Adams, G. (Ed.

) Proceedings of the British Society for Research into Learning Mathematics 36 (1) February 2016

Developing algebraic language in a problem solving environment: the role of

teacher knowledge

Abraham de la Fuente1, Tim Rowland2, Jordi Deulofeu1,

1
Universitat Autònoma de Barcelona, Spain; 2University of East Anglia, UK

This paper describes a teaching sequence designed by a team of three

teachers in Spain to enable a group of 13 to 14-year-old students to
develop algebraic language through problem solving. Problems are
introduced which provoke the thinking needed to solve systems of linear
equations, without formal instruction in standard methods. We consider
the mathematics-related knowledge that the teachers used while
implementing these tasks, using the Knowledge Quartet (KQ) model to
analyse this knowledge. In particular, we show how the connections that
the teachers make between different representations of the same concept
are key for the students to acquire algebraic language as one way to solve
certain problems.

Keywords: algebra, problem solving, teacher knowledge, Knowledge Quartet

Introduction

The research reported in this paper is taking place in a British school (following the
English national curriculum) in Barcelona, and involves the four mathematics
teachers (including the first author) who teach in the two first years of the secondary
stage. The school has three classes in each year group, taught by three teachers
teaching at the same time. Instruction is planned jointly by the team, and this situation
allows us to investigate how three different teachers implement the same material
with comparable classes.
Specifically, this research aims to understand how the teachers use their
knowledge to help students to learn to use algebraic language in a problem-solving
environment. The teachers involved agree that learning is a process in which the
learner has the autonomy to construct their own knowledge, and that teachers act as a
guide in that process.
One feature of the mathematics department is that they do not use a textbook,
but instead they design their own activities or they adapt existing material. In this
scenario there are not themes labelled “algebra”, or “linear equations”. However, the
teachers help students to construct the algebraic language inherent in units focussed
on solving a problem in a particular context, performing an investigation, or
producing a mathematical model.
This article describes the teachers’ planning, one of the learning activities, and
an analysis of one lesson, focusing in particular on how the teacher makes
connections between different representations in order to help students to solve a
particular type of problem. The Knowledge Quartet (Rowland, Turner, Thwaites &
Huckstep, 2009; Rowland, 2014) will provide a language with which to discuss the
mathematics teaching practice, with a focus on the teacher and their mathematical
knowledge in teaching.

From Conference Proceedings 36-1 (BSRLM) available at bsrlm.org.uk © the author - 1

Adams, G. (Ed.) Proceedings of the British Society for Research into Learning Mathematics 36 (1) February 2016

Analytical framework: the Knowledge Quartet (KQ)

The Knowledge Quartet identifies three categories of situations in which

teachers’ mathematics-related knowledge is revealed in the classroom: named
foundation, transformation, connection and contingency. Table 1 outlines these and
their contributory codes which arose from grounded analysis of mathematics
classroom data (Rowland et al, 2014). Each dimension is composed of a small number
of related subcategories (i.e. open codes).
Dimension Contributory codes
Foundation: awareness of purpose; adheres to textbook;
knowledge and understanding of mathematics per se and concentration on procedures; identifying
of mathematics-specific pedagogy, beliefs concerning errors; overt display of subject knowledge;
the nature of mathematics, the purposes of mathematics theoretical underpinning of pedagogy; use
education, and the conditions under which students will of mathematical terminology
best learn mathematics
Transformation: choice of examples; choice of
the presentation of ideas to learners in the form of representation; use of instructional
analogies, illustrations, examples, explanations and materials; teacher demonstration
demonstrations
Connection: anticipation of complexity; decisions about
the sequencing of material for instruction, and an sequencing; making connections between
awareness of the relative cognitive demands of different procedures and between concepts;
topics and tasks recognition of conceptual appropriateness
Contingency: deviation from agenda; responding to
the ability to make cogent, reasoned and well-informed students’ ideas; use of opportunities;
responses to unanticipated and unplanned events teacher insight during instruction
Table 1: The Knowledge Quartet – dimensions and contributory codes

School algebraic language

Our conception of teacher knowledge includes knowing what to teach, and why
(‘theoretical underpinning of pedagogy’: Rowland et al., 2009), and how to design
tasks for learning. In the light of the team’s ‘bottom up’ approach to planning
instruction, it was necessary to agree what algebra is, and what students should learn.
Following a literature review on algebraic language construction, it was decided to
use the theoretical background given by Kaput (2000) and the NCTM (2000). In
particular, algebraic school language enables students:
1. to communicate generalisations and formalisations;
2. to represent abstract structures and to reason with them;
3. to understand functions and others relations between variables;
4. to carry out mathematical modelling.
Given their agreement about these components, the teachers were clear about
the necessary objectives of the activities to be designed to introduce algebraic
language to the students. It was decided to investigate one of the activities (described
later) that the teachers had prepared to implement in their classrooms, with a
particular focus on the second and third algebraic language dimensions listed above.

Data collection and selection

During the 2014-15 academic year, all of the weekly department meetings were
video-recorded, and also all the lessons of these four teachers whose objectives
included learning some of the algebraic language dimensions.

From Conference Proceedings 36-1 (BSRLM) available at bsrlm.org.uk © the author - 2

Adams, G. (Ed.) Proceedings of the British Society for Research into Learning Mathematics 36 (1) February 2016

A set of activities implemented in the second year of secondary school

(student age 13 to14) were selected for analysis, in which the objective was to
develop the skills needed to solve simultaneous linear equations. The students
involved in the research had had no prior instruction on solving equations, but they
were accustomed to use algebra to communicate generalisations (Kaput, 2000).

Characteristics of the learning activities

The algebra-oriented learning activities devised for the second year classes had the
characteristics of problem solving tasks in the sense of Schoenfeld (1992): i.e.,
students did not already know procedures to solve the problems directly, but they had
sufficient resources and heuristics to attempt to resolve them. The intention was that
students learn how to solve linear equations by solving relevant problems and making
connections between iconic, algebraic and tabular representations of key information.

The pizzas and drinks problem

Figure 1 presents one of the problems within a sequence of tasks, presented

iconically. Students are required to find the price of one pizza and one drink using the
given information. Before presenting the problem as a whole, the teachers first
showed only one of the two conditions (e.g., that 3 pizzas and 3 drinks cost 12€), and
asked the students what more they could say about it. For example, some students said
that if they knew the price of one pizza, they could find the price of a drink; and
conversely. Other students then gave a list of possible price-solution pairs,
representing them in a table or a graph.

Figure 1: The pizzas and drinks problem: What is the price of one drink? Of one pizza?
Figure 2 shows how students solved the problem in different ways. Student 1
used proportional reasoning in the first condition, dividing by two, so that two pizzas
and three drinks cost 8.80€. Then she compared this information with the second
condition, and concluded that one pizza cost 12€ - 8.8€ = 3.20€. It then follows that
the price of one drink is 0.80€. She expressed the solution in the iconic way, and she
explained the solution orally to the class. Student 2 presented her solution in words,
and began by using the second condition to conclude that one drink and one pizza
together cost 4€. Then, in the first condition, she could identify four drink-plus-pizza
‘groups’, costing 16€. Therefore she knew that two drinks cost 17.60€ - 16€ = 1.60€,
and so one drink costs 0.80€. Student 3’s solution uses the same reasoning, but with
iconic representation.
Problems such as this one were designed to promote mathematical activities
close to the students’ own common sense and everyday meanings. The teachers also
wanted their students not to be constrained by rigid use of mathematical tools and to
give students the opportunity to construct their own ways to solve problems
(Espinoza, Barbé & Gálvez, 2009).

From Conference Proceedings 36-1 (BSRLM) available at bsrlm.org.uk © the author - 3

Adams, G. (Ed.) Proceedings of the British Society for Research into Learning Mathematics 36 (1) February 2016

Figure 2: Three different solutions of the pizzas and drinks problem

The ‘Chewbacca problem’

After several lessons working on problems like the one described above, the teachers
devised a ‘test’ for the students, consisting of the three problems shown in Figure 3.
This was the first occasion on which students were required to work with systems of
equations (problems 2 and 3) conventionally represented in algebraic symbolism. The
icons used in the first problem are difficult to draw. Therefore it might appeal to the
students to change the icons for something else: another icon they could draw quickly,
or perhaps a letter. This in turn might help them to see that the second problem is
isomorphic to the first.

Figure 3: The test items given to the students

Lesson analysis

We now analyse how one of the teachers helped students to solve the third problem
shown in Figure 3. The students had had twenty minutes to solve the three problems
individually, and this was followed by a whole-class discussion about how to solve

From Conference Proceedings 36-1 (BSRLM) available at bsrlm.org.uk © the author - 4

Adams, G. (Ed.) Proceedings of the British Society for Research into Learning Mathematics 36 (1) February 2016

the first problem, and concluding that problem 2 is the ‘same’ problem as problem 1.
The following analysis will use the dimensions and codes of the Knowledge Quartet
(Rowland et al., 2009). On the basis of this analysis, we propose an additional code in
the Connection dimension of the KQ: making connections between representations.

Decisions about sequencing - Making connections between representations

The teacher began writing the equations in problem 3 on the whiteboard, and tried to
connect that problem with the problems that they solved using iconic representations:
Teacher: We saw in previous lessons how through some information, prices, etc., we
could deduce the price of each thing, didn’t we? And we saw that this could be identified
with another kind of representation [referring to the algebraic representation].
In particular, he connected problem 3 with problem 1, because (as we shall
show) he wanted to give meaning to the letters ‘a’ and ‘b’, not because he wanted to
connect the procedures.

Choice of representations - Theoretical underpinning of pedagogy

The teacher wanted to introduce a new procedure to solve the problems. He had read
the student solutions to the test, and seen how one of the students solved the problem.
He asked that student to go to the whiteboard and explain his solution. The student
called the method “by trying”. For example, he tried replacing ‘a’ by 4. The teacher
asked what ‘a’ meant to him, and several students said the price of a pizza, or a
Chewbacca, for example. It appeared that that teacher believed that students learn by
constructing their own knowledge. He asked questions about the meaning of the
letters, connecting that problem with the problems about prices.

Identifying errors - Theoretical underpinning of pedagogy

As the student at the whiteboard replaced ‘a’ by 4, he deduced from the first equation
that ‘b’ would then be 10, and wrote: “4x2 = 8+10 = 18”. The teacher decided to
correct the error in the use of the equals symbol, believing that correct use of “=” is
important when students solve equations.

Deviation from agenda

Before the equals symbol correction incident, the student had written wrote on the
whiteboard: 4x2 = 8+10 = 18
4x4 = 16 + ___ = 29
He had found that ‘b’ would need to be 10 in the first equation, but that b=10
would not satisfy second equation. Earlier, in correcting the written test scripts, the
teacher had thought that student was simply trying two random numbers for ‘a’ and
‘b’. The student’s method was more powerful than the teacher had thought. What the
student called “by trying” was in effect a form of trail-and-improvement: “I was
trying values for ‘a’ and calculating the corresponding value of ‘b’ in the first
equation. If that value of ‘a’ and ‘b’ satisfy the second equation, then I have the
solution”.
This was clearly a contingency episode. The teacher had interpreted the
student’s written resolution incorrectly, was surprised in the classroom, and had to
revise his thinking accordingly.

From Conference Proceedings 36-1 (BSRLM) available at bsrlm.org.uk © the author - 5

Adams, G. (Ed.) Proceedings of the British Society for Research into Learning Mathematics 36 (1) February 2016

Making connections between representations - Use of instructional materials

The teacher could have used the student intervention to present the representation
of solution-pairs in two tables. Instead, he decided to ask the student to write the values
that he tried in his own way. Figure 4 compares the two representations.

Figure 4: Teacher representation vs Table of values representation

Conclusion
In solving the ‘pizzas and drinks’ problem, the students used different ways of reasoning,
and also various representations: iconic (students 1 and 3) and symbolic (student 2).
The teacher uses these types of problems to provoke student reflection about
the meaning of the letters in linear equations, and tries to show a different process that
also works to solve those problems, thereby helping them to connect different
representations by considering the meaning of the letters. It would have been
advantageous if the teacher had understood the student’s method more readily, and
exploited the situation to show yet another connection between representations: the
algebraic formulation of the equations with the tables of solution-pairs.
The findings of this research show how these teachers use problem solving
activities in the classroom, to make connections between representations, supporting
students in the construction and use of algebraic language. This component of teacher
knowledge is absent so far in the Knowledge Quartet: therefore we propose a new
code within the Connection dimension of the KQ, namely: making connections
between representations.
References
Espinoza, L. Barbé, J.and Gálvez, G. (2009). Estudio de fenómenos didácticos
vinculados a la enseñanza de la aritmética en la educación básica chilena.
Enseñanza de las ciencias, 27(2), pp. 157-168.
Kaput, J. (2000). Teaching and learning a new algebra with understanding.
University of Massachussets-Dartmouth.
National Council of Teachers of Mathematics (2000). Principles and standard for
school mathematics. Reston, Va.
Rowland, T. (2014). The Knowledge Quartet: the genesis and application of a
framework for analysing mathematics teaching and deepening teachers’
mathematics knowledge. SISYPHUS Journal of Education, 1(3), pp. 15-43
Rowland, T., Turner, F., Thwaites, A. and Huckstep, P. (2009). Developing Primary
Mathematics Teaching: reflecting on practice with the Knowledge Quartet.
London: Sage.
Schoenfeld, A. H. (1992). Learning to think mathematically: problem solving,
metacognition, and sense making in mathematics. In D. Grouws (Ed.),
Handbook for Research on Mathematics Teaching and Learning (pp. 334-
370). New York: MacMillan

From Conference Proceedings 36-1 (BSRLM) available at bsrlm.org.uk © the author - 6