. Carraher, & M. Blanton (Eds.), Algebra in the Early Grades (pp. 235–
272). Mahwah: Erlbaum.
Clements, M., & Ellerton, N. (2009). A model for improving prospective mathematics teachers’ al-
gebra content knowledge. Brunei International Journal of Science and Mathematics Education,
1(1), 68–84.
Davidov, V. (1972/1990). (J. Teller, Trans.) Types of Generalization in Instruction: Logical and
Psychological Prelims in the Structuring of School Curricula. Reston, VA: National Council of
Teachers of Mathematics.
Davis, B., Summara, D., & Simmt, E. (2006). Complexity & Education: Inquiries into Learning,
Teaching and Research. Mahwah: Lawrence Erlbaum.
Dougherty, B., & Slovin, H. (2004). Generalized diagrams as a tool for young children’s problem
solving. In M. Høines & A.-B. Fuglestad (Eds.), The 28th International Conference of the
International Group for the Psychology of Mathematics Education. Bergen: Bergen University
College.
Dürr, C. (1985). Vorschläge für intensives Arbeiten mit Variablen im Mathematikunterricht
der unteren Schuljahre (Suggestions for intensive working with variables in mathematics
lessons of the lower years). Wissenschaftliche Zeitschrift der Humboldt-Universität zu Berlin /
Mathematisch-naturwissenschaftliche Reihe, 34(7), 615–623.
Dweck, C. (2000). Self-Theories: Their Role in Motivation, Personality and Development.
Philadelphia: Psychology Press.
Gardner, H. (1985). The Mind’s New Science: A History of the Cognitive Revolution. New York:
Basic.
Gattegno, C. (1970). What We Owe Children: The Subordination of Teaching to Learning. London:
Routledge & Kegan Paul.
Gattegno, C. (1973). In the Beginning There Were No Words: The Universe of Babies. New York:
Educational Solutions.
Gattegno, C. (1975). The Mind Teaches the Brain. New York: Educational Solutions.
Gattegno, C. (1987). The Science of Education Part I: Theoretical Considerations. New York:
Educational Solutions.
Gerhard, S. (2009). Problem solving without numbers. In Proceedings of the Sixth Congress of the
European Society for Research in Mathematics Education (pp. 499–508). C.E.R.M.E.
Gerofsky, S. (1996). A linguistic and narrative view of word problems in mathematics education.
For The Learning of Mathematics, 16(2), 36–45.
Gillings, R. (1972 reprinted 1982). Mathematics in the Time of the Pharaohs. New York: Dover.
Gravemeijer, K. (1994). Developing Realistic Mathematics Education. Culenborg: Technipress.
Greeno, J. (1994). Gibson’s affordances. Psychological Review, 101(2), 336–342.
Hanson, N. (1958). Patterns of Discovery: An Enquiry into the Conceptual Foundations of Science.
Cambridge: Cambridge University Press.
Hewitt, D. (1998). Approaching arithmetic algebraically. Mathematics Teaching, 163, 19–29.
James, W. (1890 reprinted 1950). Principles of Psychology (Vol. 1). New York: Dover.
Kieran, C. (2007). Learning and teaching algebra at the middle-school through college levels:
Building meaning for symbols and their manipulation (A project of the National Council of
Teachers of Mathematics). In F. K. Lester (Ed.), Second Handbook of Research on Mathematics
Teaching and Learning (pp. 707–762). Charlotte, NC: Information Age Publishing.
Marton, F., & Booth, S. (1997). Learning and Awareness. Hillsdale: Lawrence Erlbaum.
Marton, F., & Pang, M. (2006). On some necessary conditions of Learning. Journal of the Learning
Sciences, 15(2), 193–220.
Maslow, A. (1971). The Farther Reaches of Human Nature. New York: Viking Press.
Mason, J. (2001). On the use and abuse of word problems for moving from arithmetic to algebra. In
H. Chick, K. Stacey, J. Vincent, & J. Vincent (Eds.), The Future of the Teaching and Learning of
Algebra, Proceedings of the 12th ICMI Study Conference (pp. 430–437). Melbourne: University
of Melbourne.
576 J. Mason
Mason, J. (2008). Making use of children’s powers to produce algebraic thinking. In J. Kaput, D.
Carraher, & M. Blanton (Eds.), Algebra in the Early Grades (pp. 57–94). New York: Lawrence
Erlbaum.
Mason, J., Graham, A., Pimm, D., & Gowar, N. (1985). Routes to, Roots of Algebra. Milton
Keynes: The Open University.
Mason, J., Johnston-Wilder, S., & Graham, A. (2005). Developing Thinking in Algebra. London:
Sage (Paul Chapman).
Mason, J., Stephens, M., & Watson, A. (2009). Appreciating mathematical structure for all. Math-
ematics Education Research Journal, 21(2), 10–32.
Maturana, H. (1988). Reality: The search for objectivity or the quest for a compelling argument.
Irish Journal of Psychology, 9(1), 25–82.
Maturana, H., & Varela, F. (1972). Autopoesis and Cognition: The Realization of the Living. Reidel:
Dordrecht.
Maturana, H., & Varela, F. (1988). The Tree of Knowledge: The Biological Roots of Human Un-
derstanding. Boston: Shambala.
Merton, R. (1965). On the Shoulders of Giants: A Shandean Postscript. New York: Free Press.
Molina, M., & Mason, J. (2009). Justifications-on-demand as a device to promote shifts of attention
associated with relational thinking in elementary arithmetic. Canadian Journal for Science
Mathematics and Technology Education, 9(4), 224–242.
Moses, R. P., & Cobb, C. (2001). Organizing algebra: The need to voice a demand. Social Policy,
31(4), 4–12.
Moss, J. (2002). Percents and proportion at the center: Altering the teaching sequence for rational
number. In B. Littweiller (Ed.), Making Sense of Fractions, Ratios, and Proportions (pp. 109–
120). Reston: National Council of Teachers of Mathematics.
Moss, J., & Beatty, R. (2006). Knowledge building in mathematics: Supporting collaborative learn-
ing in pattern problems. Computer-Supported Collaborative Learning, 1, 441–465.
Moss, J. (2005). Pipes, tubes, and beakers: Teaching rational number. In J. Bransford & S. Donovan
(Eds.), How Children Learn: History Science and Mathematics in the Classroom. Washington:
National Academy Press.
Newton, I. (1707). Arithmetica Universalis (Ed. William Whiston). Cambridge.
Open University (1984). EM235: Developing Mathematical Thinking. Milton Keynes: Open Uni-
versity.
Papic, M. (2007). Mathematical patterning in early childhood: an intervention study. Unpublished
Ph.D. thesis. Sydney: Macquarie University.
Piaget, J., Inhelder, B., & Szeminska, A. (1960). The Child’s Conception of Geometry (E. A. Lun-
zer, Trans). New York: Basic.
Recorde, R. (1543). The Ground of Arts: Teaching the Perfect Worke and Practise of Arithmeticke,
Both in Whole Numbers and Fractions. London: Harper, Thomas. Reprinted by New York: Da
Capo Press, 1969.
Rumi, J. (1999). C. Barks (Trans.) The Essential Rumi. London: Penguin.
Sawyer, W. (1959). A Concrete Approach to Abstract Algebra. London: Freeman.
Schliemann, A., Carraher, D., & Brizuela, B. (2007). Bringing Out the Algebraic Character of
Arithmetic: From Children’s Ideas to Classroom Practice. Mahwah: Lawrence Erlbaum Asso-
ciates.
Schmittau, J. (2003). Cultural historical theory and mathematics education. In A. Kozulin, B.
Gindis, S. Miller, & V. Ageyev (Eds.), Vygotsky’s Educational Theory in Cultural Context.
Schmittau, J. (2005). The development of algebraic thinking: A Vygotskian perspective. Zentral-
blatt Fuer Didaktik Der Mathematik (International Review of Mathematics Education), 37(1),
16–22.
Stein, M., Grover, B., & Henningsen, M. (1996). Building student capacity for mathematical think-
ing and reasoning: An analysis of mathematical tasks used in reform classrooms. American
Educational Research Journal, 33(2), 455–488.
Tahta, D. (1972). A Boolean Anthology: Selected Writings of Mary Boole on Mathematics Educa-
tion. Derby: Association of Teachers of Mathematics.
Vergnaud, G. (1981). Quelques orientations théoriques et méthodologiques des recherches

françaises en didactique des mathématiques. In Actes du Vième Colloque de PME (Vol. 2,
pp. 7–17). Grenoble: Edition IMAG.
Verschaffel, L., Greer, B., & de Corte, E. (2000). Making Sense of Word Problems. Lisse: Swets &
Zeitlinger.
Vygotsky, L. (1978). Mind in Society: The Development of the Higher Psychological Processes.
London: Harvard University Press.
Ward, J. (1706). The Young Mathematicians Guide, Being a Plain and Easie Introduction to the
Mathematics in Five Parts. . . . London.
Watson, A., & Mason, J. (2002). Student-generated examples in the learning of mathematics.
Canadian Journal of Science, Mathematics and Technology Education, 2(2), 237–249.
Watson, A., & Mason, J. (2005). Mathematics as a Constructive Activity: Learners Generating
Examples. Mahwah: Erlbaum.
Whiteside, D. (Ed.) (1972). The Mathematical Papers of Isaac Newton (Vol. V, pp. 1683–1684).
Overall Commentary on Early Algebraization:
Perspectives for Research and Teaching
Carolyn Kieran
Arithmetic itself must be viewed with ‘algebra eyes’

(Subramaniam & Banerjee, this volume)
The twenty-nine chapters of this volume on early algebraization, which include an

introduction and commentary for each of the three main parts, reveal the rich di-
versity that characterizes the rapidly evolving field of early algebra. Cai and Knuth,
in their introductory chapter, point out that the development of students’ algebraic
thinking in the earlier grades is not a new idea, but has been part of school practice
in several countries around the world since the 1950s. Nevertheless, it was not until
the mid-1990s that the idea took hold more broadly and that publications began to
reflect the interest that researchers were investing in this area. Each new collection
of writings since then has made advances on its predecessors as researchers continue
in their efforts to unpack the central notions of school algebra and reflect on how
they might be made accessible to the younger student at the elementary and middle
school levels. This latest collection is no exception. With its three parts that artic-
ulate the ways in which researchers are currently conceptualizing early algebraiza-
tion from curricular, cognitive, and instructional perspectives, this volume offers to
researchers, teachers, curriculum developers, professional development educators,
and policy makers alike some of the most recent thinking in the field.
The research that is presented within sheds light on how the term algebraization
is being considered: algebraization concerns the nature of the thinking that is basic
to algebra, along with the conceptual areas within early and middle school math-
ematics that can be exploited pedagogically in this early algebraic terrain, as well
as the ways in which teachers can help students develop such thinking. The over-
all commentary that I have been invited to write attempts to synthesize the ways in
which the researchers whose work is described in the chapters of this volume have
C. Kieran ()
Département de Mathématiques, Université du Québec à Montréal, Montreal, Canada

580 C. Kieran
been shaping this triple aspect of algebraization. Their efforts will have an impact
not only on the way in which children come to think about their mathematics at
the elementary and middle school levels, but also on the way in which high school
students come to engage with algebra.
Shaping the Notion of Algebraic Thinking within Early Algebra

The citation with which I chose to open this commentary chapter, one that is drawn
from the Subramaniam and Banerjee chapter, states that arithmetic needs to be
viewed with ‘algebra eyes.’ Elsewhere, Blanton and Kaput (2008) have referred to
this phenomenon as algebrafying and have described it as transforming and extend-
ing the mathematics normally taught in elementary school toward algebraic think-
ing, with its intrinsic feature of generality, and including within this transformation
“the establishing of classroom norms of participation so that argumentation, conjec-
ture, and justification are routine acts of discourse” (p. 362). Taken together, these
two references suggest that the developing of ‘algebra eyes’ involves seeing the gen-
eral within arithmetic and that the more global mathematical reasoning processes of
argumentation, conjecturing, and justification are routes toward this goal. However,
as will be seen from the chapters within this volume, it involves much more than
this.
More than a decade ago, Kieran (1996) offered the perspective that algebraic ac-
tivity in school consists of three components: the generational; the transformational;
and the global meta-level, which includes analyzing relationships between quanti-
ties, noticing structure, studying change, generalizing, problem solving, justifying,
proving, and predicting. Although these three types of activities were framed against
the dual backdrop of both equation-based and function-based approaches, the ways
in which they might be adapted for an early algebra context were left largely unar-
ticulated.
Kaput (2008) has proposed a slightly different perspective on algebra. In his
opening chapter of the anthology, Algebra in the Early Grades, he specified the
two core aspects of algebraic reasoning to be (i) generalization and the expres-
sion of generalization in increasingly systematic, conventional symbol systems, and
(ii) syntactically guided action on symbols within organized systems of symbols.
Each of these core aspects is deemed, according to Kaput, to be found in varying
degrees throughout the following three strands of algebra: algebra as the study of
structures arising in arithmetic and in quantitative reasoning, algebra as the study of
functions, and algebra as the application of modeling languages.
While Kieran (2004) has argued that algebraic thinking in the earlier grades could
be construed in terms of the global, meta-level activity of algebra and be engaged
in without the use of the letter-symbolic, Kaput’s main thrust has been on the over-
arching role of generalization and its gradual symbolization. In any case, Radford,
one of the chapter authors of this volume on Early Algebraization, emphasizes that
“algebraic thinking is not about using or not using notations but about reasoning in
certain ways.”
Overall Commentary on Early Algebraization: Perspectives for Research and Teaching 581
In keeping with Radford, the issue in coming to grips with algebraic thinking
centers on what is meant by “reasoning in certain ways.” As an aside, it is noted
that scholars in the field of algebra education (be it at the high school level or ear-
lier) have yet to distinguish algebraic thinking from algebraic reasoning. While
the two terms are used interchangeably within this literature, classic approaches to
the study of mathematical reasoning tend to focus, in general, on ‘forms of reason-
ing,’ be they deductive, inductive, abductive, or analogical (Jeannotte 2010). When
viewed against the lens of classical-mathematical-reasoning terminology, the term
algebraic reasoning risks being interpreted too narrowly to encompass adequately
the various and diverse approaches to early algebra that are being considered within
this volume. Thus, I have opted within this commentary to use whenever possible
that which I consider to be the broader term, algebraic thinking. Taken as a whole,
the chapters of this volume make significant strides in unpacking not only the nature
and components of such thinking but also the manner in which it might be fostered
by teachers of elementary and middle school students. Although my organizational
structure and résumé of salient ideas from the chapters—the product of a diagonal
cut through the volume—do not preserve the rich detail that constitutes the central
contributions of the authors, I nevertheless attempt to point out within each of the
sections below those chapter aspects that I consider inject something new and im-
portant into the development of the field of algebra education. The research that is
presented in this volume, research that is shaping both our ways of thinking about
the nature and components of algebraic thinking and the routes by which its growth
might be encouraged, includes the following focal themes:
• Thinking about the general in the particular
• Thinking rule-wise about patterns
• Thinking relationally about quantity, number, and numerical operations
• Thinking representationally about the relations in problem situations
• Thinking conceptually about the procedural
• Anticipating, conjecturing, and justifying
• Gesturing, visualizing, and languaging.
Thinking about the General in the Particular

One of the pioneers of a generalization approach to the teaching and learning of
algebra, John Mason, has described algebraic thinking as follows:
Algebraic thinking is rooted in and emerges from learners’ natural powers to make sense
mathematically. At the very heart of algebra is the expression of generality. Exploiting al-
gebraic thinking within arithmetic, through explicit expression of generality makes use of
learners’ powers to develop their algebraic thinking and hence to appreciate arithmetic more
thoroughly. (Mason 2005, p. 310)
Nearly a dozen chapters in this volume express ideas that resonate with Mason’s,
that is, that the expression of generality is the core of algebraic thinking. Moreover,
their focus is on both the process of generalizing that contributes to the production of
582 C. Kieran
such expressions of generality as well as the generalized product. Thus, generalizing

is considered as both a route to, and a characteristic of, algebraic thinking.
For example, Rivera and Rossi Becker in their chapter draw our attention to their
finding that “individuals tend to see and process the same pattern differently . . . and
produce different generalizations for [that pattern],” while Britt and Irwin note that
“successful application of operational strategies demands an awareness of the gener-
ality of the operational strategy.” Russell, Schifter, and Bastable speak of “generaliz-
ing and justifying”; Koellner, Jacobs, Borko, Roberts, and Schneider, of “describing
and generalizing patterns”; and Cai, Moyer, Wang, and Nie, of “the development
of students’ algebraic thinking related to . . . making generalizations.” Both the pro-
cess and product aspects of generalizing are explicitly found in Blanton and Kaput
who, in their chapter within this volume, discuss “algebraic reasoning as an activ-
ity of generalizing mathematical ideas” and propose using these generalized ideas
as “objects of mathematical reasoning.” Similarly, Cooper and Warren argue for
both grasping and expressing generalities. In addition, Radford discusses “dealing
with generality through particular examples, in a manner that Balacheff (1987) calls
‘generic example,’ a way of seeing the general through the particular, as Mason
(1996) puts it.”
Radford, however, nuances the oft-found practice among many algebra-education
researchers to identify nearly all generalization activity within this area as algebraic.
His nuanced position is presented immediately below, within the focal theme of
‘thinking rule-wise about patterns.’
Thinking Rule-Wise about Patterns

In his chapter that describes second graders’ activity with pattern generalization,
Radford argues that the process of grasping a commonality in a sequence and ex-
tending it to a few subsequent items does not mean that students are thinking al-
gebraically. He points out that chimpanzees and birds can form commonalities too.
Rather, what
characterizes thinking as algebraic is that it deals with indeterminate quantities conceived
of in analytic ways . . . indeterminacy and analyticity are in fact bound together in a schema
or rule that allows the students to deal with any particular figure of the sequence, regardless
of its size . . . the students’ rule attests to a shift in focus: the student’s focus is no longer
specifically numeric . . . for the student’s emerging understanding, what matters is not the
[numeric] result; it is the rule, that is to say, the formula—the algebraic formula. (Radford,
this volume)
Put succinctly, it is the shift from the purely numeric to the devising of a rule or
calculation method involving indeterminates that constitutes a [pattern] generaliza-
tion that is algebraic in nature. The precise articulation that Radford brings to the
discussion of what is algebraic, and what is not, within the context of pattern gen-
eralization in early algebra is one that is important for the field. He identifies not
only a distinction between students’ using the visual and the numeric in action and
their movement toward a more general kind of thinking that is neither visualized
nor experienced directly, but also a distinction between this more general form of
thought within patterning activity and algebraic thinking itself.
Additional contributions from other chapters in this volume that bear on pattern
generalization include the research by Rivera and Rossi Becker who describe middle
schoolers’ activity with more complex patterns, by Moss and McNab who discuss
second graders’ reasoning about linear function and co-variation through the inte-
gration of geometric and numeric representations of growing patterns, by Watanabe
who provides details related to the functional underpinnings of patterning within the
Japanese curriculum, and by Cai, Ng, and Moyer who do likewise with respect to
the Singaporean curriculum.
Thinking Relationally about Quantity, Number, and Numerical

Operations
Empson, Levi, and Carpenter point out that relational thinking is almost entirely
neglected in typical U.S. elementary school classrooms. This reason alone would
make all of the ten or so chapters dealing with this approach to the development of
algebraic thinking required reading, for they offer a glimpse into what is possible
within an early algebra context. However, these chapters offer even more, with their
varying theoretical and cultural frameworks and rich descriptions of student and
teacher work in this area.
According to Empson et al., relational thinking “involves children’s use of fun-
damental properties of operations and equality to analyze a problem in the context
of a goal structure and then to simplify progress towards this goal”; such thinking
is also said to include anticipating those relations and actions that move one effec-
tively toward the final goal of a given situation. These authors pit relational thinking
against algorithmic thinking about operations where the goal structure can be sum-
marized as ‘do next’. An example of relational thinking that they provide involves a
student who has to calculate 1/2 + 3/4. This student unpacks 3/4 as 1/2 + 1/4 in
anticipatory fashion and reasons that 1/2 plus another 1/2 is equal to 1, then plus
another 1/4 is 1 14 . For Empson et al., to understand arithmetic is to think relation-
ally about arithmetic, and thinking relationally about arithmetic involves the kind of
property-based thinking that is used in algebra.
Several other chapters of this volume contribute equally important perspectives
on relational thinking, especially with respect to the conceptual arena of ‘unpacking
number.’ For example, Russell, Schifter, and Bastable describe how students ben-
efit from “explicit study of the operations by examining calculation procedures as
mathematical objects that can be described generally in terms of their properties
and behaviors”; Subramaniam and Banerjee argue that “understanding and learning
to ‘see’ the operational composition encoded by numerical expressions is impor-
tant for algebraic insight”; and Cusi, Malara, and Navarra attend to both canonical
and non-canonical forms of numbers in their work with teachers of early algebra.
Similarly, Britt and Irwin promote algebraic thinking in the form of generalizing
584 C. Kieran
relationships for operations with emphasis on relational and compensating opera-

tions, by means of student tasks such as: “Jason uses a simple method to work out
problems like 27 + 15 . . . in his head. Jason’s calculation is 30 + 12 = 42. Show
how to use Jason’s method to work out 298 + 57.”
Other aspects of numerical unpacking are presented in the chapter by Cai, Ng,
and Moyer who describe the Singaporean focus on ‘doing and undoing’ within the
relationships between addition and subtraction, and between multiplication and di-
vision. They also draw our attention to the Singaporean curricular emphasis on ‘ab-
stract strategies,’ which are clearly relational in nature. In a similar vein, but with
a focus that is as much on quantity as it is on number, Watanabe synthesizes the
Japanese course of study in mathematics at the elementary school level with its
quantitative relations strand and attention to the ‘writing and interpreting of mathe-
matical expressions.’
The notion that algebra is about insight into quantities and their relationships is
also reflected in the chapter by Subramaniam and Banerjee, who maintain that alge-
bra is not so much a generalization of arithmetic as it is a foundation for arithmetic
and who affirm that “arithmetic itself must be viewed with ‘algebra eyes’.” Britt and
Irwin, as well, argue that the origins of algebraic thinking precede understanding
of arithmetic and thus these researchers focus on developing such thinking in stu-
dents from their earliest years in school. The ultimate embodiment of this position
is found in the chapter by Schmittau. She first reminds us of Vygotsky’s assertion
that “the student who has mastered algebra attains ‘a new higher plane of thought,’
a level of abstraction and generalization that transforms the meaning of the lower
(arithmetic) level.” According to Schmittau, Davydov did not want students to wait
until the secondary level of schooling and so sought to introduce theoretical or alge-
braic thinking earlier in the school experience. Schmittau describes the way in which
students thereby begin the study of algebraic structure, even before they learn about
number, by means of a focus on the theoretical (quantitative) characteristics of real
objects.
While the stance of Schmittau is quite exceptional within this volume, much of
the research within the theme of relational thinking could be said to have its roots
in activity involving quantities. For example, Ellis states: “Quantities are attributes
of objects or phenomena that are measurable; it is our capacity to measure them—
whether we have carried out the measurements or not—that makes them quantities.”
Ellis, whose research is situated within a functional approach, argues further that a
focus on functional relationships between quantities, rather than on numbers discon-
nected from meaningful referents, can ground the study of algebra, and functions in
particular, in students’ experiential worlds.
The multiple ways in which the above chapters open up the ‘relational thinking’
perspective on early algebra contribute substantially toward counteracting the tra-
ditional view of arithmetic as being simply about number facts and algorithms for
number operations. Students who come to see number and its operations in terms of
their inherent structural relations, that is, as objects that can be compared relation-
ally in terms of their components, and who can use the fundamental properties of
operations and equality within the kinds of activities that are described in this vol-
ume, could be said to be seeing their arithmetic with ‘algebra eyes’. In high school
algebra, students are often called upon to look for relationships in symbolic expres-
sions in terms of underlying structure, such as for example, seeing x 6 − 1 both as
((x 3 )2 − 1) and as ((x 2 )3 − 1), and so being able to factor it in two ways (either as
a difference of squares or as a difference of cubes). Even if literal symbols are not
considered a constituent part of algebraic thinking within early algebra, it is clear
that the unpacking of quantity, number, and numerical operations and seeing such
unpacked objects in terms of their underlying structure has its parallels in the seeing
of relationships in literal expressions at the high school level.
Thinking Representationally about the Relations in Problem

Situations
A strongly held belief in algebra education is the notion that problem-solving con-
texts are foundational to algebraic activity. This stance is based to a certain extent on
historical grounds whereby algebra grew in status to become the privileged tool for
expressing general methods for solving whole classes of problems. However, the
difficulties that students experience in generating equations to represent the rela-
tionships found in word problems is well known (Kieran 2007). Thus, research that
leads to alternate forms of representation that both embody that which equations
represent as well as prove to be more accessible to students, in particular younger
students, is of great interest. Although much of the early algebraic activity related
to the already described relational-thinking frame involves, at least implicitly, story
problem contexts as opposed to purely numeric contexts, the Singaporean pictorial
equation (or model method, as it is sometimes called), presented in the chapter by
Cai, Ng, and Moyer, and also referred to by Watanabe, offers an analytic method
for dealing with indeterminates in the representing of relationships in a problem
situation—one that is well suited to the younger student.
The Singaporean approach, as described by Cai, Ng, and Moyer, focuses on the
use of pictorial equations so as “to analyze parts and wholes, generalize and specify,
and do and undo.” It is believed that, if children are provided with a means to visu-
alize a problem, they will come to see the structural underpinnings of the problem.
An example of the pictorial equation, which is drawn from the Cai, Ng, and Moyer
chapter, is provided in Fig. 1.
The authors point out that, as students move to the higher grades of elemen-
tary school, the pictorial equations are used to solve algebra problems involving
unknowns, emphasizing that the rectangles allow students to treat unknowns as if
they were knowns. To solve for the unknown, students undo the operations that are
implied by the pictorial equation. It is intended that pictorial equations provide a
smooth transition to the more abstract forms of equations with their literal-symbolic
notation that are encountered in the formal algebra of high school.
Another noteworthy approach to problem representation that is highlighted in the
chapters by Cai, Ng, and Moyer and by Li, Peng, and Song involves the combining
of various representations to encourage abstraction of central algebraic ideas. The
586 C. Kieran
Fig. 1 Pictorial equation, Raju and Samy shared $410 between them. Raju re-
drawn from the Cai, Ng, and ceived $100 more than Samy. How much money did
Moyer chapter Samy receive?
2 units = $410 - $100
= $310
1 unit = $155
Samy received $ 155
Chinese approach to developing algebraic thinking, which is described in both these

chapters, provides students with opportunities to represent a quantitative relation-
ship in a combination of different ways—an approach that the latter authors refer
to as “teaching teaching with variations with variation.” It is expected that students
will use both arithmetic and algebraic approaches (from Grade 5 onward), and com-
pare them. The authors suggest that the use of multiple approaches (which include
the arithmetic, algebraic, pictorial, as well as other approaches for other types of
situations) can foster a deeper understanding of the relationship between quantities,
as well as their representation.
Thinking Conceptually about the Procedural

High school algebra has traditionally been viewed as a domain of school mathemat-
ics that is dominated by the procedural and where the notion of a conceptual com-
ponent has been considered nothing short of an oxymoron. In his commentary on
the instructional part of this volume, Mason argues that “a blinkered, procedurally
oriented perspective on what school algebra is and could be inhibits and obstructs
the take up of a richer and broader vision of what school algebra could be, and as
far as I am concerned must be if mathematics education is going to develop.”
One of the central issues related to this widespread procedural orientation in
algebra has been the lack in the past of any significant forward movement with
respect to the question of that which might constitute the conceptual aspects of al-
gebraic procedures. However, recent theoretical perspectives (e.g., Artigue 2002;
Lagrange 2003) are offering a nuanced rethinking of the procedural in terms of the
conceptual. Artigue and Lagrange argue that the learning of procedures has within
itself a conceptual component. They point out that the technical activity of students,
during the period of elaboration of techniques, contains an epistemic (i.e., concep-
tual) element that is so intertwined with the technical that one co-develops with the
other. Examples (drawn from Kieran to appear) of conceptual understanding of al-
gebraic procedures include: being able to see a certain form in algebraic expressions
and equations (e.g., seeing that x 2 + 5x + 6 and x 4 + 7x 2 + 10 are both of the form
ax 2 + bx + c); being able to see relationships, such as the equivalence between

factored and expanded expressions; and being able to see through algebraic trans-
formations to the underlying change in form of the algebraic object and being able
to explain and justify these changes.
Many of the research studies described in this volume reflect implicitly these new
perspectives with their emphasis on the conceptual aspects of early algebra, as seen
for example, in their attention to the structural face of arithmetic operations, viewed
not just as procedures for calculation but also as relational objects. Such perspectives
are beginning to break down the old dichotomy between the procedural and the
conceptual by including a focus on the conceptual aspects of procedural operations.
However, as is seen below, the breaking down of this old dichotomy between the
procedural and conceptual brings with it some difficulties in naming and describing
approaches to the teaching of algebra that are primarily procedurally oriented.
For example, functional and so-called structural approaches to curricula for mid-
dle schoolers are compared in the chapter by Cai, Moyer, Wang, and Nie. The func-
tional approach is described as emphasizing the ideas of change and variation in
situations and contexts, as well as the representation of relationships between vari-
ables, while the ‘structural’ approach is described as avoiding contextual problems
so as to concentrate on working abstractly with symbols and following procedures in
a systematic way. The authors assert that this latter approach uses “naked equations
and [emphasizes] procedures for solving equations . . . all hallmarks of a structural
focus.” From their observations of classroom instruction, the authors report that the
teaching of the functional approach involved a much higher level of conceptual em-
phasis while the so-called structural approach involved a much higher level of pro-
cedural emphasis. In particular, they found that a larger percentage of high cogni-
tive demand tasks (procedures with connections) was implemented in the functional
approach classrooms, while a larger percentage of low cognitive demand tasks (pro-
cedures without connections or involving memorization) was implemented in the
structural approach classrooms.
However, Cai, Moyer, Wang, and Nie’s use of the term structural is at variance
with the way in which this term is used in other chapters of this volume. While Cai
et al. associate structural with a low-cognitive-demand procedural (i.e., algorithmic)
orientation, other authors tend to use the term within a more relational, conceptual
focus, for example: “algebraic structure emerges in young children’s reasoning and
can, with the help of the teacher, be made explicit” (Empson et al.); “pupils focus
. . . on relations, that is, on the structure of the sentence” (Cusi et al.); “the spe-
cific movement back and forth between these two representations, geometric and
numeric, ultimately supported students to gain not only flexibility with, but also a
structural sense of, two-part linear functions” (Moss and McNab); and “meaning
is encoded in the structure or relationships between the components” (Cooper and
Warren). The contrast between Cai et al.’s use of the term structural and the way
in which it is used by other authors in the same volume is but one example that
suggests a need for a more common terminology, but even more important is the
urgency to grapple with the meanings of, and relation between, the procedural and
the conceptual in early algebra.
588 C. Kieran
The conceptual versus the procedural is also the theme of a study reported by
Knuth, Alibali, Weinberg, Stephens, and McNeil, who compare relational thinking
with that which they describe as ‘operationally’-oriented thinking. They report that,
despite having learned within a function-based curriculum, only a minority of the
middle-school students that were tested demonstrated a relational understanding of
the equal sign. Knuth et al. thus recommend that the concept of equivalence be
given much more attention than it currently receives in the development of algebraic
thinking at both the elementary and middle school levels.
Comparison between relational and ‘procedural’ emphases in instruction also
constitutes the basis of the analysis in the chapter by Smith. She contrasts two 8th
grade lessons on the topic of simultaneous equations. Smith describes one lesson
as showing a procedural approach to the topic with students focusing on getting an-
swers through a series of routine steps. The other lesson emphasized building gener-
alized solution methods and understanding the relationships represented in systems
of equations. Smith notes that the latter lesson “shows how problems that appear
procedural can still be completed with conversations that provide rich mathematical
connections, allowing students to begin to connect the relations and generalizations
which characterize algebra.” This observation by Smith is an important one that
is consistent with the opening remarks of this section of my overall commentary:
the procedural can be approached in a conceptual manner. Similarly, the chapter
by Ellerton and Clements describes how procedures for solving decontextualized
linear and quadratic equations and inequalities can be conceptualized in connected,
relational ways. Both of these chapters, which offer a vision on how the learning
of so-called formal algebraic procedures can be rendered conceptual, touch upon an
area where further research is crucial, not only for high school algebra but also for
early algebra.
Anticipating, Conjecturing, and Justifying

Up to now in this commentary, the characterization of the nature and components
of algebraic thinking as reflected in the chapters of this volume has been the main
thrust. These characteristics have included thinking about the general in the particu-
lar, thinking rule-wise about patterns, thinking relationally about quantity, number,
and numerical operations, thinking representationally about the relations in problem
situations, and thinking conceptually about the procedural. Clearly, one of the main
routes to the development of such algebraic thinking is generalizing, a process that
was touched upon in an earlier section. In addition to generalizing, other routes to
the development of algebraic thinking that are emphasized within this volume in-
clude anticipating, conjecturing, explaining, and justifying. Still other chapters add
questioning, wondering, and discussing to this list.
With respect to the role of anticipating within algebraic activity, Boero (2001)
has elsewhere argued that:
A common ingredient of all the processes of transformation (without, before and/or after
formalisation) is anticipation. In order to direct the transformation in an efficient way, the
subject needs to foresee some aspects of the final shape of the object to be transformed re-
lated to the goal to be reached, and some possibilities of transformation. This ‘anticipation’
allows planning and continuous feed-back. (p. 99)
Reflecting this point of view, Empson, Levi, and Carpenter in their chapter empha-
size the role played by anticipation within relational thinking. As was described
briefly above, the example they provide of a student’s relational approach to adding
1/2 and 3/4 included anticipatory thinking. They argue further that the solution “in-
volved thinking flexibly about both the quantity 3/4 and about the operation, taken
into account concurrently rather than separately as a series of isolated steps.” An-
other perspective on anticipation and the role it can play is highlighted in the chapter
by Moss and McNab. In the report of their study of 2nd graders, they discuss how
the process of designing and presenting their own growing patterns to classmates
provided the students with the opportunity to anticipate how their classmates might
respond. According to Moss and McNab, this kind of anticipation and planning adds
an extra metacognitive dimension to students’ algebraic thinking, thereby enriching
the learning potential of the activity.
Conjecturing, generalizing, and justifying are central to the developing of alge-
braic thinking, according to Blanton and Kaput. In their chapter within this volume,
these authors suggest further that tasks ought not only to involve these processes but
also build upon systematic variation in the values of problem parameters: “Delib-
erately transform single-numerical-answer arithmetic problems to opportunities for
pattern building, conjecturing, generalizing, and justifying mathematical relation-
ships by varying the given parameters of a problem.” But how, Blanton and Kaput
ask rhetorically, does this transformation lead to algebraic thinking or, specifically,
functional thinking? They respond: “Varying a problem parameter enables students
to generate a set of data that has a mathematical relationship, and using sufficiently
large quantities for that parameter leads to the algebraic use of number.”
Other chapters that signal the importance of justifying within the development of
algebraic thinking include that of Russell, Schifter, and Bastable, who describe stu-
dents’ constructing of mathematical arguments to justify general claims for classes
of numbers. The authors point out that, although younger students lack the tools of
formal proof, they do have available to them ways of representing the operations—
drawings, models, or story contexts that they can use to represent specific numerical
expressions, but which can also be extended to model and justify general claims.
They argue specifically that the development of representations for the operations
is critical to connecting arithmetic and algebra. This is clearly an area that invites
further research—research on the ways in which operations might effectively be rep-
resented by drawings and models, and used as tools for justifying general claims,
within the context of early algebra.
Subramaniam and Banerjee, in a historical passage within their chapter, offer a
quote attributed to Bhaskara: “Mathematicians have declared algebra to be computa-
tion attended with demonstration: else there would be no distinction between arith-
metic and algebra.” The way in which Indian mathematicians in the past thought
about algebra provides, according to Subramaniam and Banerjee, the foundation
for the way in which the two of them conceptualize algebraic thinking in terms of
justification:
590 C. Kieran
Algebra involves taking a different attitude or stance with respect to computation and the
solution of problems, it is not mere description of solution, but demonstration and justi-
fication. Mathematical insight into quantitative relationships combined with an attitude of
justification or demonstration, leads to the uncovering of powerful ways of solving complex
problems and equations.
Both anticipation and justification are inherent to the theoretical frame presented by
Morselli and Boero in their chapter. These authors use Habermas’ theory of rational-
ity as a tool for analyzing students’ use of algebraic language in mathematical mod-
eling and proving. In their adaptation of Habermas’ construct of rational behavior,
the authors propose the following three dimensions of rational behavior: epistemic
rationality, which concerns both “coherency between the algebraic model and the
modeled situation” and the “manipulation rules of the system of signs”; teleological
rationality, which consists of the “transformations and interpretations that are useful
to the aims of the activity”; and communicative rationality, which includes “not only
communication with others (explanation of the solving processes, justification of
the performed choices, etc.) but also communication with oneself.” As the authors
point out, students may carry out certain operations correctly and thereby satisfy
the requirements of epistemic rationality; however, they may not have adequately
anticipated the aims of the activity and thereby do not satisfy the requirements of
teleological rationality. In addition, this model with its communicative-rationality
dimension allows for a focus on explanation and justification.
Other routes considered important in the fostering of algebraic thinking include
questioning and discussing. For example, Izsák, in his chapter on the complexity
of students’ thinking in the act of generating and interpreting problem representa-
tions, recommends that teachers elicit this student thinking and engage in classroom
conversations that include explicit comparisons of different approaches, thereby en-
couraging the emergence of more powerful algebraic representations. Other related
pedagogical interventions considered important by the authors for developing al-
gebraic thinking within problem-solving and problem-representation situations are
proposed in the chapter by Koellner, Jacobs, Borko, Roberts, and Schneider: posing
questions to move the students forward in their thinking, having students explain and
justify their own thinking, and probing more deeply into relevant and challenging
ideas.
However, just as little is known about the way that students generalize (Radford,
this volume), even less is known about the ways that students come to anticipate,
conjecture, and justify. The manner is which students’ engagement with these pro-
cesses leads to algebraic thinking is an area of research that could prove fruitful for
years to come.
Gesturing, Visualizing, and Languaging
Although generalizing has already been discussed in terms of being both a char-
acteristic of and a route to algebraic thinking, we return to it once more, even if
briefly—this time using the lens of gesturing, visualizing, and languaging, as sug-
gested in the chapters by Radford, by Moss and McNab, and by Cooper and Warren.
From his study on patterning with 2nd graders, Radford notes that the progression
in the grasping of the regularity within a pattern linked two kinds of components,
both a spatial and a numerical one. This link was mediated by a complex inter-
action of various senses, such as the visual, the motor, and the aural, as well as
by language and rhythm. As students began to think about larger figures, gestures
and words helped them to visualize these non-present figures. They generalized and
could produce both spatial descriptions of the unspecified figures and the sought-for
numerical totals by means of their calculators, even if the majority of the students
were not yet stating explicitly the operations being used in terms of unknown num-
bers. Throughout, both the teacher and the students made extensive use of gestures,
acting out, rhythm, and words—most of the senses, in fact. In a related way, the
study by Moss and McNab, which also involved 2nd graders in a patterning se-
quence, highlights the centrality of the visual in interaction with the numeric in
evoking students’ initial algebraic thinking. Similarly, the roles played by the ki-
naesthetic, the visual, and the verbal are underlined by Cooper and Warren in their
studies of generalization among 3rd to 5th graders.
Elsewhere, Radford (2010) has contrasted his view of the ‘sensuous’ nature of
thinking with that of a purely mental conception of thinking: “Thinking is consid-
ered a sensuous and sign-mediated reflective activity embodied in the corporeality
of actions, gestures, and artifacts . . . the adjective sensuous refers to a conception
of thinking that is inextricably related to the role that the human senses play in it.
Thinking is a versatile and sophisticated form of sensuous action where the various
senses collaborate in the course of a multi-sensorial experience of the world” (p. 4).
The cultural-semiotic lens that Radford brings to his analysis of the role played
by the senses in arriving at a pattern generalization in the context of early algebraic
thinking provides a valuable viewpoint on the process of generalization. This view
broadens considerably existing perspectives on the mental nature of the generalizing
process, opening up the construct to the consideration of factors that up to now have
largely been ignored, and so suggests an area for further research in the study of
algebraic thinking with younger students.
The View of Algebraic Thinking that Emerges from this Volume

The authors of the chapters in this volume provide support for their point of view
that algebra in elementary and early middle school is not all about literal symbols
but rather is about ways of thinking—thinking about the general in the particular,
thinking rule-wise about patterns, thinking relationally about quantity, number, and
numerical operations, thinking representationally about the relations in problem sit-
uations, and thinking conceptually about the procedural. The processes that consti-
tute these ways of thinking include generalizing, anticipating, conjecturing, justify-
ing, gesturing, visualizing, and languaging. The conceptual areas within early and
middle school mathematics that serve as the terrain for such thinking involve not
592 C. Kieran
the traditional content of high school algebra but rather the content of arithmetic,
including elements of function and change. However, the arithmetic being engaged
in is far removed from the usual fare of number facts, algorithms for number op-
erations, and single-numerical-answer problems. The emphasis is rather on seeing
within arithmetic not only its inherent regularities, equivalences, multiple ways of
conceptualizing numerical relations and analyzing and representing quantitative re-
lationships, but also its functional face involving patterning, analyzing how quan-
tities vary, and identifying correlations between problem variables. As Kilpatrick
points out in his commentary on the curricular part of the volume, “if curriculum is
a topic list, nothing changes; but if curriculum is the set of experiences that learners
have, then the change can be profound.”
An additional, but non-negligible, thread running through almost all the chap-
ters is that algebraic thinking does not develop unaided in students. The role of the
teacher is crucial. For example, Blanton and Kaput emphasize that “it requires an
‘algebra sense’ by which teachers can identify occasions in children’s thinking to ex-
tend conversations about arithmetic to those that explore mathematical generality”;
Radford, within a patterning context, points to the importance of the teacher asking
students to come up with an idea of how to find the total before using actual num-
bers, thereby encouraging the emergence of the generic aspects of the spatial config-
uration; and Russell, Schifter, and Bastable recount a teacher’s pivotal requesting of
her students to “make a picture, draw a model, but not use any particular numbers.”
Sriraman and Lee remark in their commentary on the cognitive part of the volume
that “algebraic thinking can be cultivated from the early grades on if teachers are
cognizant of non-symbolic modes of reasoning.” The examples that are provided
throughout the volume of the ways in which teachers are instrumental in assisting
their students to come to think algebraically about their arithmetic point to the com-
plexity of being “cognizant of non-symbolic modes of reasoning.” It involves being
cognizant of not only the characteristics and components of algebraic thinking, as
well as the centrality of certain process-related routes to the development of such
thinking, but also novel approaches to tasks, forms of questioning, key examples to
focus on, appropriate ways of reacting to students’ responses, and a manner of capi-
talizing on students’ contributions so as to help make them accessible to the class at
large. As has been emphasized several times throughout this volume, students learn
to see algebraically because appropriate learning environments have been designed
and put into place according to specific mathematical and pedagogical ideas. De-
spite the considerable advances that have been made in this field of early algebra, as
reflected in the chapters of this volume, much still remains to be done.
References
Artigue, M. (2002). Learning mathematics in a CAS environment: The genesis of a reflection about
instrumentation and the dialectics between technical and conceptual work. International Jour-
nal of Computers for Mathematical Learning, 7, 245–274. doi:10.1023/A:1022103903080.
Balacheff, N. (1987). Processus de preuve et situations de validation [Proving processes and vali-
dation situations]. Educational Studies in Mathematics, 18, 147–176.
Blanton, M. L., & Kaput, J. J. (2008). Building district capacity for teacher development in alge-
braic reasoning. In J. J. Kaput, D. W. Carraher, & M. L. Blanton (Eds.), Algebra in the Early
Grades (pp. 361–388). New York: Routledge.
In R. Sutherland, T. Rojano, A. Bell, & R. Lins (Eds.), Perspectives on School Algebra (pp. 99–
119). Dordrecht, The Netherlands: Kluwer Academic.
Jeannotte, D. (2010). L’apport pour la formation et la pratique enseignante: analyse et synthèse de
différents modèles de raisonnement mathématique dans la littérature scientifique [Contributing
to teacher training and teaching practice: Analysis and synthesis of different models of mathe-
matical reasoning in the scientific literature]. In V. Freiman (Ed.), Actes du congrès Groupe des
Didacticiens des Mathématiques 2010. Moncton, New Brunswick, Canada: Comité scientifique
GDM.
Kaput, J. J. (2008). What is algebra? What is algebraic reasoning. In J. J. Kaput, D. W. Carraher,
& M. L. Blanton (Eds.), Algebra in the Early Grades (pp. 5–17). New York: Routledge.
Kieran, C. (1996). The changing face of school algebra. In C. Alsina, J. Alvarez, B. Hodgson,
C. Laborde, & A. Pérez (Eds.), Eighth International Congress on Mathematical Education:
Selected Lectures (pp. 271–290). Seville: S.A.E.M. Thales.
Kieran, C. (2004). Algebraic thinking in the early grades: What is it? The Mathematics Educator,
8(1), 139–151.
Kieran, C. (2007). Learning and teaching algebra at the middle school through college levels:
Building meaning for symbols and their manipulation. In F. K. Lester Jr. (Ed.), Second Hand-
book of Research on Mathematics Teaching and Learning (pp. 707–762). Charlotte, NC: Infor-
mation Age Publishing.
Kieran, C. (to appear). The false dichotomy in mathematics education between conceptual under-
standing and procedural skills: An example from algebra. In K. Leatham (Ed.), Vital Directions
in Mathematics Education Research. New York: Springer.
Lagrange, J.-B. (2003). Learning techniques and concepts using CAS: A practical and theoretical
reflection. In J. T. Fey (Ed.), Computer Algebra Systems in Secondary School Mathematics
Education (pp. 269–283). Reston, VA: National Council of Teachers of Mathematics.
Mason, J. (1996). Expressions of generality and roots of algebra. In N. Bednarz, C. Kieran, &
L. Lee (Eds.), Approaches to Algebra (pp. 65–86). Dordrecht, The Netherlands: Kluwer Aca-
demic.
Mason, J., with Graham, A., & Johnston-Wilder, S. (2005). Developing Thinking in Algebra. Lon-
don: Sage Publications.
Radford, L. (2010). Signs, gestures, meanings: Algebraic thinking from a cultural semi-
otic perspective. In V. Durand-Guerrier, S. Soury-Lavergne, & F. Arzarello (Eds.),
Proceedings of the 6th Congress of the European Society for Research in Math-
ematics Education (2009). Lyon, Fr: Institut National de Recherche Pédagogique.
http://www.inrp.fr/editions/editions-electroniques/cerme6/plenary-1.
Author Index
A Bankov, K., 380, 408

Ageyev, V., 567, 576 Bardini, C., 314, 321
Ainley, J., 486, 507, 564, 574 Barker, D., 278, 300, 341, 364
Alatorre, S., 304, 310, 320, 321, 329, 336, 340, Barkin, S., 9, 22, 279, 300
346, 356, 363, 365 Barneveld, G. V., 8, 22
Alcock, L., 459, 481 Bartolini Bussi, M. G., 469, 479
Alibali, M., 248, 255, 257, 274, 276 Bastable, V., 44, 46, 49, 67, 68, 69, 147, 148,
Alibali, M. W., 122, 123, 261, 273, 274, 276, 158, 260, 275
512, 527 Battey, D., 411, 427
Allen, F. B., 381, 407
Baturo, A. R., 188, 195, 211, 212
Almog, N., 381, 389, 408
Bazzini, L., 381, 388, 389, 407, 408, 460, 479,
Alsina, C., 173, 184, 485, 509, 580, 593
485, 507
Alvarez, J., 485, 509, 580, 593
Beatty, R., 9, 22, 278, 279, 298–300, 304, 321,
Alvarez, J. M., 173, 184
567, 568, 574, 576
Amit, M., 279, 298
Becker, J., 304, 320, 336, 354, 365
Anghileri, J., 488, 507
Anichini, G., 469, 479 Becker, J. R., 154, 158, 244, 258, 278, 301,
Arafeh, S., 516, 526 324, 329, 336, 341, 347, 363, 365
Araujo, C., 226, 237 Bednarz, N., vii, x, 162, 165, 173, 183, 184,
Arcavi, A., 243, 254, 258, 460, 479, 485, 507, 218, 220, 235, 236, 260, 275, 278, 279,
512, 526 291, 295, 300, 315, 318, 320, 321, 329,
Archambault, R., 410, 427 341, 364, 365, 459, 480, 514, 528, 582,
Artigue, M., 586, 592 593
Arzarello, F., 314, 320, 460, 469, 479, 485, Behr, M. J., 9, 21, 31, 41, 64, 68, 189, 211,
507, 591, 593 242, 257, 410, 411, 426, 427, 428, 534,
Ausubel, D. P., 532, 555 555
Awtry, T., 103, 106 Bell, A., 38, 40, 318, 321, 343, 346, 366, 368,
370, 372, 373, 485, 507, 588, 593
B Bennett, J., 557, 558, 573, 574
Baek, J., 413, 425, 426 Berliner, D. C., 426, 427
Bagni, G. T., 371, 372 Bernardz, N., 38, 40
Bailey, R., 164, 166, 168, 183 Biggs, J. B., 530, 555
Balacheff, N., 315, 320, 370, 372, 457, 479, Biggs, J. G., 530, 555
582, 592 Bishop, A., 193, 194, 212, 329, 364
Ball, D. L., 430, 431, 451 Bishop, J., 341, 347, 364
Banerjee, R., 89, 98, 102, 103, 105, 105, 106 Blair, W., vii, xi, 216, 236, 260, 276

DOI 10.1007/978-3-642-17735-4, © Springer-Verlag Berlin Heidelberg 2011
596 Author Index
Blanco, L. J., 382, 384, 386, 407 C

Blanton, M. L., 7–10, 12, 14, 17–19, 21, Çağlayan, G., 247, 249, 254, 256
21–23, 44, 57, 68, 69, 111, 123, 138, Cai, J., viii, x, 26, 27, 29, 30, 36–39, 40, 89,
147, 148, 154, 156, 157, 158, 216, 226, 106, 111, 123, 162, 164, 165, 170, 172,
230, 237, 260, 275, 276, 298, 299, 303, 174, 176, 183, 184, 298, 299, 484, 486,
320, 430, 431, 451, 452, 486, 488, 507, 507, 529, 530, 532, 548, 555, 556
508, 513, 514, 526, 527, 563, 567, 568, Calfee, R. C., 426, 427
575, 576, 580, 593 Callebaut, W., 371, 372
Bloedy-Vinner, H., 193, 211 Cannizzaro, L., 371, 372
Blömeke, S., 380, 408 Carlson, M., 215, 218, 219, 223, 224, 226, 235
Blum, W., 459, 480 Carlson, M. P., 218, 235
Boaler, J., 563, 574 Carpenter, T. P., 7–9, 16, 18, 19, 22, 26, 40, 44,
Bobis, J., 461, 481 64, 68, 162, 173, 183, 188, 191, 207,
Boero, P., 90, 99, 106, 381, 407, 454–458, 477, 213, 218, 237, 242, 258, 260, 261, 273,
479, 480, 481, 566, 574, 588, 593 274, 275, 276, 410, 411, 413, 424–426,
426–428, 486, 507, 508, 511–514, 526,
Booker, G., 485, 486, 508–510
527, 561, 570, 574
Booth, L. R., 241, 256, 387, 407, 485, 507
Carr, W., 453, 480
Booth, S., 565, 575
Carraher, D. W., viii, x, 7–9, 12, 21, 21–23, 26,
Borasi, R., 488, 507 36, 41, 44, 68, 69, 138, 139, 147, 148,
Borko, H., 431, 435, 436, 441, 442, 446, 451, 154, 157, 158, 190, 192, 212, 214, 216,
452 218, 220, 226, 230, 235, 237, 244, 256,
Bose, A., 92, 106 260, 275, 275, 276, 278, 279, 282, 289,
Bourbaki, N., 73, 85 297, 299, 301, 303, 317, 320, 430, 452,
Boyd, B. A., 6, 23 486, 508, 512, 513, 525, 527, 561, 563,
Boyer, C., 8, 21 567, 568, 574–576, 580, 593
Brader-Araje, L., 260, 276 Carver, S., 280, 296, 299
Bransford, J., 280, 300, 568, 576 Case, R., 279, 280, 296, 299, 300
Bransford, J. D., 426, 426 Cassundé, M. A., 226, 237
Breda, A., 127, 130 Castro, E., 488, 510, 526, 528
Brenwald, S., 515, 527 Cavanagh, M., 278, 299
Brinker, L., 423, 426 Cedillo, T., 380, 408
Britt, M. S., 138, 146–149, 152, 153, 157 Chaiklin, S., 96, 106
Brizuela, B. M., 9, 12, 21–23, 138, 139, 157, Chalouh, L., 173, 184
190, 192, 212, 214, 218, 220, 226, 237, Chavajay, P., 162, 185
244, 256, 275, 275, 289, 301, 303, 304, Chazan, D., 173, 183, 218, 222, 235, 512, 527
317, 320, 508, 568, 574, 576 Chen, Z., 530, 555
Brousseau, G., 572, 574 Cheong, I. P., 383, 384, 393, 401, 406, 407
Brown, A. L., 426, 426 Chevallard, Y., 459, 480
Brown, J. S., 249, 250, 257, 410, 428 Chiappini, G., 460, 479, 485, 507
Chick, H. L., vii, x, xi, 7, 22, 26, 41, 91, 94, 95,
Brown, R., 192, 213
98, 106, 156, 157, 188–190, 192, 193,
Brownell, W. A., 368, 372
197, 198, 203, 206, 208, 212, 214, 314,
Bruce, C., 279, 299
321, 329, 336, 363, 486, 495, 507–510,
Bruner, J. S., 162, 183, 193, 212 512, 513, 527, 528, 530, 532, 556, 559,
Bruzuella, B., 279, 299 575
Bryant, R., 218, 237 Choppin, J., 56, 68
Bryant, R. L., 56, 68 Chukwu, O., 383, 384, 393, 401, 406, 407
Bryk, A. S., 176, 185 Ciarrapico, L., 469, 479
Buck, J. C., 222, 235 Cifarelli, V., 192, 204, 212
Bunning, K., 431, 435, 452 Clark, K. K., 431, 435, 441, 442, 446, 451
Burns, R. B., 195, 212 Clarke, B., 509
Burrill, G., 240, 256 Clarkson, P., 190, 192, 197, 204, 214, 383,
Bussi, M. G. B., 221, 235, 371, 373 384, 393, 408
Author Index 597
Clement, J., 66, 68, 241, 254, 256, 387, 388, de Corte, E., 559, 577
407 de Jong, T., 514, 527
Clements, M. A., 381, 383–389, 393, 395, 400, De la Torre, E., 488, 510
401, 403, 404, 406, 407, 408, 568, 575 De Villiers, M., 454, 457, 468, 480, 481
Cnop, I., 430, 451 DeLoache, J. S., 424, 428
Cobb, C., 430, 452, 563, 576 DeStefano, L., 128, 129
Cobb, P., 140, 158, 190, 212, 222, 225, 236, Detterman, D. K., 162, 183
435, 451, 485, 486, 488, 508–510 Dewan, H. K., 90, 106
Cockburn, A., 564, 574 Dewey, J., 410, 427
Cockburn, A. D., 388, 389, 408, 430, 452, 486, Dhindsa, H. S., 383, 384, 393, 401, 406, 407
507 Dienes, Z. P., 188, 190, 212, 370, 372
Cocking, R. R., 173, 177, 183, 426, 426 diSessa, A. A., 243–245, 254, 256, 258, 435,
Coe, E., 215, 219, 235 451
Cogan, L., 380, 408 Dobbs, D., 382, 400, 407
Cohen, D. K., 431, 451 Dominguez, H., 410, 413, 415, 427
Colebrooke, H. T., 92–94, 106 Donovan, S., 280, 300, 568, 576
Colliins, A. M., 426, 427 Doorman, L. M., 332, 364
Confrey, J., 8, 22, 190, 195, 212, 218, 222, Dorfler, W., 278, 299
225, 235–238, 435, 451 Dörfler, W., 327, 329, 335, 364
Cooney, T., 215, 219, 236 Douek, N., 454, 460, 477, 479, 480
Cooper, T. J., 188–190, 192, 193, 195, 197, Dougherty, B. J., 9, 21, 123, 123, 138, 154,
198, 203, 204, 209, 211, 211, 212, 214, 157, 190, 192, 205, 212, 218, 226, 235,
244, 258, 279, 299, 304, 322 237, 279, 282, 299, 303, 320, 324, 329,
Copes, L., 222, 235 365, 388, 407, 486, 488, 507–509, 567,
Cortina, J., 329, 340, 346, 356, 365 575
Cortina, J. L., 304, 310, 320, 321, 336, 363, Dougherty, G., 204, 214
365 Douglas, E. C., 381, 407
Coxford, A. F., 6, 23, 31, 41, 111, 124, 188, Dowker, A. D., 423, 427
214, 240, 256, 262, 276 Downton, A., 383, 384, 393, 408
Cramer, K., 14, 22 Doyle, W., 182, 183
Cuoco, A. A., 9, 22, 173, 183 Dremock, F., 7–9, 18, 19, 22
Curcio, F. R., 9, 22, 26, 41 Dretske, F., 329, 364
Cusi, A., 508 Dreyfus, T., 187, 188, 191, 193, 212, 218, 221,
Czarnocha, B., 501, 508, 509 236, 238, 327, 364, 457, 461, 462, 480,
488, 510
D Driscoll, M., 34, 41, 432, 434, 447, 451
da Ponto, J. P., 220, 237 Dubinsky, E., 8, 23, 56, 68, 215, 218, 219, 236,
Da Rocha Falcão, J. T., 486, 508 238, 242, 257, 458, 480
Dacey, L., 278, 299 Dubsinsky, E., 193, 213
Damon, W., 343, 364 Durand-Guerrier, V., 591, 593
Dapueto, C., 455, 480 Dürr, C., 567, 575
Darling-Hammond, L., 431, 451 Duval, R., 187, 188, 191, 193, 207, 212, 331,
Daro, P., 128, 129 341, 364, 458, 478, 480
Datta, B., 93–95, 106 Dweck, C., 562, 575
David, M. M., 488, 507
Davidov, V., 567, 575 E
Davis, B., 573, 575 Earnest, D., 9, 21, 138, 139, 157, 190, 192,
Davis, R. B., 195, 212 212, 226, 237, 244, 256, 279, 282, 299,
Davydov, V. V., 71, 73, 78, 80, 83, 85, 126, 303, 317, 320, 568, 574
129, 192, 205, 209, 210, 212, 356, 364 Economopoulos, K., 44, 68, 222, 236
Dawson, S., 488, 510 Edwards, L., 314, 320
Day, R., 164, 166, 168, 183 Eiteljorg, E., 431, 435, 436, 441, 451, 452
de Castro, B., 380, 406, 407 Elkonin, D. B., 72, 85
598 Author Index
Ell, F., 144, 149, 157, 159 Frykholm, J. A., 431, 435, 441, 451
Ellerton, N., 568, 575 Fuglestad, A. B., 381, 389, 407, 488, 489, 509,
Ellerton, N. F., 383, 384, 386, 387, 393, 395, 510, 515, 527
401, 403, 404, 406, 407, 408 Fuglestad, A.-B., 567, 575
Ellis, A., 329, 364 Fujii, T., 91, 95, 96, 98, 106, 123, 123, 138,
Ellis, A. B., 216, 219, 220, 223, 234, 236 152, 157, 193, 197, 203, 208, 212, 318,
Empson, S. B., 410, 411, 413, 415, 417, 424, 320, 388, 407, 513, 527
427 Fung, A. F., 548, 556
English, L., 188, 191, 192, 195, 206, 208, 212, Furinghetti, F., 312, 319, 321, 370, 371, 372,
220, 236, 340, 343, 364, 367, 370, 371, 373, 400, 408, 460, 480, 485, 507, 508
373, 454, 477, 480, 488, 497, 509 Fuson, K. C., 102, 106
English, L. D., 162, 184, 194, 212, 278, 279,
299 G
Erlwanger, S., 64, 68, 189, 211, 534, 555 Gadowsky, K., 221, 238
Evans, K. M., 381, 407 Gallimore, R., 417, 427
Even, R., 430, 451 Garcia, R., 371, 373
Garcia-Cruz, J. A., 329, 364
F Gardner, H., 162, 183, 566, 575
Falkner, K. P., 261, 273, 276, 410, 427, 512, Garnier, H., 417, 427, 450, 452, 515, 527
527 Garrote, M., 382, 384, 386, 407
Fan, L., 529, 530, 532, 555, 556
Gattegno, C., 565, 567, 575
Farenga, S. J., 218, 222, 236
Gearhart, M., 415, 428
Feng, J. H., 380, 408
Gelman, R., 343, 344, 364
Fennema, E., 6, 19, 22, 23, 26, 41, 173, 184,
Gentner, D., 191, 212
190, 213, 215, 218, 219, 236, 237, 242,
Gerhard, S., 567, 575
258, 368, 373, 413, 424, 426, 426–428,
Gerofsky, S., 559, 575
570, 574
Gillings, R., 565, 575
Fernández, M., 18, 21
Fernandez, M., 303, 320 Gindis, B., 567, 576
Ferrara, F., 314, 321 Ginsburg, 278, 299
Ferrari, P. L., 454, 477, 480 Givvin, K. B., 417, 427, 450, 452, 515, 516,
Ferreri, S., 486, 508 527
Ferrini-Mundy, J., 278, 299, 450, 452 Glaser, R., 243, 254, 258, 423, 428
Fey, J. T., 163, 167, 184, 240, 250, 251, 256, Godino, J. D., 56, 68
257, 586, 593 Goldenburg, P., 173, 183
Figueras, O., 102, 106, 336, 363, 365 Goldin, G., 9, 22
Filloy, E., 139, 157, 193, 194, 212, 310, 318, Gómez, J. C., 308, 320
320, 367, 373, 459, 480, 485, 508 Gonzales, P., 514, 515, 527
Findell, B., ix, xi, 240, 249, 256, 257, 426, 427 Gonzales, P. A., 514, 515, 528
Findell, C., 278, 299 Goodrow, A., 226, 237
Fiori, A., 371, 372 Goodson-Espy, T., 192, 212
Fischer, K. W., 162, 185 Gorbov, S. F., 71, 80, 83, 85
Fitzgerald, W., 240, 250, 251, 257 Gould, S. J., 371, 373
Fitzgerald, W. M., 163, 167, 184 Gowar, N., 565, 569, 576
Fonzi, J., 488, 507 Graça Martins, M. E., 127, 130
Franke, M. L., 16, 19, 22, 26, 40, 44, 64, 68, Graham, A., 147, 153, 158, 329, 365, 564, 565,
162, 173, 183, 260, 274, 275, 411, 413, 569, 576
425, 427, 486, 507, 508, 511, 526, 561, Grandau, L., 274, 276
574 Grandsard, F., 430, 451
Freiman, V., 581, 593 Grant, E. J., 188, 212
Freudenthal, H., 8, 22 Gravemeijer, K., 244, 245, 254, 256, 563, 575
Frey, P., 164, 166, 168, 183 Gravemeijer, K. P. E., 332, 364
Friel, S. N., 163, 167, 184, 240, 250, 251, 257 Gray, E., 485, 508
Frykholm, J., 431, 435, 452 Green, L. J., 381, 407
Author Index 599
Greenes, C. E., vii, x, 9, 22, 96, 106, 110, 111, Herstein, I. N., 411, 427
113, 122, 123, 124, 125, 129, 173, 183, Heuvel-Panhuizen, M., 9, 23
278, 279, 299, 300, 318, 320, 413, 425, Hewitt, D., 138, 153, 157, 295, 299, 565, 567,
426, 512, 513, 526, 527 575
Greeno, J. G., 426, 427, 431, 452, 558, 575 Hiebert, J., 182, 183, 188, 191, 207, 213, 411,
Greer, B., 271, 276, 368, 373, 559, 577 417, 426, 427, 428, 450, 452, 514–516,
Grevholm, B., 509 527, 528
Griffin, S., 280, 299 Higgins, J., 144, 149, 157, 159
Grinstead, P., 216, 220, 236 Hirabayashi, I., 220, 236, 485, 507
Gronn, D., 383, 384, 393, 408 Hironaka, H., 114, 123
Grossman, P., 432, 452 Hirsch, C., 240, 256
Grouws, D. A., viii, x, 39, 41, 102, 106, 188, Hitt, F., 187, 191, 193, 207, 212
191, 207, 213, 218, 236, 240, 257, 260, Hoch, M., 461, 480
261, 271, 276, 278, 299, 369, 373, 426, Hodgson, B., 173, 184, 485, 509, 580, 593
427, 485, 508, 512, 525, 526, 527, 530, Hoey, B., 247, 258
556 Hoines, M. J., 279, 298, 299, 300, 381, 389,
Grover, B. W., 168, 171, 182, 185, 570, 576 407, 488, 489, 509, 510, 515, 527, 567,
Gu, L., 530–532, 555 575
Guimarães, F., 127, 130 Hollerback, J., 329, 364
Guimarães, H. M., 127, 130 Holligsworht, H., 516, 527
Gutierrez, A., 90, 99, 106 Hollingsworth, H., 417, 427, 450, 452, 515,
Guzmán, J. C., 514, 515, 527 527
Holyoak, K., 347, 365
H Horne, M., 383, 384, 393, 408
Habermas, J., 454, 480 Houang, R., 380, 408
Hadas, N., 488, 510 Houang, R. T., 128, 130
Hadjidemetriou, C., 430, 452 Howard, A. C., 164, 166, 168, 183
Haeckel, E., 370, 373 Hoyles, C., 278, 279, 297, 299, 300
Halevi, T., 221, 236
Hsieh, 454, 457, 481
Halford, G., 188, 191, 192, 206, 208, 212
Hsu, E., 215, 219, 235
Halford, G. S., 188, 191, 213
Huang, R., 529, 530, 532, 555, 556
Hall, R., 243, 256
Huberman, A. M., 440, 452
Hamley, H. R., 8, 22
Huinker, D., 111, 123
Hammer, D., 244, 254, 256
Husen, T., 529, 556
Hammerness, K., 432, 452
Hutchens, D. T., 164, 166, 168, 183
Hanna, G., 454, 457, 468, 480, 481
Hwang, S., 36, 40
Hanson, N., 573, 575
Harel, G., 8, 23, 56, 68, 192, 193, 198, 203,
208, 213, 218, 219, 222, 236–238, 458, I
480 Iaderosa, R., 96, 106, 486, 509
Harper, E., 241, 256, 485, 508 Inhelder, B., 411, 428, 566, 576
Hart, E., 240, 256 Irwin, K. C., 138, 139, 144, 146–149, 152,
Hart, K., 485, 509 153, 157, 159
Hart, K. M., 241, 257, 387, 407 Iwasaki, H., 329, 364
Hartnett, P., 343, 344, 364 Izsák, A., vii, x, 110, 111, 122, 123, 125, 129,
Hatrup, R. A., 140, 158 218, 237, 244–247, 249, 254, 256, 424,
Healy, L., 278, 297, 299, 300 427
Henningsen, M. A., 168, 171, 182, 185, 570,
576 J
Herbert, K., 192, 213 Jacobs, J., 417, 427, 431, 435, 436, 441, 442,
Herscovics, N., 139, 157, 188, 189, 213, 534, 446, 450, 451, 452
541, 555, 556 Jacobs, J. K., 431, 435, 441, 451, 515, 516, 527
Hershkovitz, R., 330, 364 Jacobs, S., 215, 219, 235
Hershkowitz, R., 222, 237, 274, 276, 488, 510 Jacobs, V., 411, 413, 427
600 Author Index
Jacobson, M. J., 223, 237 300, 315, 318, 320, 321, 329, 341, 364,
James, W., 566, 575 365, 380, 381, 383, 388, 389, 391, 407,
Janvier, B., 459, 480 430, 431, 452, 459, 480, 484, 485, 508,
Janvier, C., 242, 243, 254, 256, 257, 410, 426, 509, 512, 515, 525, 526, 527, 530, 556,
427 562, 575, 580, 582, 585, 586, 593
Jaquet, F., 486, 509 Kieran, E., 57, 68
Jaworski, B., 488, 489, 509, 510 Kieran, E. J., 56, 68
Jeannotte, D., 581, 593 Kieran, K., 485, 508
Jocelyn, L., 514, 515, 527 Kieren, T., 145, 158
Johanning, D., 278, 301, 341, 342, 360, 366 Kieren, T. E., 145, 158
Johnston-Wilder, S., 147, 153, 158, 329, 365, Kilpatrick, J., ix, x, 8, 16, 23, 110, 111, 122,
564, 565, 569, 576 123, 125, 126, 128, 129, 130, 188, 189,
Jones, G., 162, 184 213, 222, 237, 240, 257, 369, 373, 426,
Jones, G. A., 371, 373 427, 431, 452, 469, 479
Jonsen Hoines, M., 298, 299 Kirschner, P., 514, 527
Junk, D., 410, 413, 415, 427 Kirshner, D., 103, 106
Klahr, D., 280, 296, 299
K Knoll, S., 514, 515, 528
Kaijser, S., 312, 319, 321 Knuth, E., viii, x, 26, 40, 122, 123, 162, 183,
Kaiser, G., 367, 373 243, 248, 255, 257, 274, 276, 298, 299,
Kalchman, M., 280, 296, 299 330, 364
Kaldrimidou, M., 92, 106, 354, 365, 529, 530, Knuth, E. J., 512, 527
556 Ko, P. Y., 530, 532, 556
Kanselaar, G., 514, 527 Koedinger, K. R., 430, 452
Kaput, J. J., vii, xi, 6–9, 11, 12, 14, 17–19, 21, Koehler, J., 410, 424, 426, 427
21–23, 26, 41, 44, 56, 66, 68, 69, 111, Koellner, K., 431, 435, 436, 441, 452
123, 138, 147, 148, 154, 156, 157, 158, Kolpakowski, T., 244, 254, 256
173, 184, 190, 193, 213, 215, 216, 218, Kosslyn, S. M., 329, 364
219, 223, 226, 230, 236–238, 242, 243, Koyama, M., 275, 275, 278, 301, 508
254, 257, 260, 275, 276, 298, 299, 303, Kozma, R. B., 223, 237
320, 368, 373, 430, 431, 451, 452, 458, Kozulin, A., 567, 576
480, 485, 486, 507, 508, 512–514,
Krabbendam, H., 8, 22
526–528, 563, 567, 568, 575, 576, 580,
Kratka, M., 188, 212
593
Kratká, M., 278, 301
Karp, A., 380, 382, 407
Krátká, M., 304, 321, 322, 336, 341, 347, 363
Kasimatis, E., 247, 258
Krátkná, M., 304, 320
Kastberg, D., 514, 515, 527
Krill, D., 274, 276
Katz, V., 92, 93, 106, 107
Krutetskii, V. A., 188, 213
Katz, V. J., 94, 106, 162, 184
Kawanka, T., 514, 515, 528 Ktorza, D., 261, 276
Kelly, A. E., 190, 195, 212 Kuchemann, D. E., 241, 257, 262, 274, 276,
Kemmis, S., 453, 480 278, 299, 387, 407, 485, 509
Kendal, M., vii, xi, 26, 41, 512, 513, 527, 528 Kuhn, D., 343, 364
Kenney, P. A., 26, 41 Kuiper, J., 96, 106
Kerslake, D., 241, 257, 410, 413, 427 Kyeleve, I. J., 383, 384, 393, 401, 406, 407
Khng, K., 525, 527 Kysh, J., 247, 258
Khng, K. H., 525, 527
Kho, T. H., 32, 41 L
Kibler, D., 243, 256 Laborde, C., 173, 184, 485, 509, 580, 593
Kieran, C., vii, x, 26, 37–39, 40, 41, 64, 68, 90, Lacampagne, C., vii, xi, 216, 236, 260, 276
99, 106, 111, 122, 123, 162, 165, 173, Lachance, A., 190, 195, 212
183, 184, 188, 189, 213, 218, 220, 235, Ladson-Billings, G., 260, 276
236, 240, 241, 248, 255, 257, 260, 261, Lagrange, J., 279, 299
273, 275, 276, 278, 279, 291, 295, 299, Lagrange, J.-B., 586, 593
Author Index 601
Lambdin, D. V., viii, xi Linchevski, L., 96, 102, 106, 139, 157, 188,
Lampert, M., 191, 213, 432, 452 189, 213, 218, 237, 278, 297, 301, 341,
Landau, M., 9, 21 366, 400, 408, 461, 480, 485, 509, 534,
Langrall, C., 162, 184, 341, 342, 347, 366 541, 555, 556
Langrall, C. W., 514, 528 Linn, R., 128, 129
Lannin, J., 190, 192, 197, 213, 244, 257, 278, Lins, R., 7, 22, 318, 321, 343, 346, 366, 368,
300, 330, 341, 347, 364 370, 372, 373, 512, 513, 528, 588, 593
Lannin, J. K., 279, 285, 291, 297, 299 Lins, R. C., 485, 509
Lapointe, A. E., 529, 556 Littweiller, B., 568, 576
Lappan, G., 163, 167, 184, 240, 250, 251, 257, Livneh, D., 102, 106, 461, 480
278, 299 Lo, M. L., 530, 532, 556
Lara-Roth, S., 226, 237 Lobato, J., 220, 236, 329, 364
Larsen, S., 215, 219, 235 Lochhead, J., 66, 68
Laughlin, C., 170, 184 London McNab, S., 279, 300
Lave, J., 162, 184, 431, 452 Loveless, T., vii, xi, 128, 130
Lawson, J., 308, 321 Lubienski, S., 430, 451
Lee, J., 79, 85
Lee, K., 36, 41 M
Lee, L., vii, x, 38, 40, 162, 165, 173, 183, 184, Ma, F., 541, 556
218, 220, 235, 236, 260, 275, 278, 279, Macedo, S., 226, 237
288, 291, 295, 300, 315, 318, 320, 321, MacGregor, M., 99, 107, 147, 158, 192, 193,
329, 341, 364, 365, 459, 480, 485, 509, 214, 220, 236, 237, 262, 276, 278, 288,
582, 593 301, 317, 320, 343, 346, 364, 366, 388,
408, 459, 478, 481, 513, 526, 528
Lee, S., 529, 556
Maggio, M., 486, 508
Lehrer, R., 244, 245, 254, 256, 424, 426, 426,
Maher, C. A., 195, 212, 341, 365
435, 451
Malara, N. A., 96, 106, 459, 481, 486, 488,
Leinhardt, G., 221, 236, 241, 249, 257
495, 497, 501, 508, 509
Leinhardt, L. G., 140, 158
Malisani, E., 525, 528
Lemons-Smith, S., 170, 176, 183, 184
Malloy, C. E., 260, 276
Lesgold, S., 96, 106
Mammana, C., 330, 341, 364
Lesh, R. A., 9, 21, 31, 41, 190, 195, 212, 242, Mariotti, M. A., 221, 235
257, 371, 373, 410, 426, 427 Mark, J., 173, 183
Lester, F., 26, 36, 41, 56, 68, 240, 257, 561, Markman, A. B., 191, 212
574 Martin, G., 56, 68
Lester, F. K., viii, xi, 317, 320, 380, 381, 383, Martin, H. T., 424, 427
388, 389, 391, 407, 512, 525, 526, 527, Martin, L., 530, 556
562, 575 Martin, W. G., 8, 16, 23, 222, 237, 431, 452
Lester, F. K. Jr., viii, xi Martinez, M., 244, 256, 278, 279, 297, 299
Lester Jr., F. K., 430, 431, 452, 585, 593 Martino, A. M., 341, 365
Leung, F. K. S., 529, 556 Martinón, A., 329, 364
Levi, L., 16, 19, 22, 26, 40, 44, 64, 68, 162, Marton, F., 530, 532, 555, 556, 565, 575
173, 183, 260, 261, 273, 274, 275, 276, Maslow, A., 557, 575
410, 411, 413, 417, 424, 425, 427, 486, Mason, J., 21, 22, 138, 147, 153, 158, 173,
508, 511–514, 526, 527, 561, 574 184, 190, 213, 220, 236, 278, 279, 291,
Levin, M., 514, 527 295, 300, 315, 321, 329, 341, 365, 369,
Lew, H., 111, 123 373, 488, 489, 509, 530, 556, 559,
Lew, H. C., 89, 106, 484, 486, 507 563–567, 569, 575–577, 581, 582, 593
Li, S., 529, 530, 532, 555, 556 Matos, J. F., 220, 237
Liebenberg, R., 96, 106 Maturana, H., 567, 573, 576
Liljedahl, P., 173, 185, 221, 238 Matz, M., 249, 250, 257, 410, 428
Lim, T. H., 388, 389, 391, 393, 401, 406, 407 Mavlankar, A. T., 90, 107
Lin, 454, 457, 481 Mayer, R. E., 173, 184
Lin, F. L., 220, 236, 485, 507 McClain, K., 164, 166, 168, 183
602 Author Index
McDonagh, A., 383, 384, 393, 408 Mulligan, J. T., 279, 300
McDonald, M., 432, 452 Mumford, D., 92, 107
McDougall, D., 274, 276, 279, 300 Muñoz, R., 220, 236, 329, 364
McIntosh, A., 140, 158
McKnight, C. C., ix, xi, 26, 41 N
Mclain, K., 190, 212 Nakahara, T., 275, 275, 278, 301, 508
McLaughlin, M., 516, 526 Nardi, E., 388, 389, 408, 430, 452, 486, 507,
McNeil, N. M., 122, 123, 248, 255, 257, 261, 564, 574
273, 274, 276, 512, 527 Nathan, M. J., 430, 452
Mead, N. A., 529, 556 Navarra, G., 459, 481, 486, 495, 509
Medin, D., 343, 364 Nelson, M., 431, 435, 441, 451
Meira, L., 244, 257, 486, 508, 510 Nemirovsky, R., 215, 219, 222, 236, 242, 257,
Mellin-Olsen, S., 329, 364 258, 314, 321, 410, 428
Méndez, A., 304, 310, 320, 321, 329, 340, 346, Neria, D., 279, 298
356, 365 Nesher, P., 188, 189, 213
Mendicuti, T. N., 485, 486, 508–510 Ness, D., 218, 222, 236
Menzel, B., 486, 510 Newstead, K., 381, 389, 408
Merton, R., 572, 576 Newton, I., 560, 576
Mestre, J. P., 173, 177, 183 Ng, S. F., 27, 34, 36, 41, 89, 106, 111, 123,
Metzger, W., 326, 365 484, 486, 507
Mevarech, Z., 274, 276 Nicéas, L., 226, 237
Mewborn, D., 56, 68, 218, 237, 430, 451
Nichols, E., 64, 68, 189, 211, 534, 555
Mikulina, G. G., 71, 80, 83, 85
Nie, B., 164, 165, 170, 172, 174, 176, 183,
Miles, M. B., 440, 452
184, 530, 532, 555
Miller, S., 567, 576
Nisbet, S., 162, 184, 195, 213
Millman, R., 509
Niss, M., 459, 480
Millsaps, G. M., 11, 22, 485, 508
Nohda, N., 220, 236, 485, 507
Mishanski, J., 401, 408
Noney, K., 56, 68
Mitchelmore, M., 279, 300, 461, 481
Noony, K., 218, 237
Mitchelmore, M. C., 279, 300
Norman, D. A., 455, 481
Mok, I. A. C., 530, 532, 548, 556
Mokros, J. R., 222, 236, 242, 257 Noss, R., 278, 297, 300
Molina, M., 526, 528, 566, 576 Novotna, J., 188, 192, 194, 212, 214
Monk, G., 66, 68 Novotná, J., 278, 301, 304, 320–322, 336, 341,
Monk, G. S., 215, 219, 236 347, 363
Monk, S., 44, 69, 147, 148, 158, 219, 236, 242, Nunes, T., 226, 237
257
Moore-Harris, B., 164, 166, 168, 183 O
Moraova, H., 188, 212 Oehrtman, M., 223, 224, 226, 235
Moraová, H., 278, 301, 304, 320–322, 336, Ogonowski, M., 222, 236
341, 347, 363 Ohlsson, S., 189, 213
Morita, E., 516, 527 Okamoto, Y., 279, 280, 299
Morris, A., 41, 89, 106, 111, 123, 484, 486, Olive, J., 247, 249, 254, 256
507 Oliveira, P. A., 127, 130
Morris, A. K., 192, 213 Oliver, A., 381, 389, 408
Morselli, F., 454, 455, 457, 460, 477, 479, 480, Olivier, A., 96, 106, 278, 297, 301, 341, 366
481 Omanson, S., 423, 428
Moschkovich, J., 243, 254, 257 Onghena, P., 26, 41
Moses, R. P., 35, 39, 41, 430, 452, 563, 576 Onslow, B., 485, 507
Moss, J., 9, 22, 278–280, 296, 298–300, 304, Ortega, M., 128, 129
321, 568, 576 Orton, A., 147, 158, 173, 185, 192, 213, 220,
Moyer, J. C., 89, 106, 111, 123, 164, 165, 170, 237, 278, 300, 341, 365
172, 174, 176, 183, 184, 484, 486, 507 Orton, J., 147, 158, 173, 185, 192, 213, 220,
Mukherjee, A., 90, 107 237, 278, 300, 341, 365
Author Index 603
Osherson, D., 329, 364 Prescott, A., 279, 300

Ott, J. M., 164, 166, 168, 183 Presmeg, N., 314, 321
Owens, D., 11, 22 Price, E., 478, 481
Owens, D. T., 173, 184, 485, 508 Price, J., 164, 166, 168, 183
Owens, K., 317, 320, 346, 364 Puig, L., 310, 318, 320, 367, 373, 460, 481
Purdy, D., 485, 507
P Putnam, R., 140, 158, 431, 452
Pahlke, E., 514, 515, 527 Putnam, R. T., 191, 213
Paine, L., 380, 408
Pang, M., 565, 575 R
Papic, M., 567, 576 Radford, L., 138, 158, 188, 192, 194, 197, 203,
Parenti, L., 455, 480 206–208, 213, 244, 258, 284, 295, 301,
Park, K., 336, 363, 529, 556 310–312, 314, 318, 319, 321, 326, 327,
Partelow, L., 514, 515, 527 329, 340, 341, 346, 347, 356, 365, 370,
Pateman, N. A., 9, 21, 123, 123, 204, 214, 218, 371, 373, 460, 481, 486, 510, 514, 528,
226, 235, 237, 279, 282, 299, 303, 320, 591, 593
324, 329, 365, 388, 407, 488, 507, 509 Raizen, S. A., ix, xi, 26, 41
Pedemonte, B., 458, 479, 480, 481 Rajagopalan, S., 90, 107
Pehkonnen, E., 329, 364 Rasmussen, C. L., 218, 237
Peled, I., 191, 192, 213, 226, 237 Raudenbush, S. W., 176, 185
Pelfrey, R., 164, 166, 168, 183 Recio, A. M., 56, 68
Penny, B., 317, 320 Recorde, R., 569, 576
Perera, J. S. H. Quintus, 383, 384, 393, 401, Redden, T., 192, 214
406, 407 Reed, M., 11, 22
Pérez, A., 173, 184, 580, 593 Reed, M. K., 485, 508
Perez, A., 485, 509 Resnick, L., 423, 426, 427, 428
Perry, B., 346, 364, 461, 481 Resnick, L. B., 140, 158, 162, 185
Persson, J., 218, 235 Reys, B. J., 121, 124, 140, 158
Peterson, J., 382, 400, 407 Reys, R. E., 121, 124, 140, 158
Peterson, P. L., 191, 213 Richardson, V., 430, 451
Philipp, R., 262, 276 Richland, L., 347, 365
Philipp, R. A., 6, 23 Richmond, D. E., 381, 407
Phillips, E., 278, 299 Rickart, C. E., 381, 407
Phillips, E. D., 163, 167, 184, 240, 250, 251, Rittle-Johnson, B., 261, 273, 276
257 Rivera, F., 147, 154, 155, 158, 278, 301, 304,
Phillips, G. W., 529, 556 320, 324, 329, 330, 336, 341, 347, 354,
Piaget, J., 189, 213, 371, 373, 411, 428, 566, 363, 365
576 Rivera, F. D., 154, 158, 244, 258
Pierce, R., 383, 384, 393, 408, 461, 481 Roberts, D. L., 126, 130
Pimm, D., 382, 401, 408, 565, 569, 576 Roberts, S. A., 431, 436, 441, 452
Pirie, S., 145, 158, 530, 556 Robertson, M. E., 339, 340, 366
Pirie, S. E. B., 145, 158 Robutti, O., 314, 320, 371, 372, 469, 479
Pittman, M., 431, 435, 442, 446, 451, 452 Roche, A., 383, 384, 393, 408
Pittman, M. E., 431, 435, 441, 451 Roey, S., 515, 527
Pligge, M., 410, 428 Rogoff, B., 162, 185
Plofker, K., 91–93, 95, 107 Rojano, T., 139, 157, 310, 318, 320, 321, 336,
Ponte, J. P., 127, 130, 488, 510 343, 346, 363, 365, 366, 367, 368, 370,
Post, T., 242, 257, 410, 426, 427 372, 373, 459, 480, 588, 593
Post, T. R., 9, 21, 31, 41 Romberg, T., 6, 23, 26, 41, 173, 184, 413, 428
Potari, D., 488, 510 Romberg, T. A., 7–9, 18, 19, 22, 215, 218, 219,
Pradhan, H. C., 90, 107 236, 237, 242, 258, 368, 373, 424, 426,
Pratt, D., 564, 574 426
Pratt, K., 485, 507 Romnberg, T., 190, 213
604 Author Index
Roper, T., 278, 300, 341, 365 Schoenfeld, A. H., 56, 68, 218, 238, 242, 243,
Rosamund, F., 222, 235 254, 257, 258, 458, 480, 488, 510
Roschelle, J., 223, 237, 243, 258 Schram, T., 450, 452
Rose, B., 488, 507 Schubring, G., 319, 321
Rosebery, A., 410, 428 Schulte, A. P., 188, 214
Ross, J.A., 279, 300 Schwank, I., 96, 106
Roth, M.-W., 314, 321 Schwartz, D., 424, 427
Rubenstein, R., vii, x, 96, 106, 110, 111, 122, Schwartz, J., 8, 23, 568, 575
123, 125, 129, 173, 183, 282, 301, 318, Schwartz, J. L., 138, 157
320, 413, 425, 426, 512, 513, 526, 527 Schwarz, B., 488, 510
Rumbaugh, D. M., 308, 321 Schwarz, B. B., 222, 237
Rumi, J., 572, 576 Schwarz, J. L., 9, 22
Runesson, U., 530, 532, 556 Schwille, J., 380, 408
Russel, S. J., 222, 236 Scudder, K. V., 424, 428
Russell, S. J., 44, 46, 49, 67, 68, 69, 147, 148, Seeger, F., 319, 321
158 Segalis, B., 191, 192, 213
Seltzer, 415, 428
S Senk, S. L., 38, 41, 163, 180, 185
Sabena, C., 314, 321 Seo, 278, 299
Sadovsky, P., 534, 556 Seo, D., 336, 363
Sáiz, M., 304, 310, 320, 321, 329, 340, 346, Sepulveda, A., 336, 363, 365
356, 365 Serfati, M., 318, 321, 322
Sakonidis, H., 92, 106, 354, 365, 529, 530, 556 Serrano, A., 514, 515, 528
Sallee, T., 247, 258
Serrazina, L., 127, 130
Santillan, M., 380, 408
Sessa, C., 534, 556
Santos, L., 488, 509
Sew, H., 336, 363
Santos, M., 187, 191, 193, 207, 212
Sfard, A., 173, 185, 188, 189, 192, 194, 204,
Sarma, K. V., 94, 107
214, 218, 237, 294, 301, 400, 408, 541,
Sasman, M., 278, 297, 301, 341, 366
556
Sasman, M. C., 96, 106
Sharry, P., 194, 212
Savage-Rumbaugh, E. S., 308, 321
Sherin, B., 244, 245, 254, 256, 258
Saveleva, O. V., 71, 80, 85
Sherin, M. G., 488, 510
Savyelyeva, O. V., 83, 85
Sherin, S., 254, 258
Sawyer, W., 569, 576
Saxe, G., 415, 428 Shigematsu, K., 220, 236, 485, 507
Scandura, J. M., 189, 193, 214 Shillolo, G., 9, 22, 279, 300
Scardamalia, 279, 299 Shin, I. H., 380, 408
Schauble, L., 435, 451 Shoenfield, A. H., 215, 219, 236
Schifter, D., 8, 16, 23, 26, 41, 44, 49, 57, 67, Shteingold, N., 9, 22
68, 69, 147, 148, 158, 222, 237, 260, Shulman, L. S., 436, 452
275, 431, 452 Shulte, A., 31, 41
Schifter, D. V., 44, 46, 68 Shulte, A. P., 111, 124
Schliemann, A. D., viii, x, 9, 12, 21–23, 26, 36, Siegler, R., 343, 364
41, 138, 139, 157, 190, 192, 212, 214, Sierpinska, A., 469, 479
218, 220, 226, 235, 237, 244, 256, 275, Silver, E. A., 9, 21, 26, 41
275, 278, 279, 289, 297, 299, 301, 303, Simmt, E., 573, 575
304, 317, 320, 508, 512, 525, 527, 561, Sims-Knight, J., 66, 68
568, 574–576 Singh, A. N., 93–95, 106
Schmidt, W., 380, 408 Skemp, R. R., 188, 214
Schmidt, W. H., ix, xi, 26, 41, 128, 130 Slaughter, Ml., 56, 68
Schmittau, J., 41, 72, 83, 84, 85, 89, 106, 111, Slavit, D., 96, 107, 140, 158, 222, 237
123, 484, 486, 507, 567, 576 Sleeman, D., 249, 250, 257, 261, 276, 410, 428
Schneider, C., 431, 435, 436, 441, 451, 452 Sleeman, D. H., 410, 428
Schoen, H., 240, 256 Slovin, H., 567, 575
Author Index 605
Small, M., 278, 299 Streefland, L., 413, 428

Smith, C., 488, 507 Stroup, W., 223, 237
Smith, E., 8, 16, 22, 23, 222, 225, 235–237 Stylianides, A. J., 458, 481
Smith, G. C., 381, 407 Stylianou, D. A., 57, 68, 488, 507
Smith, J., 216, 226, 230, 237, 243, 254, 258 Subramaniam, K., 98, 102, 103, 106, 107
Smith, J. P., 243, 258, 422, 428 Sugiyama, Y., 114, 123
Smith, M., 515, 516, 528 Summara, D., 573, 575
Smith, N., 218, 235 Sun, X., 529, 530, 556
Smith, S. T., 308, 321 Sutherland, J., 56, 68
Solomon, J., 410, 428 Sutherland, R., 193, 194, 212, 318, 321, 343,
Song, N., 530, 548, 555, 556 346, 366, 368, 370, 372, 373, 456, 480,
Sophian, C., 424, 428 566, 574, 588, 593
Sorde Marti, T., 453, 481 Swafford, J. O., ix, xi, 240, 257, 341, 342, 347,
Soury-Lavergne, S., 591, 593 366, 426, 427, 514, 528
Sousa, H., 127, 130 Swain, H., 381, 407
Southwell, B., 317, 320, 346, 364 Swan, M., 485, 507
Sowder, L., 56, 68, 458, 480 Swars, S. L., 170, 176, 183, 184
Spagnolo, F., 371, 372, 525, 528 Sykes, G., 431, 436, 451, 452
Srinivas, M. D., 94, 107 Szeminska, A., 411, 428, 566, 576
Sriraman, B., 367, 370, 371, 373 Sztajn, P., 56, 68, 218, 237
Stacey, K., vii, x, 26, 41, 91, 94, 95, 98, 99,
106, 107, 147, 156, 157, 158, 188, 189,
T
192, 193, 197, 203, 208, 212, 214, 220,
Tabachnikova, N. L., 83, 85
236, 237, 262, 276, 278, 279, 288, 291,
Tagg, A., 144, 159
297, 301, 317, 320, 343, 346, 347, 364,
Tahta, D., 559, 576
366, 459, 461, 481, 486, 495, 507–510,
Tairab, H. H., 388, 407
512, 513, 526, 527, 528, 559, 575
Takahashi, A., 112, 114, 121, 122, 123, 124
Stacy, K., 7, 22
Tall, D. O., 187, 188, 191, 193, 194, 212, 214,
Stafylidou, S., 410, 413, 428
327, 364, 485, 508
Stancavage, F., 128, 129
Taplin, M. L., 339, 340, 366
Steele, D., 278, 301, 341, 342, 360, 366
Steen, L. A., 173, 185 Tatto, M. T., 380, 408
Steffe, L., 218, 237 Tempesta, I., 486, 508
Steffe, L. P., 72, 73, 85, 138, 140, 158, 192, Thomas, G., 144, 149, 157, 159
205, 209, 210, 212 Thompson, A. G., 6, 23, 516, 528
Stehlikova, N., 188, 212 Thompson, D. R., 38, 41, 163, 180, 185
Stehlíková, N., 304, 320–322, 336, 341, 347, Thompson, P., 216, 226, 230, 237
363 Thompson, P. W., 6, 23, 218, 238, 516, 528
Stehlková, N., 278, 301 Thornton, C., 162, 184
Stein, M., 570, 576 Tierney, C., 222, 236
Stein, M. K., 168, 171, 182, 185, 221, 236, Tinker, R. F., 242, 257
241, 249, 257 Tirosh, D., 371, 373, 381, 389, 408
Steinberg, R., 261, 276 Törner, G., 371, 373
Stephens, A., 248, 255, 257, 274, 276 Townsend, B., 278, 300, 341, 364
Stephens, A. C., 122, 123, 512, 527 Treffers, A., 332, 366
Stephens, M., 91, 95, 96, 98, 106, 138, 152, Trinick, T., 144, 149, 157, 159
157, 193, 197, 203, 208, 212, 318, 320, Tripathi, P., 88, 107
513, 527, 566, 576 Truxaw, C., 243, 256
Stevenson, H. W., 529, 556 Tsamir, P., 381, 388, 389, 408
Stiff, L. V., 26, 41 Tsui, A. B. M., 530, 532, 556
Stigler, J., 347, 365 Turner, E. E., 410, 413, 415, 427
Stigler, J. S., 528 Twisk, J. W. R., 150, 159
Stigler, J. W., 514–516, 528 Tzanakis, C., 312, 319, 321
Stinson, D. W., 170, 176, 183, 184 Tzekaki, M., 92, 106, 354, 365, 529, 530, 556
606 Author Index
U Watson, A., 278, 301, 530, 556, 564, 566, 576,

Underhill, R. G., 8, 22 577
Ursini, S., 193, 214 Watterson, G. A., 381, 407
Usiskin, Z., 35, 39, 41, 111, 114, 124, 163, Wearne, D., 182, 183, 450, 452, 515, 527
180, 185, 188, 214, 262, 276 Weber, K., 459, 481
Uttal, D. H., 424, 428 Wenger, E., 243, 256, 431, 452
Wheeler, D., 278, 288, 300, 485, 509
V White, D., 218, 237
Vaisenstein, A., 67, 68 White, D. Y., 56, 68
Vaiyavutjamai, P., 382–386, 388, 389, 391, Whiteside, D., 560, 577
393, 397, 400, 401, 406, 408 Wiegel, H., 218, 237
Valentine, C., 410, 428 Wiegel, H. G., 56, 68
Van de Walle, J., 423, 428 Willard, T., 164, 166, 168, 183
van den Heuvel-Panhuizen, M., 96, 98, 107, Williams, E., 343, 364
190, 214, 346, 365, 486, 507 Williams, J., 430, 452
van Dooren, W., 26, 41 Williams, T., 514, 515, 527
van Dormolen, J., 329, 364 Willoughby, S., 282, 301
van Merriënboer, J., 514, 527 Wilmore, E., 114, 124
van Oers, B., 244, 245, 254, 256 Wilson, M., 215, 219, 236
Varela, F., 567, 576 Wilson, R. L., 188, 214
Vergnaud, G., 411, 428, 566, 577 Winicki-Landman, G., 430, 452
Verschaffel, L., 26, 41, 244, 245, 254, 256, Wirszup, I., 128, 130
559, 577 Wittenberg, L., 44, 68
Vielhaber, K., 164, 166, 168, 183 Woepcke, F., 309, 322
Viète, F., 310, 322 Wong, K. Y., 388, 407
Villani, V., 330, 341, 364 Wong, N., 529, 530, 532, 555, 556
Vincent, J. L., x, 91, 94, 95, 98, 106, 156, 157, Woo, J., 336, 363
188–190, 192, 193, 197, 198, 203, 206, Wood, T., 488, 510
208, 212, 214, 314, 321, 329, 336, 363, Wozniak, R. H., 162, 185
559, 575 Wright, R., 140, 159
Vincent, Jl., 486, 495, 507–510
Vincent, Jn., 486, 495, 507–510 Y
Vinner, S., 218, 238 Yackel, E., 190, 212, 488, 510
Vlassis, J., 530, 556 Yamaguchi, T., 329, 364
von Glasersfeld, E., 140, 158, 242, 254, 257, Yaroshchuk, V. L., 128, 130
485, 508 Yerushalmy, M., 218, 238
Vosniadou, S., 410, 413, 428 Yitschak, D., 274, 276
Vygotsky, L. S., 12, 23, 71–73, 84, 85, 85, Yoshida, M., 112, 114, 121, 122, 123, 124
563, 577 Yoshikawa, S., 114, 124
Young-Loveridge, J., 144, 149, 157, 159
W
Wagner, S., 485, 508 Z
Walker, R. J., 381, 407 Zan, R., 488, 497, 509
Wang, C., 172, 174, 183 Zandieh, M. J., 218, 238
Wang, N., 170, 172, 174, 176, 183, 184 Zaslavski, O., 278, 297, 301
Ward, J., 144, 159, 560, 577 Zaslavsky, O., 221, 222, 236, 238, 241, 249,
Warren, B., 410, 428 257, 341, 366
Warren, E. A., 188–190, 192–195, 197, 198, Zazkis, R., 173, 185, 221, 238
203, 204, 206, 209, 211, 211–214, 220, Zeringue, J., 411, 413, 427
236, 244, 258, 278, 279, 299, 301, 304, Zhang, D., 548, 556
322, 340, 343, 364 Zilliox, J. T., 9, 21, 123, 123, 190, 192, 204,
Watanabe, T., 112–114, 121, 122, 123, 124 205, 212, 214, 218, 226, 235, 237, 279,
Watkins, A., 240, 256 282, 299, 303, 320, 324, 329, 365, 388,
Watkins, D. A., 530, 555 407, 488, 507, 509
Subject Index
A Algebraic equation, 33, 165, 261, 274, 370,

Abstraction, 71, 72, 74, 75, 126, 189, 192, 194, 378, 529–531, 533, 534, 544, 546,
209–211, 278, 279, 294, 295, 305, 548
307, 308, 341, 369, 370, 424, 531, Algebraic language, 288, 346, 347, 378,
547, 584, 585 453–463, 477, 478, 484, 485, 487,
Algebra, vii–x, 3–5, 25–29, 31–33, 35–40, 492, 493, 495, 503, 538, 567, 590
43–45, 51, 58, 63, 64, 66, 67, Algebraic reasoning, 430
71–74, 77, 79, 80, 84, 87–103, 105, Algebraic thinking, vii, ix, x, 3, 7, 14, 17–19,
109–114, 121–123, 125–129,
25–27, 32, 34, 35, 37, 40, 43, 72,
135–139, 146, 147, 152–156,
76, 85, 89, 90, 111, 121, 122, 126,
161–165, 168, 173, 180, 181,
127, 129, 135–141, 146–154, 156,
187–190, 195, 215–218, 220–224,
157, 161–165, 172–174, 176, 180,
226, 227, 230, 232, 233, 235,
181, 190, 193, 196, 206, 218, 240,
239–247, 249, 253–256, 259–263,
279, 295, 298, 303–305, 308, 309,
268, 274, 275, 277, 278, 303, 304,
311–314, 316–319, 367–370, 372,
308–312, 317–319, 330–332, 334,
377, 378, 387, 431, 432, 447, 459,
341, 342, 346, 347, 367–372, 377,
378, 409–412, 417, 418, 420, 422, 460, 484, 486, 487, 492, 505, 530,
423, 426, 429–432, 436, 438–443, 538, 543–548, 550, 579–586,
447, 450, 455–461, 463–467, 469, 588–592
470, 473, 477–479, 484–487, Arithmetic, viii–x, 4, 26, 29–32, 35–40, 43–45,
490–497, 502, 504–506, 511–513, 50, 51, 58, 61, 63, 67, 71–74, 77,
515, 516, 525, 526, 530, 532, 533, 87–100, 105, 111, 114, 115,
537, 541, 543–548, 557, 559–571, 120–123, 125–127, 129, 135–140,
573, 574, 579–592 146, 147, 152–154, 156, 188, 190,
Algebra curriculum, 153, 260 193–195, 201, 208, 210, 244, 255,
Algebra education, 382, 387, 388, 391, 581, 260, 261, 275, 282, 291, 303–305,
585 308, 309, 311, 317–319, 367–370,
Algebra for all, vii 409–412, 417, 424, 426, 429, 431,
Algebra instruction, 122, 239, 377, 430, 450, 459, 461, 462, 469, 484–487,
515 490–495, 497, 504, 505, 512, 513,
Algebra language, 454 537, 559, 561–563, 565–569, 571,
Algebra learning, 161, 163, 165, 278, 369 580, 581, 583, 584, 586, 589, 592
Algebra teacher learning, viii Arithmetic and algebraic approaches, 29, 31,
Algebraic, 30, 50, 59, 429–432, 447, 450, 491, 38, 586
581, 582, 588, 589 Arithmetic curriculum, 129, 136

608 Subject Index
Arithmetic operations, 43–45, 129, 216, 242, Elementary mathematics curriculum, 32, 39,
369, 587 123
Arithmetic properties, 455, 484 Elementary school algebra, 4, 109, 122
Elementary schools, 109
C Embodiment, 74, 370, 584
Chinese classroom, 29, 529 Equal sign, ix, 37, 64, 66, 122, 241, 242, 248,
Chinese curriculum, 28–32, 38, 39 249, 255, 261–270, 273, 274, 491,
Classroom instruction, 162, 164, 169, 170, 512, 513, 525, 526, 530, 534, 535,
244, 255, 437, 450, 511, 514–516, 588
587 Equality, 63, 78, 79, 140, 166, 168, 261, 269,
Cognitive processes, 242, 243, 254, 489 304, 409–412, 417–421, 423–425,
Conceptual understanding, 88, 161, 171, 176, 435, 463, 477, 484, 495, 496, 505,
181, 369, 459, 586 512, 513, 515, 520, 525, 537, 539,
Critical reflections, 488, 505 549, 560, 562, 565, 583, 584
Curriculum, 3–8, 12, 18, 21, 26–30, 32–36, Equation, 28, 29, 31, 32, 38, 39, 52, 62, 75,
38–40, 44, 63, 67, 71–81, 84, 85, 77–80, 84, 89, 91–95, 110, 113,
87, 88, 90, 99–101, 109, 111–114, 120, 121, 126, 128, 129, 153, 162,
119, 121–123, 125–129, 135–137, 165, 166, 168, 169, 173, 174, 177,
139, 144, 153, 154, 156, 161, 181, 182, 187, 188, 190, 191, 195,
163–169, 174, 176, 178, 180–182, 198–209, 211, 220, 230, 241–244,
222, 250, 260, 270, 275, 277, 285, 246–248, 250, 255, 261, 263–270,
295–297, 304, 305, 317, 332, 368,
305, 372, 377–384, 386–396,
369, 380–382, 400, 410, 424, 445,
398–401, 404, 406, 415, 420, 422,
451, 465, 485, 554, 555, 579, 583,
445, 466, 506, 511, 513, 516, 517,
588, 592
520, 522–526, 530, 533–541, 543,
E 544, 547, 548, 550, 551, 553, 560,
Early algebra, viii, x, 7, 9, 12, 15, 16, 44, 67, 564, 568, 569, 585–588, 590
125, 126, 135, 187, 188, 190, 206, Extensive, 73
209, 260, 275, 277, 317, 372, 377, External, 244
378, 483–487, 490, 491, 496, 497,
501, 504–506, 579–585, 587–589, F
592 Fraction, 28, 30, 32, 39, 59, 63, 66, 67, 92, 94,
Early algebra curriculum, 18 109, 110, 125, 127, 143, 144, 148,
Early algebra instruction, 377 191, 197, 206, 368, 378, 409–418,
Early algebra learning, 25, 27 422–426, 561, 563, 571
Early algebra variable, 513 Function, 4, 27–29, 34, 35, 44, 84, 92, 101,
Early algebraic thinking, 19, 138, 188, 591 109, 111–113, 121, 128, 129, 135,
Early algebraization, viii, ix, 367, 368, 370, 136, 206, 211, 215, 216, 218–220,
372, 579, 580 222, 224–227, 230, 235, 243–245,
Effective teaching, 188, 494 277, 278, 282, 283, 286, 287, 291,
Elementary, 3, 4, 25–29, 32, 35–40, 43, 44, 46, 293, 295–297, 330, 346, 433, 435,
51, 56–59, 63, 67, 68, 71–74, 84, 440, 468, 484, 491, 494, 502, 506,
109, 111, 112, 120–123, 137, 139, 512, 525, 538, 546, 559, 580, 584,
141, 147, 156, 187, 193, 195, 207, 592
241, 244, 259, 261, 274, 275, 277, Function machine, 187, 190, 193, 198–201,
279, 303, 313, 317, 318, 324, 368, 206, 207, 209–211, 281–285, 291,
377, 409–411, 417, 418, 422, 423, 293, 296–298
426, 438, 455, 462, 465, 549, Function-based, 588
579–581, 584, 585, 588, 591 Functional, 121, 435, 583, 584, 587
Elementary algebra, 485 Functional approach, 34, 136, 161, 164, 165,
Elementary arithmetic, 483, 484 182, 218, 286, 369, 584, 587
Elementary curriculum, 28, 29, 40, 45, 72, 73, Functional thinking, 5, 7–9, 11, 15–21, 126,
84, 85, 412, 423 135, 153, 188, 190, 195, 198, 211,
Elementary mathematics, 27, 44, 411 215, 218, 222, 589
Subject Index 609
G 412, 417, 423, 425, 426, 429–432,

Generalization, 35, 39, 44, 47, 48, 51, 53, 435, 440, 445, 450, 451, 454,
55–57, 59, 61–64, 71, 72, 74, 87, 89, 457–460, 462, 463, 483–487, 489,
91, 94, 105, 125, 126, 138, 139, 145, 491, 492, 494, 495, 501, 504–506,
149, 153, 154, 161, 162, 173, 174, 513, 529–533, 536, 538, 540–542,
178–182, 222, 230, 278, 279, 285, 544–550, 555, 557, 562, 565,
295, 305, 310, 317–319, 323–335, 567–569, 572, 574, 581, 583, 586,
339–342, 344, 346, 347, 351, 352, 588, 589, 592
355, 356, 358, 360, 362, 363, Learning algebra, 98, 147, 189, 260, 409–411,
368–370, 485, 491, 500, 502, 513, 413, 426, 459, 460, 530, 532, 545,
514, 516, 580–582, 584, 588, 591 560, 571
Generalizing, ix, 32, 33, 43, 44, 51, 67, 127, Learning arithmetic, 369
135, 147, 148, 155, 188, 192, 193, Learning fractions, 409, 411, 413, 426
197, 199, 203, 204, 207, 210, 223, Learning representations, 370
277–279, 297, 298, 305, 327, 332, LieCal project, 161, 163, 164, 174, 181
335, 336, 341, 344, 347–352, 354, Linear function, 136, 201, 216, 218–220, 223,
356, 357, 369, 370, 435, 487, 227, 231, 243, 244, 277, 280, 290,
512–514, 516, 525, 538, 567, 296, 467, 583, 587
580–582, 584, 588–591 Linguistic, 8, 12, 15, 16, 19, 313, 327, 368,
Geometric growing patterns, 281 370, 378, 454, 483, 487, 490,
492–495, 506, 538
H Longitudinal study, 136, 147, 149, 150, 153,
Habermas, 378, 453–455, 457, 461, 462, 161, 323, 329, 337, 362
467–469, 473, 477–479, 590
History of mathematics, 370 M
Mathematical thinking, 541
I Mathematics education, vii–ix, 4, 162, 163,
Induction, 126, 191, 192, 356, 547 239–242, 259, 260, 262, 273, 275,
Inequalities, 110, 113, 126, 181, 377, 379–390, 278, 317, 332, 367, 370, 371, 454,
392–401, 404, 406, 513, 568, 588 468, 469, 477, 529–531, 557, 572,
International perspective, 27, 39 574, 586
Mathematics instruction, 67, 356, 430, 438,
J 450, 511, 514, 515
Japan, x, 3, 4, 109, 111–114, 119, 121–123, Mathematics teachers, 27, 169, 331, 401, 432,
126, 128, 378, 388, 495, 511, 530, 531, 569
514–516, 525, 562, 568, 583, 584 Mathematics teaching, viii, 206, 429, 436, 451,
Judgment, 272 530, 532, 549
Justification, 51, 53, 55, 56, 94, 98, 105, 216, Mental strategies, 141, 143, 144
217, 230, 271–273, 275, 310, 323, Metacognitive, 298, 395, 483, 488, 493,
325, 327, 335, 355, 356, 360, 363, 499–501, 589
453, 456, 467, 478, 493, 562, 566, Middle grades, 46, 58, 67, 68, 88, 135, 136,
580, 589, 590 180, 244, 260, 327, 363, 411
Middle-school mathematics, 163, 164, 387,
L 396
Layers of generality, 154, 311 Middle-school mathematics students, 387
Learning, viii, 28, 29, 35, 38, 44, 46, 66, 72, Middle-school mathematics teachers, 379, 383,
73, 87–89, 91, 95, 97, 98, 104, 121, 384, 387, 396, 401
135–137, 140, 141, 144, 153, 155, Multiplication, 32, 34, 45, 46, 48–51, 56–58,
162–165, 171, 173, 174, 180, 182, 61, 62, 110, 113, 115, 116, 127,
187, 188, 190, 191, 194–196, 136, 141–144, 146, 201–203, 209,
206–209, 211, 235, 239, 240, 246, 266, 277, 280, 282, 285, 288–290,
247, 249, 253, 255, 262, 278–280, 295, 298, 316, 334, 342–344, 353,
285, 297, 298, 303–305, 312, 329, 357, 412, 416, 417, 419–421,
330, 344, 369, 371, 377, 378, 411, 424–426, 464, 521, 565, 584
610 Subject Index
N 117–120, 126, 136, 138, 140, 165,

New Zealand Numeracy Project, 137–140, 215–218, 222–224, 226, 228–230,
149, 152, 153, 369 232–235, 248, 250, 251, 256, 261,
263, 264, 274, 278, 298, 310, 318,
O 332, 372, 411–417, 420, 422–424,
Operation, ix, 31, 33, 34, 37–39, 43–46, 48, 484, 492, 494, 513, 541, 543, 549,
51, 55, 57–59, 61, 63, 64, 67, 74, 580, 582, 584, 586, 589, 592
78, 89, 91, 92, 94–96, 98–102, 105,
110, 112, 113, 117, 121, 122, 127,
129, 137–140, 147, 155, 166, 189, R
194, 198, 199, 201–203, 208–210, Random coefficients analysis, 151
216, 217, 235, 261, 274, 289, 295, Rationality, 453–461, 463–469, 471, 473, 474,
314, 323, 332, 333, 340, 342–344, 477–479, 533, 590
348, 368, 369, 409–413, 415–421, Relational thinking, 91, 98, 215, 409–413,
423–425, 460, 472, 493, 495, 502, 416, 417, 421–426, 490, 565, 566,
525, 543, 561, 563, 565, 566, 581, 571, 583, 584, 588, 589
583–585, 587–592 Representations, 6, 9, 12, 25, 26, 28, 29, 39,
46, 57, 58, 63, 64, 76, 90, 94, 99,
P 110, 121, 122, 126, 136, 154, 155,
Pattern generalization, 235, 305, 312, 317, 187, 188, 190–196, 198–200,
323, 327–332, 334, 335, 337–343, 202–211, 218, 220, 221, 226, 230,
347, 348, 352–354, 357, 358, 360, 233, 239–247, 249–251, 253–256,
362, 363, 370, 582, 583, 591 261, 277–280, 282, 285, 293, 296,
Patterning, 188, 190, 195, 198, 210, 277–279, 297, 326, 331, 350, 368, 370, 372,
285, 288, 289, 293–296, 298, 304, 415, 430, 433, 435, 442, 444, 445,
323, 325, 329, 334, 340, 351, 357, 450, 458, 470, 471, 473–476, 483,
358, 360, 363, 370, 583, 591, 592 484, 487, 496, 497, 502, 505, 532,
Patterning curriculum, 280 533, 538–540, 542–549, 554, 583,
Patterning problem solving, 279 585, 587, 589, 590
Pictorial equations, 25, 32, 33, 39, 585 Representations using, 121
Problem solving, ix, 31, 39, 111, 123, 162,
163, 263, 298, 416, 457, 525, 532,
533, 537, 542–544, 548, 555, 561, S
580 School algebra, 7, 8, 93, 95, 109–112, 122,
Procedural knowledge, 161, 171, 541 125, 173, 180, 218–220, 368, 370,
Professional development, 44, 49, 194, 279, 380, 383, 388, 406, 409, 410, 420,
377, 410, 429, 431, 432, 486, 570, 423, 461, 530, 557, 559, 560, 564,
571, 579 574, 579, 585, 586, 588, 592
Proof, 51, 56–58, 63, 278, 457–459, 479, 485, School algebra curriculum, 3
589 School algebraic thinking, 327
Properties, 33, 39, 45, 46, 48, 49, 57, 63, 73, Semiotics, 138
74, 81, 82, 89, 110, 121, 146, 147, Solving equations, 32, 84, 110, 181, 420, 460,
168, 222, 230, 297, 340, 350, 354, 512, 537, 539, 540, 587
357, 369, 409–413, 417–426, 438, Structural approach, 136, 161, 164–166, 369,
461, 484, 486, 488, 505, 512, 520, 587
559, 560, 563, 565, 566, 568, 571, Structure of activity, 560
583, 584 Student learning, 196, 435, 548
Psychological tool, 71, 74, 76, 79–85 Symbolic expressions, 91, 96, 97, 99, 105,
Psychological tool function, 80 216, 585
Q
Quantitative reasoning, 216, 218, 224, 230, T
580 Teacher, vii
Quantities, 3, 29, 31, 47, 64, 65, 73, 74, 76–82, Teacher education, 377, 379, 572
84, 89, 91–95, 99–101, 112–114, Teacher learning, viii, ix, 17, 19, 194
Subject Index 611
Teaching with variation, 529–532, 538, 152, 154, 165, 166, 168–170, 176,
543–546, 548, 586 188–190, 197, 199, 208, 211,
Theoretical concept, 74, 85 223–227, 250, 259–266, 270, 271,
Transition, 26, 27, 31, 35, 37, 40, 45, 58, 66, 273–275, 281, 310, 311, 315, 318,
88–91, 99–101, 125, 126, 139, 193, 346, 349, 433, 455, 456, 466, 484,
230, 233, 244, 346, 429, 431, 512, 491, 492, 496, 504, 512, 513, 517,
513, 525, 534, 537, 565, 568, 571, 525, 526, 545–547, 551, 552, 587,
585 592
V Video Study, 511, 514, 515
Variable, 28, 29, 32, 34, 38, 66, 82–84, 89–91, Vygotskian perspective, 71
95, 96, 99, 105, 118–120, 126, 138, Vygotsky, 4, 126, 563, 584
Editors and Contributors
Martha W. Alibali is a professor of psychology at the University of Wisconsin-

Madison. Her research addresses three major themes: (a) mechanisms of knowledge
change in the development of mathematical reasoning, (b) the role of visual scaf-
folding (including gestures, diagrams, and other inscriptions) in instructional com-
munication, and (c) the role of gesture in perceptual and spatial thinking. She is a
fellow of the Association for Psychological Science and a governing board member
of the Cognitive Development Society.
Rakhi Banerjee has a PhD in Mathematics Education from the Tata Institute of
Fundamental Research, Mumbai, India. At present, she is an Assistant Professor in
the Tata Institute of Social Sciences, Mumbai, India. She has varied research in-
terests in the area of Mathematics Education. Particularly, teaching and learning of
Mathematics at elementary level and related classroom processes; meaning mak-
ing and the use of symbols in elementary mathematical understanding; nature and
role of teacher knowledge, values, and beliefs in teaching practices. Her teaching
includes courses in cognitive development and mathematics education.
Virginia Bastable has been the Director of the SummerMath for Teachers program
at Mount Holyoke College since 1993 after working as a secondary mathemat-
ics teacher for more than twenty years. In collaboration with Deborah Schifter of
EDC and Susan Jo Russell of TERC, Bastable co-authored the professional devel-
opment series, Developing Mathematical Ideas. Current work includes developing a
book and on-line course for elementary teachers, Connecting Arithmetic to Algebra.
Bastable and Schifter also lead the Mathematics Leadership Program, designed to
provide support for teacher-leaders, math specialists, math coaches to help educa-
tors reconceive the way mathematics instruction takes place.
Maria Blanton is Professor in the STEM Department at University of Mas-
sachusetts Dartmouth and Senior Research Scientist at the Kaput Center for Re-
search and Innovation in STEM Education. Her early algebra research has focused
on children’s understanding of functions and developing teacher practice that fosters
algebraic thinking. Blanton is currently PI on the NSF project Developing Algebra

614 Editors and Contributors
Ready Students for Middle Grades and is co-editor (with J. Kaput and D. Carraher)
of the research volume Algebra in the Early Grades (2008) and author of Alge-
bra and the Elementary Classroom: Transforming Thinking, Transforming Practice
(2008).
Joanne Rossi Becker is a Professor of Mathematics at San José State University in
the U.S. Her research interests include gender and mathematics, teacher professional
development, and generalization in algebra. She currently teaches content and meth-
ods courses for prospective teachers and directs several professional development
projects. In addition she is fiscal agent for the Silicon Valley Mathematics Initiative,
an umbrella group for professional development in the San Jose area. Among its
noteworthy efforts is the Mathematics Assessment Collaborative that provides per-
formance assessments, and concomitant professional development, in grades 2–10
for thousands of students.
Paolo Boero was born in 1941; formerly a mathematician, at present he is associate
professor of Mathematics Education at the Mathematics Department of the Genoa
University. His research interests in Mathematics Education concern: linguistic and
representation aspects of the teaching and learning of mathematics (the title of a
national research project leaded by him in the last decade); approach to proof and
proving; concept construction in contexts. He leads a group of school teachers and
university researchers engaged in research for classroom innovation in primary and
lower secondary school. He serves as associate editor of Educational Studies in
Mathematics.
Hilda Borko (Ph.D., UCLA, 1978) is a professor of education at Stanford Univer-
sity. Her research explores the process of learning to teach, with an emphasis on
changes in novice and experienced teachers’ knowledge and beliefs about teaching
and learning, and their classroom practices as they participate in teacher education
and professional development programs. Dr. Borko’s teaching interests are in the
related areas of classroom instructional processes, and teacher learning.
Murray Britt was a Principal Lecturer at The University of Auckland, now re-
tired. Originally from Christchurch New Zealand and the University of Canterbury,
he has specialities in mathematics education research and curriculum development.
His research has focused on mathematics teacher professional development, and the
learning and teaching of algebra in elementary and secondary schools. His expertise
in these areas is largely derived from working extensively with teachers and stu-
dents in the Pacific region: Fiji at the University of the South Pacific, Papua New
Guinea (Unesco/UNDP), and New Zealand at The University of Auckland. He has
contributed to the development and writing of national curriculum documents in
mathematics in Papua New Guinea and New Zealand and has published student text
material in: the UK, the USA, Papua New Guinea, Australia, and New Zealand.
Jinfa Cai is a Professor of Mathematics and Education. He is interested in how
students learn mathematics and solve problems, and how teachers can provide and
create learning environments so that students can make sense of mathematics. He
received a number of awards, including a National Academy of Education Spencer
Editors and Contributors 615
Fellowship, an American Council on Education Fellowship, an International Re-

search Award, and a Teaching Excellence Award. He has been serving on the Ed-
itorial Boards for several international journals, such as the Journal for Research
in Mathematics Education. He was a visiting professor in various institutions, in-
cluding Harvard University. Currently, he serves as a Program Director at the U.S.
National Science Foundation (2010–2011) and a co-chair of American Educational
Research Association’s Special Interest Group on Research in Mathematics Educa-
tion (AERA’s SIG-RME) (2010–2012).
Thomas Carpenter is Emeritus Professor of Curriculum and Instruction (Mathe-
matics Education) at the University of Wisconsin-Madison and Director of Diver-
sity in Mathematics Education—Center for Learning and Teaching. He served as
Director of the National Center for Improving Student Learning and Achievement
in Mathematics and Science and as editor of the Journal for Research in Mathemat-
ics Education. His research integrates the study of the development of children’s
mathematical thinking, instruction that supports that development, and professional
development that fosters instruction that leads to learning with understanding. His
recent research focuses on the development of algebraic thinking in the elementary
school, in particular the development of relational thinking, generalization, mathe-
matical representations, and proof.
M. A. (“Ken”) Clements has been Professor in the Mathematics Department at
Illinois State University since 2005. After teaching in elementary and secondary
schools, he then taught in four Australian universities (Monash, Deakin, Newcastle,
and Victoria), and at Universiti Brunei Darussalam. He served as a consultant in
Brunei, China, India, Malaysia, Papua New Guinea, South Africa, Thailand, and
Vietnam. He co-edited two international handbooks of mathematics education and
will be editor-in-chief for the forthcoming third handbook. He has written or edited
24 books and more than 200 articles, and co-authored a UNESCO book (with Nerida
Ellerton) on mathematics education research.
Annalisa Cusi is a young PhD in Mathematics. In her research she faces the prob-
lem of promoting a renewal in the didactic of algebra at secondary school level.
She, in particular, analyzes the teaching-learning processes from the point of view
of the interactions between teacher and students, highlighting teachers’ attitudes and
behaviours which turn out to be suitable for helping students develop a conscious
vision of algebraic language. Currently one of her main research interests is to an-
alyze how teachers’ participation in research projects affects the evolution of their
systems of conceptions and convictions and fosters their professional growth.
Nerida F. Ellerton has been Professor within the Mathematics Department at Illi-
nois State University since 2002. She holds two doctoral degrees—the first in phys-
ical chemistry and the second in mathematics—and has taught in elementary and
secondary schools and at four universities. She was Dean of Education at the Univer-
sity of Southern Queensland (1997–2002), and has served as consultant and led re-
search projects in numerous countries, including Australia, New Zealand, the United
States, China, Malaysia, the Philippines, Brunei, Thailand, and Bangladesh. She has
written or edited 12 books and has had more than 150 articles published in refereed
journals or books.
Amy Ellis is assistant professor of mathematics education at the University of
Wisconsin-Madison. Dr. Ellis conducts research on students’ reasoning, particularly
as it relates to mathematical generalization, proof, and the development of algebraic
thinking. Her current research is supported by two NSF-funded projects that focus
on (a) investigating relationships between students’ quantitative reasoning and their
proof practices, and (b) studying students’ inductive and deductive reasoning about
problems in mathematics and the natural world. Dr. Ellis teaches a range of under-
graduate and graduate courses in mathematics education and enjoys working with
both pre-service and in-service mathematics teachers.
Susan B. Empson is an Associate Professor of Mathematics and Science Educa-
tion at The University of Texas at Austin. Her research on the learning and teach-
ing of fractions is the topic of her forthcoming book from Heinemann, Extending
Children’s Mathematics: Fractions and Decimals, co-authored with Linda Levi. Her
work has been supported by the National Science Foundation and the Spencer Foun-
dation, and published in such journals as Cognition and Instruction, Journal for Re-
search in Mathematics Education, Teaching Children Mathematics, and Journal of
Mathematics Teacher Education. She earned her PhD in Mathematics Education at
the University of Wisconsin-Madison. Before going back to graduate school, she
was a high school mathematics teacher in New York City and in the Peace Corps, in
Morocco.
Kathryn Irwin was a Senior Lecturer at the University of Auckland, now retired.
Initially from the United States, she has degrees from Carleton College, Minnesota,
Harvard University, and the University of Auckland. Her specialty has been re-
search and teaching of mathematics education; in particular, researching how chil-
dren learn, what causes difficulties in their understanding of mathematics, and how
these difficulties can be overcome. Her research has covered how students learned
between the ages of 4 and 15. She has been engaged in joint research with Murray
Britt for about 25 years.
Andrew Izsák received his Ph.D. in Science and Mathematics Education from the
University of California, Berkeley. Currently, he is Associate professor of mathe-
matics education at the University of Georgia. His research focuses on teachers’ and
students’ understanding and use of inscriptions as representations in mathematical
problem solving. He has concentrated on algebra and fractions as the mathematical
contexts for his investigations. In service of this research, he also has interest in
developing new methods for studying teaching and learning in classroom settings.
Jennifer Jacobs (Ph.D., UCLA, 1999) is a research faculty associate in the Insti-
tute of Cognitive Science at the University of Colorado at Boulder. Her areas of
specialization are teaching practices and beliefs and mathematics professional de-
velopment.
Carolyn Kieran is Professor Emerita of Mathematics Education at the Université
du Québec à Montréal. Her research focus is the teaching and learning of algebra,
with special emphasis on the use of computing technology. The results of her re-
search have been published in over 150 articles and book chapters. Her contributions
to the international mathematics education research community include serving as
President of the International Group for the Psychology of Mathematics Education,
as a member of the Mathematics Learning Study committee which published Adding
it up, and as a member of the Board of Directors of the National Council of Teachers
of Mathematics.
Jeremy Kilpatrick was appointed Regents Professor in 1993. Before joining the
faculty at Georgia in 1975, he taught at Teachers College, Columbia University. He
served two terms as Vice President of the International Commission on Mathemati-
cal Instruction. In 2007, he received the Felix Klein Medal from the ICMI honoring
lifetime achievement in mathematics education. He also received the 2003 Lifetime
Achievement Award from the National Council of Teachers of Mathematics. He is
a National Associate of the National Academy of Sciences, a Fellow of the Ameri-
can Educational Research Association, and a Member of the National Academy of
Education.
Eric Knuth is a professor of mathematics education at the University of Wisconsin-
Madison. His research concerns the meaningful engagement of students in mathe-
matical practices and the design of curriculum and instruction that foster the devel-
opment of increasingly more sophisticated ways of engaging in these practices. By
mathematical practices, he includes such practices as justifying and proving mathe-
matical claims and using algebraic representations appropriately, flexibly, and effi-
ciently to model and to solve problems. His work has been published in the Journal
for Research in Mathematics Education, Journal of Mathematics Teacher Educa-
tion, Cognition and Instruction, and the Mathematics Teacher, among others.
Karen Koellner (Ph.D., Arizona State University, 1998) is an associate professor
in Mathematics Education at Hunter College, City University of New York. Her
research focuses on mathematics teacher development in conjunction with students’
mathematical thinking.
Kyeong-Hwa Lee is Associate Professor at Seoul National University. Her research
and teaching interests embrace mathematics education, and particularly gender is-
sues, gifted education, probability and statistics education, analogical reasoning, and
teacher education. She has been in charge of several research projects; for example,
WISE (Women into Science and Engineering Project), APDMT (Advance Plan-
ning for Development of Mathematics Textbook), and Improvement of Teaching
and Learning Methods Course in Mathematics to Establishing Practical Perspec-
tive.
Linda Levi is The Director of Cognitively Guided Instruction (CGI) Professional
Development Initiatives at Teachers Development Group. She is especially inter-
ested in helping teachers and other educators develop the capacity to lead CGI
professional development to teachers in their region. She is working on state-wide
mathematics teacher professional development initiatives Arkansas and Iowa De-
partments of Education. Her research on the learning and teaching of fractions is
the topic of her forthcoming book from Heinemann, Extending Children’s Mathe-
matics: Fractions and Decimals, co-authored with Susan Empson. She earned her
PhD in Mathematics Education at the University of Wisconsin-Madison. Before go-
ing back to graduate school, she was a high school mathematics teacher in Buffalo,
NY.
Jing Li is a doctoral candidate at the School of Mathematics and Statistics, South-
west University. He is also an associate professor at the School of Mathematics and
Information of Langfang Teachers’ College. His research interests include teach-
ing mathematics with variation, mathematics curriculum reform, and strategies of
mathematics teaching.
Nicolina A. Malara is full professor of Mathematics Education at Modena & Reg-
gio E. University (Italy). She is author of about 150 publications including some
monographic studies. Since 1994 she addressed the question of the renewal of the
teaching of algebra. Her studies have been devoted to an early linguistic and socio-
constructive approach to algebra as a tool for thinking. Currently her studies are
focused on the teacher’s role in leading the students to face modelling and proving
activities and are aimed at the individuation of methods and tools for promoting the
teachers development in this new teaching field.
John Mason was retired by the Open University in 2009 having taught mathematics
for some 50 years. He spent 40 years at the Open University writing materials for
distance taught courses for 10 years in mathematics, and for 30 years in mathematics
education at all phases from primary to tertiary and undergraduate to PhD. His core
interest is in thinking mathematically, and supporting those who wish to foster and
sustain mathematical thinking in others. He has published widely, usually promoting
a phenomenological approach to capturing the lived experience of mathematical
thinking so as to be sensitised to the struggles of others. He is a Senior Research
Fellow at the Department of Education at the University of Oxford, and professor
emeritus at the Open University.
Susan London McNab is a PhD candidate at the Ontario Institute for Studies in Ed-
ucation, University of Toronto, in primary mathematics education with a focus on
young children’s generalizing and early algebra learning. Other research interests
include mathematical modeling and girls’ mathematics learning. Susan is a former
elementary and intermediate mathematics classroom teacher with a diverse back-
ground that includes studies in music and architecture. She is currently a lecturer at
Ryerson University in Toronto.
Nicole McNeil is an assistant professor of psychology at the University of Notre
Dame. She is interested in the mechanisms that propel and constrain the develop-
ment of problem solving, quantitative reasoning, and symbolic understanding. Two
questions motivate her work: (1) Why are some domains of knowledge, such as
mathematics, so difficult for children (and adults) to learn? and (2) How do domain
experience and practice affect learning and problem solving? She is interested in
theoretical issues related to the construction and organization of knowledge, as well
as practical issues related to learning and instruction.
Francesca Morselli graduated in Mathematics at the University of Genova (2002)

and obtained her PhD in Mathematics at the University of Torino (2007), with a
thesis in mathematics education on the role of students’ mathematical culture in
proving. She is research assistant at the University of Genova. Her research inter-
ests concern proof, affective factors influencing mathematics teaching and learning,
teacher professional development. She currently works in a research project, whose
aim is to develop and experiment didactical activities with a strong argumentative
component. Since 2004, she is involved in programs of pre and in-service teacher
education.
Joan Moss is an Associate Professor at the Department of Human Development
and Applied Psychology, Ontario Institute for Studies in Education, University of
Toronto. Moss a former elementary school teacher obtained a Ph.D. in at OISE in
2000, where she worked with Robbie Case on the development of students’ under-
standing of rational number. Recently, she has worked on innovative approaches to
support young students’ generalizations. Other research foci include metacognition
and collaborative learning, computer supported knowledge building in mathemat-
ics and adaptations of Lesson Study for professional development. Joan is a board
member of For the Learning of Mathematics and a co-author of the mathematics
textbook series Real Math 2007.
John (Jack) Moyer is a professor of mathematics specializing in mathematics edu-
cation in the Department of Mathematics, Statistics and Computer Science at Mar-
quette University. He received his M.S. in mathematics in 1969, and his Ph.D. in
mathematics education in 1974, both from Northwestern University. Since 1980 he
has been an investigator or director of more than 70 private- and government-funded
projects. The majority of the projects have been conducted in formal collabora-
tion with the Milwaukee Public Schools to further the professional development of
Milwaukee-area middle school teachers and the mathematics development of their
students.
Giancarlo Navarra has worked in the field of Educational Mathematics as a Junior
High School Teacher-Resercher and as a Contract Professor at the Science Fac-
ulty of the University of Modena and Reggio Emilia (Italy). His main topic of re-
search involves the teaching and learning of arithmetic and algebra in Elementary
and Middle Schools. He is scientific responsible, together with N. Malara, of the
ArAl project, devoted to a linguistic and early approach to algebra, of which he
coordinates the net of experimentations at national level. He is author of more than
hundred publications regarding this field, several of which also at international level.
Swee Fong Ng is Associate Professor of Mathematics Education at the National In-
stitute of Education, Nanyang Technological University, Singapore. She works ex-
tensively with both pre-service primary mathematics teachers as well as in-service
mathematics teachers. Her other responsibilities include teaching and supervising
at the master and doctoral level. Her general interest is looking at ways to help im-
prove the teaching and learning of mathematics across the curriculum. The teaching
and learning of algebra is her special interest. Currently she is involved in a longi-
tudinal study investigating the development of algebraic thinking, beginning with

kindergartners.
Bikai Nie is a research associate and a part-time student in the Statistics Program for
a Master’s degree at the University of Delaware. Previously, he taught at the School
of Mathematics and Statistics in Central China Normal University. He has ten years
of experience teaching mathematics in both middle school and high school. He re-
ceived his Ph. D. in Mathematics Education from East China Normal University.
Aihui Peng is an associate professor at Southwest University. She is currently a
postdoctoral fellow at Umea Mathematics Education Research Centre, Umea Uni-
versity. She received her PhD in mathematics education in 2007. Her research fo-
cuses on teacher knowledge for analyzing students’ mathematical errors, error anal-
ysis, and influence of different cultures on mathematics teaching and learning.
Luis Radford is full professor at Laurentian University, in Sudbury Ontario,
Canada. He teaches at École des science de l’éducation, in the pre-service teachers’
training program. He conducts classroom research with teachers from Kindergarten
to Grade 12. His research interests include the development of algebraic thinking,
the relationship between culture and thought, the epistemology of mathematics, and
semiotics. He has been co-editor of two special issues of Educational Studies in
Mathematics, one on the history of mathematics and mathematics education and the
other on multimodality and gestures. He co-edited the book Semiotics in mathemat-
ics education: epistemology, history, classroom, and culture, published by Sense
Publishers. He received the Laurentian University 2004-05 Research Excellence
Award.
F. D. Rivera is an Associate Professor in the Department of Mathematics at San Jose
State University. He conducts research in algebraic thinking at the elementary and
middle school levels. He has published widely in the area of pattern generalization
and has a book-in-press entitled, Toward a Visually-Oriented School Mathematics
Curriculum, that will be published by Springer.
Sarah A. Roberts (Ph.D., University of Colorado at Boulder, 2009) is an assistant
professor in Mathematics Curriculum and Instruction at Iowa State University. Her
research focuses on equity in mathematics, particularly related to supporting English
language learners.
Susan Jo Russell is a principal scientist at the Education Research Collabora-
tive at TERC, Cambridge, MA, where she has worked for over 25 years directing
projects focused on computer education, mathematics for special needs students,
professional development in mathematics, research on students’ and teachers’ un-
derstanding of mathematics, and curriculum design for elementary students. She
co-directed the development and revision of the NSF-funded elementary curricu-
lum, Investigations in Number, Data and Space and, with Deborah Schifter and
Virginia Bastable, the professional development materials, Developing Mathemati-
cal Ideas. With Schifter and Bastable, she is developing a book and on-line course
for teachers in grades 1–6, Connecting Arithmetic and Algebra.
Deborah Schifter is a principal research scientist at the Education Development

Center, Newton, MA. She has worked as an applied mathematician; has taught
elementary, secondary, and college level mathematics; and, since 1985, has been
a mathematics teacher educator and educational researcher. She authored Recon-
structing Mathematics Education: Stories of Teachers Meeting the Challenge of Re-
form and edited a two-volume anthology of teachers’ writing, What’s Happening in
Math Class? She was a writer for The Mathematical Education of Teachers as well
as the second edition of the K-5 curriculum, Investigations in Number, Data, and
Space. Dr Schifter produced, with Virginia Bastable and Susan Jo Russell, a pro-
fessional development curriculum series for elementary and middle-grade teachers
called Developing Mathematical Ideas.
Jean Schmittau holds a Ph.D. from Cornell University in Educational Psychology,
Cognitive Development, and Mathematics, and is a professor at the State Univer-
sity of New York at Binghamton. She is the editor of Investigations in Mathematics
Learning and directs the Teacher Leader Quality Partnership, a project to improve
mathematics instruction in New York state. Her interest in Vygotskian psychology
led to research in Russia with students who had been instructed with V.V. Davy-
dov’s mathematics curriculum, and subsequently to editing its English translation
and researching its first implementation in a U.S. school setting.
Craig Schneider (Ph.D., University of Colorado at Boulder, 2008) is a Senior In-
structor with CU Teach, the University of Colorado at Boulder’s secondary math
and science teacher licensure program. His interests including teacher professional
development and discourse practices of linguistically diverse students.
Margaret Smith is an Associate Professor at Iona College, where she teaches un-
dergraduate and graduate mathematics education classes. She has contributed to the
TIMSS-R Video Study and consulted for the LessonLab Research Institute, includ-
ing an on-line professional development platform for teachers and research for the
ALFA (Algebra Learning for All) professional development program. Her research
focuses on understanding how students engage in mathematical processes and how
teachers help achieve this student engagement.
Naiqing Song is the executive vice-president of Southwest University and the di-
rector of the Center of Southwest Basic Education of the Ministry of Education. He
has authored or coauthored more than 10 books and 70 articles, and edited 8 series
of mathematics textbooks. He received National Distinguished Teacher Award in
2009, the first prize for Outstanding Teaching Achievement in 2009, and the first,
second and third prizes for Research in Humanities and Social Sciences in Higher
Education in 2009, 2006, 1989. He specializes in curriculum reform, textbook writ-
ing, and educational experiments.
Bharath Sriraman is Full Professor of Mathematics at The University of Montana
with an honorary appointment in the Faculty of Central/Southwest Asian Studies.
He has published 250+ journal articles, books, book chapters, reviews and commen-
taries in mathematics education, educational philosophy , gifted/talented education
and mathematics. He is the series Co-editor of Advances in Mathematics Education
(Springer). Among his recent major works include Theories of Mathematics Educa-
tion: Seeking New Frontiers (Springer, 2010) and The First Sourcebook on Nordic
Research in Mathematics Education (Information Age Publishing, 2010).
Ana Stephens is an associate professor at the Wisconsin Center for Education Re-
search at the University of Wisconsin-Madison. She is interested in the development
of algebraic reasoning. Her current research examines the development of algebraic
thinking in grades 3–7, focusing in particular on the development of curricular learn-
ing progressions. Her work has been published in the Journal for Research in Math-
ematics Education, Journal of Mathematics Teacher Education, and the Mathemat-
ics Teacher, among others.
K. Subramaniam is associate professor of mathematics education at the Homi
Bhabha Centre for Science Education, Tata Institute of Fundamental Research in
Mumbai, India. His main areas of work currently are teaching fractions for under-
standing ratio and proportion, symbolic arithmetic and beginning algebra, profes-
sional development of mathematics teachers, and the role of visuo-spatial thinking
in learning science and mathematics. He also has interest in cognitive science and
philosophy, especially in relation to education and to maths learning. He has con-
tributed to the development of the national curriculum framework in mathematics,
and to the development of mathematics textbooks at the primary level.
Ning Wang received her Ph.D. in Educational Measurement and Statistics from the
University of Pittsburgh. She also received a master’s degree in Research Method-
ology and another master’s degree in mathematics education. Currently, she is am
Associate Professor at the Center for Education, Widener University, teaching re-
search methodology courses at the Master’s and Doctoral level. Dr. Wang has ex-
tensive experience in the validation of assessment instruments in various settings,
scaling using Item Research Theory (IRT), and conducting statistical data analysis
using Hierarchical Linear Modeling and Structural Equation Modeling. In particu-
lar, she is interested in the applications of educational measurement, statistics, and
research design techniques into the exploration of the issues in the teaching and
learning of mathematics.
Elizabeth Warren is a professor and an academic with an ardent commitment to
academic and research leadership, and service to others. She has an excellent track
record with respect to research projects in mathematics education and particularly in
school algebra. Her PhD investigated Interactions between instructional approaches,
students’ reasoning processes and their understanding of elementary algebra. She
has presented several invited International Keynote addresses on this area of re-
search. In the last five years from her research in the algebraic domain she has been
sole author on 14 refereed papers and co-authored 12 refereed papers, and a text
book series for use in the elementary school (Algebra for all, Origo Press).
Tad Watanabe is an Associate Professor of Mathematics and Mathematics Educa-
tion in the Department of Mathematics & Statistics at Kennesaw State University,
Kennesaw, Georgia. His research interests include teaching and learning of multi-
plicative concepts such as fractions and proportions, the development of mathemat-
ical knowledge for teaching in prospective elementary school teachers, and various
Japanese mathematics education practices, including curriculum materials and les-
son study. He has published a number of articles both in English and in Japanese. At
Kennesaw State University, he teaches mathematics content courses for prospective
elementary school teachers.
Aaron Weinberg is an assistant professor of mathematics at Ithaca College, where
he teaches in the mathematics department and graduate program in education. His
research interests include studying students’ personal semiotic systems and the ways
they contrast with cultural semiotic systems in algebra; using ideas from reader-
oriented theory to analyze calculus texts; and studying the development of students’
statistical reasoning, including their conceptions of sampling distributions and their
informal inferential reasoning.

Early Algebraization

Uploaded by

Copyright:

Available Formats

Early Algebraization

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Early Algebraization

Uploaded by

Copyright:

Available Formats

Advances in Mathematics Education

For other titles published in this series, go to

ISSN 1869-4918 e-ISSN 1869-4926

Library of Congress Control Number: 2011922614

© Springer-Verlag Berlin Heidelberg 2011

Cover design: deblik

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Early Algebraization: A Global Dialogue from Multiple Perspectives is the second

Acknowledgements We gratefully acknowledge the assistance of the National Science Founda-

Part I: Curricular Perspective

Functional Thinking as a Route Into Algebra in the Elementary Grades . 5

Developing Students’ Algebraic Thinking in Earlier Grades: Lessons

Lessons from Chinese and Singaporean School Mathematics . . . . 34

Developing Algebraic Thinking in the Context of Arithmetic . . . . . . . 43

The Role of Theoretical Analysis in Developing Algebraic Thinking:

The Arithmetic-Algebra Connection: A Historical-Pedagogical Perspective 87

Shiki: A Critical Foundation for School Algebra in Japanese Elementary

Commentary on Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Part II: Cognitive Perspective

Preface to Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Algebraic Thinking with and without Algebraic Representation:

Examining Students’ Algebraic Thinking in a Curricular Context:

Highlights of the Differences between CMP and Non-CMP

Years 2 to 6 Students’ Ability to Generalise: Models, Representations

Algebra in the Middle School: Developing Functional Relationships

Fostering a Focus on Quantities . . . . . . . . . . . . . . . . . . . 234

Representational Competence and Algebraic Modeling . . . . . . . . . . 239

Middle School Students’ Understanding of Core Algebraic Concepts:

An Approach to Geometric and Numeric Patterning that Fosters Second

Numeric Representations: Function Machine . . . . . . . . . 282

Grade 2 Students’ Non-Symbolic Algebraic Thinking . . . . . . . . . . . 303

Formation of Pattern Generalization Involving Linear Figural Patterns

Cognitive Issues Surrounding Pattern Generalization: What We

Commentary on Part II . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

Part III: Instructional Perspective

Preface to Part III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

Prospective Middle-School Mathematics Teachers’ Knowledge

The Algebraic Nature of Fractions: Developing Relational Thinking

Professional Development to Support Students’ Algebraic Reasoning:

Using Habermas’ Theory of Rationality to Gain Insight into Students’

Research Advances . . . . . . . . . . . . . . . . . . . . . . . 477

Theoretical Issues and Educational Strategies for Encouraging Teachers

A Procedural Focus and a Relationship Focus to Algebra: How U.S.

Teaching Algebraic Equations with Variation in Chinese Classroom . . . 529

The Improvement of Understanding of Equation . . . . . . . 535

Commentary on Part III . . . . . . . . . . . . . . . . . . . . . . . . . . . 557

Overall Commentary on Early Algebraization: Perspectives for

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

Editors and Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . 615

The Role of Theoretical Analysis in Developing Algebraic Thinking:

The Arithmetic-Algebra Connection: A Historical-Pedagogical Perspective 87

Shiki: A Critical Foundation for School Algebra in Japanese Elementary

Commentary on Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Jinfa Cai and Eric Knuth

Although it is widely accepted that we should expect students in early grades to

J. Cai, E. Knuth (eds.), Early Algebraization, Advances in Mathematics Education, 3

Maria L. Blanton and James J. Kaput

J. Cai, E. Knuth (eds.), Early Algebraization, Advances in Mathematics Education, 5