Creating Meaning for and with the Graphing Calculator

Helen M. Doerr; Roxana Zangor

Educational Studies in Mathematics, Vol. 41, No. 2. (Feb., 2000), pp. 143-163.

Stable URL:
http://links.jstor.org/sici?sici=0013-1954%28200002%2941%3A2%3C143%3ACMFAWT%3E2.0.CO%3B2-3

Educational Studies in Mathematics is currently published by Springer.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/springer.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org
Thu Jan 17 14:14:48 2008
 HELEN M. DOERR and ROXANA ZANGOR

CREATING MEANING FOR AND WITH THE GRAPHING

CALCULATOR

ABSTRACT. In this study, we seek to describe how the meaning of a tool was co-
constructed by the students and their teacher and how the students used the tool to con-
struct mathematical meaning out of particular tasks. We report the results of a qualitative.
classroom-based study that examined (1) the role, knowledge and beliefs of a pre-calculus
teacher, (2) how students used graphing calculators in support of their learning of math-
ematics, (3) the relationship and interactions between the teacher's role, knowledge and
beliefs and the students' use of the graphing calculator in learning mathematics, and (4)
some constraints of the graphing calculator technology that emerged within the classroom
practice. We fhund five patterns and modes of graphing calculator tool use emerged in
this practice: computational tool, transformational tool, data collection and analysis tool.
visualizing tool, and checking tool. The results of this study suggest that nature of the math-
ematical tasks and the role, knowledge and beliefs of the teacher influenced the emergence
of such rich usage of the graphing calculator. We also found that the use of the calculator
as a personal device can inhibit communication in a small group setting. while its use as a
shared device supported mathematical learning in the whole class setting.

KEY WORDS: graphing calculator, tools, pre-calculus, technology. teacher knowledge

Functions and graphs have been the focus of numerous research studies
over the past decade. The study of students' understanding of the concept
of function, and their abilities to create and interpret graphical representa-
tions, was given strong impetus by the advent of computers and their ready
availability in some classrooms. This led to many computer-based studies
that analyzed students' reasoning with and about linked, dynamic multiple
representations of functions (e.g., Confrey and Doerr, 1996; Leinhardt,
Zaslavsky and Stein, 1990; Moschkovich, Schoenfeld and Arcavi, 1993;
Yerushalmy, 1991). Yet despite the limitations of the graphing calculator
when compared to a full-screen computer program, the calculator's low
cost, portability and ease of use have resulted in its widespread use for
teaching about functions and graphs in secondary schools in the United
States.

L6 Educational Studies in Mathematics 41: 143-163, 2000.

IfqO 2000 Kluuer Academic Publishers. Prinred In the Netherlands.
144 H.M. DOERR AND R. ZANGOR

Over the past decade, secondary mathematics teachers have moved to

adapt the graphing calculator into their practice. The National Council of
Teachers of Mathematics's (NCTM) curriculum standards (NCTM, 1989)
recommend using the graphing calculators to provide students with new
approaches, such as the use of multiple representations, to the investigation
of mathematical ideas. While it might appear that practice has moved in-
dependently of and more quickly than research, there is in fact no shortage
of research studies on the use of the graphing calculator. See Dunham and
Dick (1994), Wilson and Krapfl (1994), and Penglase and Arnold (1996)
for extensive reviews. Many, if not most, graphing calculator studies are
quasi-experimental in design and seek to answer the question of whether or
not graphing calculators are effective in achieving certain instructional ob-
jectives, which are often left unchanged from traditional paper-and-pencil
approaches (Adams, 1997; Quesada and Maxwell, 1994; Ruthven, 1990:
Shoaf-Grubbs, 1994). Many such studies compare the use of the graph-
ing calculator to the use of paper and pencil on the same set of tasks,
giving only limited insight into how and why students use graphing cal-
culators in the instructional context. One such insight is provided in the
work of Ruthven (1990), who found that students developed three distinct
approaches to symbolizing a graph: an analytic-construction approach that
exploits mathematical knowledge, a graphic-trial approach that compares
successive expression graphs with the given graph, and a numeric-trial
approach guided by the coordinates of the given graph.
The relationship between teachers' knowledge and pedagogical strate-
gies and their use of the graphing calculator is largely unexamined. Many
studies (Drijvers and Doorman, 1996; Hollar and Norwood, 1999; Porzio,
1997) do not report or describe the role of the teacher in the classroom
or the teachers' knowledge and skill with the graphing calculator or the
teachers' beliefs about the efficacy of or h n d s of uses of graphing calcu-
lators in mathematics learning. In their review of the research, Wilson and
Krapfl (1994) suggest that there is a need to better understand teachers'
conceptions and a need 'to focus on the qualitative aspects of knowledge
construction of students using graphing calculators' (p. 261). As Penglase
and Arnold (1996) point out in their extensive review of the literature,
many studies fail to distinguish carefully between the tool and the in-
structional context and assessment procedures, leading to inconsistent find-
ings regarding the effectiveness of graphing calculators in the learning of
functions. These reviewers suggest the need for studies which attempt to
address graphing calculator use within particular learning environments.
In a study of teachers' roles when using graphing calculator technology
in a high school pre-calculus class, Farrell (1996) found that there was an
 CREATING MEANING FOR AND WITH THE GRAPHING CALCULATOR 145

overall shift in teachers' roles from task setter and explainer to consult-
ant, fellow investigator, and resource. These shifts occurred within each
class as the teachers moved from not using the graphing calculator and
computer technology to the use of it; the shifts were also accompanied
by a decrease in lecture and an increase in group work. Contrary to these
findings, other research has found that teachers did not change their meth-
ods or approaches when using the graphing calculator in teaching about
transformations of the parabola, except to provide more visual examples
(Simmt, 1997). The teachers in Simmt's study did not use the graphing
calculator to facilitate discussion, encourage students to conjecture or to
prove ideas. The finding that these teachers did not shift their role when
using the graphing calculator may be accounted for in part by the short
period of time they used the technology. 4 to 10 class periods (Simmt.
1997, p. 273). The teachers in Farrell's study (who had shifted their roles)
were observed using the graphing calculator near the end of a year long
course. However, the more important variable in influencing the teacher's
role with the graphing calculator may well have been the teachers' attitudes
and beliefs about mathematics and mathematics education.
This relationship was explored in a study by Tharp, Fitzsimmons and
Ayers (1997) who found that teachers who held a less rule-based view-
point about learning mathematics were more willing to adopt the use of
calculators as an integral part of instruction than those teachers who held a
rule-based view of mathematics learning. These researchers found that the
non-rule-based teachers used more inquiry learning in their classrooms and
that their students freely used the calculators as they wished. The teachers
in Simmt's (1997) study were described as having a view of mathematics
that 'strongly favors 'traditional' approaches to problems' (p. 287), sug-
gesting that they may be more rule-based and less likely to adopt a more in-
tegrated use of graphing calculator technology. The findings of Tharp et al.
suggest that there is an important relationship anlong the teacher's know-
ledge and beliefs, pedagogical strategies, and how students use graphing
calculators.
In this paper, we report the results of a qualitative, classroom-based
study that examined (1) the role, knowledge and beliefs of a pre-calculus
teacher, (2) how students used graphing calculators in support of their
learning of mathematics. (3) the relationship and interactions between the
teacher's role, knowledge and beliefs and the students' use of the graph-
ing calculator in learning mathematics, and (4) some limitations and con-
straints of the graphing calculator technology that emerged within the class-
room practice.
 H.M. DOERR AND R. ZANGOR

2 . THEORETICAL
FRAMEWORK

The theoretical framework guiding this research follows a perspective in

which the psychological aspects of learning are coordinated with the social
aspects though students' interactions with mathematical tasks, with each
other and with their teacher within the social context of the classroom
(Cobb and Yackel, 1996; Meira, 1995; Roth and McGinn, 1997). For our
purposes, critical aspects of this social context include the tools and the
norms for tool usage which emerge as students and teacher interact with
the tool and with each other. It is through these interactions that the mean-
ing of the graphing calculator as a tool for mathematical learning within
the classroom is constructed by both teacher and students. As Hiebert et
al. (1997) have observed, 'Students must construct meaning for all tools.
. . . As you use a tool, you get to know the tool better and you use the tool
more effectively to help you know about other things.' (p. 54, emphasis
added). Hiebert et al. argue that meaning is constructed for the tool as it
is used and that learners construct mathematical meaning w1itlz the tool.
The meaning that is constructed for the tool does not precede the tool's
use in constructing mathematical meaning, but rather the meaning for the
tool and the meaning with the tool are intertwined through their use by
participants with mathematical tasks (Meira, 1998).
As Lave and Wenger (1991) have argued, the meaning of any tool is
intricately tied to the cultural practices within which it is used for some
social purpose. The features of a tool are not something in and of them-
selves, but rather are constituted by the actions and activities of people.
This view has serious consequences for interpreting the results of much of
the research on graphing calculator use, which often appears to take the
features and nature of the tool to be given or self-evident, as the details
of how and why the tool was used are often not reported. In contrast, we
seek to describe the emergence of the features of the tool in the practice of
the participants as they interacted with each other and with mathematical
tasks. For this study, the essential question is how do students and their
teacher interpret and make use of the graphing calculator as a tool that is
part of a specific cultural practice, namely their mathematics class?
In this study, we seek to describe how one teacher's knowledge and
beliefs about the graphing calculator were reflected in her pedagogical
strategies. We then describe how these strategies led to the co-construction,
with the learners, of a particular set of ways in which the graphing calcu-
lator became a tool for mathematical learning. We closely examine how the
students opted to use the graphing calculator to support their mathematical
learning throughout their year-long pre-calculus course. Lastly, we discuss
our findings on how the graphing calculator, as with any tool or technology,
enabled and constrained both pedagogical practices and student learning.

3. METHODOLOGY
AND DATA ANALYSIS

A pre-calculus curriculum based on modeling problems in an enhanced

technology environment (using graphing calculators, calculator-based meas-
urement probes for motion, temperature and pressure, and computer soft-
ware) provided a rich setting for studying the role of the teacher and the
patterns and modes of graphing calculator use by the students and the
teacher. This classroom-based, observational case study of two pre-calculus
classes took place in a suburban school. The classes were taught by the
same teacher and each met for five sessions for a total of 270 minutes per
week. One class had 17 students and the other 14 students, all between
15 and 17 years of age. This selection of a single teacher in two similar
classes allowed us to focus in-depth on the role and knowledge of the
teacher and the interactions with and among the students. The teacher
had 20 years of teaching experience and was skilled in the use of the
graphing calculator. The classes were observed over three units of study
on linear functions (four weeks), exponential functions (ten weeks) and
trigonometric functions (seven weeks).
All of the students had either TI-82 or TI-83 graphing calculators. These
devices are rich in graphing and statistical functionality, although lack-
ing in symbolic algebra capability. In general, the students had used their
calculators for well over a year before taking this course and were quite
familiar with its functions. The classroom was equipped with a computer,
printer and a 'graph link cable' that could be used to transfer pictures of
the calculator graph to the computer for printing. The link cable also could
be used to transfer data and programs between the computer and calcu-
lator, but this feature was rarely used. On the other hand, students readily
transferred data and programs between calculators using the calculator-to-
calculator link cable. The classroom was equipped with a view screen that
allowed the calculator screen (but not the keystrokes) to be projected using
a standard overhead projection unit.
Classroom instructional activities regularly alternated between model-
ing problems investigated by the students within a small group and whole
class discussion for sharing progress, discussing solution methods and ex-
tending results. The instructional tasks were designed so that the students
would create quantitative systems which describe and explain the patterns
and structures in an experienced situation and which can be used to make
predictions about the situation (Doerr and Tripp, 1999). The students were
148 H.M. DOERR AND R. ZANGOR

asked to interpret data, to find meaningful representations of the data (typ-

ically tables, graphs and equations), and to generalize relationships beyond
the particular situation at hand. The concept of the rate of change of a func-
tion and the transformations of exponential and trigonometric functions
were addressed throughout the problem situations.
All class sessions were observed by two or more members of the re-
search team. Extensive field notes, transcriptions of audio-taped group
work, transcriptions of video-taped whole class discussion, and interviews
and planning sessions with the teacher constituted the data corpus for this
study. In an observational study such as this, there are necessarily some
difficulties in observing students' use of the graphing calculator. As Wil-
liams (1993) has pointed out, it is not easy to see the students' calculator
screens, much less to capture the keystrokes they are using. However, the
mathematical investigations were specifically designed to be completed by
small groups of students and then later shared in a whole class discussion.
This format generated discussion within the small groups where students
explained to each other what they did on their calculator, or one student
would show another how to do a particular task, or students would compare
graphical or numerical results. In this way, students' calculator use was
made visible to the researchers. However. there were occasions where it
simply was not clear what the students had done with their calculators. The
use of the graphing calculator in the whole class setting was much easier
to observe, since the teacher and the students regularly used the overhead
projection view screen and explained what they were doing as they were
doing it.
The data analysis was completed in two phases. In the first phase, these
data were analyzed and coded for the patterns and modes of graphing
calculator use by both the teacher and the students throughout the instruc-
tional units. Upon completion of this initial coding, a descriptive profile
of the teacher. including her knowledge, role in the class, and beliefs was
compiled. This profile was shared with the teacher for her corroboration
and feedback. The initial coding of the data on the patterns and modes
of student use led to the creation of five categories of graphing calculator
use throughout the three instructional units. The few discrepancies in cod-
ing between the researchers were resolved by comparing the episode in
question to similarly coded episodes and to other episodes involving the
same students. Once we completed the initial set of categories we refined
our descriptions within each of these categories by re-examining the field
notes and transcripts.
In the second phase of the data analysis, we sought to refine the teacher
profile by re-examining the evidence in the field notes and transcripts in
light of her feedback. Since the teacher substantially corroborated the pro-
file, this led to only minor changes in that profile. With our refined set
of categories of graphing calculator use, we sought to identify limitations
that the technology imposed on the teachers' pedagogical practices and on
the students' use of the tool to support their learning. This phase of the
analysis yielded a detailed description of the limitations and constraints of
the tool. We then completed our analysis by identifying the links between
the teacher's role, knowledge and beliefs and the students' patterns and
modes of calculator use.

We begin with a brief description of the role, knowledge and beliefs of

the teacher as they were reflected in her pedagogical strategies and in her
interactions with the students. We then present the patterns and modes of
graphing calculator use that emerged as the students interacted with the
teacher, with each other and with the mathematical tasks. Finally, we dis-
cuss our findings as to how the graphing calculator enabled and constrained
the students' learning and the teacher's pedagogical practice.

4.1. The reacher

The teacher was particularly skilled in using the graphing calculator, as
was demonstrated throughout the instructional units by her own use of the
calculator and by her ease in answering the students' occasional questions
on a particular calculator procedure. The teacher was familiar with the
programming features of the graphing calculator and she had written a
short program that would calculate the average slope from an ordered list
of ordered pairs of data. She felt comfortable reading programs that the
students would occasionally write. The teacher had also devised her own
method for finding a numerical approximation of derivative of any given
function. Since this was a pre-calculus course, she did not want to use
any 'black box' or built in functions of the graphing calculator to find the
derivative, but rather she wanted to make the creation of the rate of change
function explicitly visible to the students through the use of the familiar
slope formula.
The teacher's confidence in her knowledge about the calculator's cap-
abilities and its potential uses for student learning was also reflected in
the willingness with which she encouraged the students to freely use the
calculator in their work. She actively encouraged the students to take over
the use of the calculator on the overhead projection unit during class dis-
150 H.M. DOERR AND R. ZANGOR

cussions. The teacher herself showed flexibility in her use of the calculator.
While using the overhead projection unit, if a student suggested an altern-
ative viewing window or another section of a table to examine or a similar
expression for comparison, the teacher was willing to take the student's
suggestion.
The teacher believed that the calculator presented certain mathematical
limitations. She raised the issue of the validity of the calculator results by
asking questions such as 'Does the calculator always tell the truth?' and
'To what extent should we believe the calculator?' For example, in one
task, the students were investigating a decay situation where they began
with a small cup full of small, disk-shaped candies with an 'M' on one
side. The candies were spilled onto a table and those with the 'M' showing
were removed on each trial until none remained. (See Doerr (1998) for a
fuller description of this mathematical task.) The students concluded that
an exponential function models this decay process. One student observed
that even though in their experiment they ended up with zero candies, the
exponential model did not attain a zero value since 'you can divide by two
infinitely without getting zero'. But another student, who was manipulating
the calculator on the overhead screen, scrolled down the table for their
exponential function until, for very large values of x,the function appeared
to reach the value zero. This generated considerable discussion among the
students about the calculator having limitations and 'not always telling the
truth'. The students began to see the calculator as a tool that should be
checked based on their own understandings of mathematical results.
The teacher believed that the calculator would be a helpful tool for the
students to use in finding meaningful responses to problem situations. Two
specific calculator-based methods (regression analysis and curve fitting by
modifying parameters) were regularly used by three of the 3 1 students to
solve problems where part of the task was to find an equation of a function
to represent the data set for a given phenomena. These two methods did not
become very popular among the students as a whole, but rather remained
in use extensively by just three students. The teacher did not explicitly
discourage these methods by telling the students not to use them, but rather
she required a meaningful explanation of how the numerical results related
to the problem situation. This meaningful interpretation of the result could
not be given by the students who used either regression analysis or a curve
fitting approach. These students did not see their findings as estimates of
a mathematically determined model or of particular parameters directly
related to the problem situation. Those students who used the calculator's
regression functions were focused on the immediacy of obtaining some
numbers (coefficients) to use in an equation and not on making sense of
 CREATING MEANING FOR AND WITH THE GRAPHING CALCULATOR 15 1

TABLE I

Patterns and modes of graphing calculator use

Role of the Graphing Calculator Description of Student Actions

Computational Tool evaluating numerical expressions, estimating

and rounding
Transformational Tool changing the nature of the task
Data Collection and Analysis Tool gathering data. controlling phenomena, finding
patterns
Visualizing Tool finding symbolic functions, displaying data, in-
terpreting data, solving equations
Checking Tool confirming conjectures, understanding multiple
symbolic forms

the result's meaning through a more mathematical analysis of the problem

situation. The teacher's belief that the calculator could be a helpful tool to
find meaningful representations of problem situations led to a de-valuing
of the regression coefficients that were easily generated by the calculator
and of curve fitting by modification of parameters.

4.2. The stzidents

Through our analysis of the data, we identified five categories of patterns
and modes of calculator use by the students (see Table I). The focus of
our analysis is on how and why these features were used in the problem
contexts and instructional situations. The graphing calculator can in certain
situations take on more than one role simultaneously.
We will carefully describe each of these roles of the graphing calculator
and its relationship to the teacher's role, knowledge and beliefs. We will
then present the findings on the limitations and constraints of the graphing
calculator in this particular teaching and learning environment.

4.3. The calculator as a computational tool

As a computational tool, the graphing calculator was routinely used by stu-
dents to evaluate numerical expressions. Two computational issues arose:
(1) correctly entered symbols and parentheses, and (2) the precision of
the computational results. When incorrectly entering symbols, the students
encountered a variety of error conditions. For example, when one student
entered 'log(-1)', an error message appeared and the student had to ac-
knowledge the error and explain the nature of it. This student explained
the calculator's error message by referring to the domain of the logarithm
152 H.M. DOERR AND R. ZANGOR

function and to the accepted mathematical fact that the logarithm of a neg-
ative number does not exist. We interpret that this call for an explanation
for calculator results (whether errors or not) had become part of the math-
ematical norms established by the teacher, whereby the accepted truth or
falsehood of a statement had to be supported by mathematical reasoning or
justification, not by an appeal to any authority ascribed to the calculator.
A second computational issue emerged as the students attended to the
need for rounding (or approximating) the results given by the calculator.
The need for an approximation of the numerical results given by the cal-
culator was closely linked to the interpretation the students gave to the
problem situation. Their interpretation of the constraints and restrictions
imposed by the context of the problem determined whether and how they
saw the range of the possible numerical values of the result. This happened,
for example, in recursive discrete processes of exponential growth and
decay. One group of students discussed whether to round the result at each
stage or at the end of the process in a population growth problem. They
argued that the fractions at each stage could add up to an integer at the end
and that therefore this could affect the final size of the population. For price
discount problems. the students relied on their common sense knowledge
of store prices and rounded the discounted prices to the nearest dollar, but
only at specified time intervals. They knew that a discount would not be
made daily, but more likely at the end of each week. The mathematical
meanings of the numerical results were interpreted by the students so as to
make descriptive sense of the experienced phenomena.
While the students attended to the issue of rounding in the problem
situations just described, there were other situations in which the students
simply worked with whatever number of decimal places the calculator
happened to provide. For example, the students readily worked with num-
bers having many decimal places that represented measurements for dis-
tances generated by a motion detector, water temperatures measured by
a temperature probe, pressure sensed by a pressure belt and meter, and
heights calculated using the sine function. The students would record and
calculate with these numbers, even though they appeared to recognize that
'real-life' measurements could not be taken to 6 or 8 decimal places of
accuracy. They showed no tendency to round the numbers and provide
more 'realistic' answers. On the contrary, the large number of decimal
places appeared to be taken as the more realistic answer.

4.4. The calculntor as a transformatioizal tool

One of the most significant uses of the calculator in this setting was as
a transformational tool, whereby tedious computational tasks were trans-
formed into interpretative tasks. As we noted earlier in our profile of the
teacher, this teacher regularly asked students about the meaning of the
coefficients and constants in any equations they used. By her focusing the
students' attention on the interpretation of the result, rather than on the
actual computation, the students attended to making sense of the result
and validating it in the context of the problem situation. This transform-
ation of tasks from computational to interpretative was especially evident
throughout our analysis of the students' investigations of several rate of
change problems.
As a computational task, determining the rate of change for a given
function for small variations of the independent variable on the entire do-
main would be tedious to perform without a calculator. Using a paper and
pencil approach for this task would restrict the study of the rate of change
to a local property instead of the global view of the rate of change, which
in turn leads to an understanding of the rate of change as a function itself.
Many students used their graphing calculator to construct 'rate functions'
of the form p(x) = [y,(x+Ax)-y1 (x)]/(Ax), where y1 (x)is the original
function and A x is an interval for the independent variable. The students
usually set this value to .I or .01. This syntax for the rate function had been
devised by the teacher and introduced to the students through the use of
the familiar formula for slope. This form was initially used with standard
function notation and then became instantiated in the calculator notation
given above.
The students and the teacher continued to use both paper-and-pencil
and the calculator for estimating the rate of change of a function. However,
it was the calculator-based form of the numerical estimate of the rate of
change of a function that transformed tasks from a computational focus
to an interpretative one. For example, the students were given the task of
finding a function to describe the vertical position of a point on a rotating
Ferris Wheel. relative to the hub of the Ferris Wheel. The students were
also asked to find a function to describe the corresponding rate of change in
the vertical position with respect to time. Nearly all groups of students used
the 'rate function' to construct the rate of change function, j 2 , after they
had found a suitable expression for the vertical position function, J I .This
computation of the rate function generated the table of values for the rate-
function and its corresponding graph. From this, they were able to validate
in terms of their experience with the physical models of the Ferris Wheel
the periodicity of both the vertical position function and the rate of change
of that function. This linked the interpretation of the resulting equation
with the experienced phenomena and shifted the focus of the task from the
154 H.M. DOERR AND R. ZANGOR

details of each pairwise computation for rate to a global interpretation of

the function over its domain and in relationship to the problem's context.

4.5. The calculator as a data collection and analysis tool

During activities with Calculator Based Lab (CBL) devices, motion detect-
ors, pressure belts, temperature probes, and special programs, the calcu-
lator was used as a tool for data collection and analysis. The collected data
was stored, compared, and re-collected until the students decided that they
had acquired a 'satisfactory' data set. Through this kind of use of the tool,
students needed to engage in understanding the context of the activity and
in deciding, through a process of conjecturing, refining, and negotiating,
what constituted a 'satisfactory' set of data. In several of these activities,
we found that students repeated the same experiment many times, until
the set of data they had collected graphically matched an expectation they
had about the behavior of the phenomena. There were two distinct modes
of repetition that emerged. On the one hand, in working with temperature
probes and the motion of a bouncing ball, the students acted with relat-
ively little direct control over the gathering of the data. In this case, the
students would simply repeat an experiment hoping to acquire a data set
that more nearly matched their expectations. On the other hand, there were
several activities that evoked a much greater sense of direct control over
the phenomena.
In one such activity, the students were asked to use a pressure belt to
gather data that represented the pattern of 'normal' breathing and to find
a function that could be used to describe that pattern. Since many of the
students expected to obtain a sine-type plot for their data, they purposefully
adjusted their breathing in terms of both the amount of air inhaled and
the rate at which it was inhaled. They attempted to control their breathing
and hence the data collection device and the resulting data in such a way
that the data were closer to expected shape and numerical characteristics
(i.e. without jagged peaks and almost constant amplitude). In this way,
the graphing calculator and the attached pressure belt became a tool that
supported the students in controlling the data collection through cycles
of interpretation of the graphical results and purposeful changes to the
physical phenomena of breathing.

4.6. The calculator as a visualizing tool

As a visualizing tool, we found that the students used the graphing calcu-
lator in four ways: (1) to develop visual parameter matching strategies to
find equations that fit data sets, (2) to find appropriate views of the graph
and determine the nature of the underlying structure of the function, (3) to
link the visual representation to the physical phenomena, and (4) to solve
equations. Whenever the students had a set of data, they were quick to
plot the data; this inclination to visualize data could be seen throughout all
three units, but most especially in the exponential and trigonometric units.
A small number of students frequently used a visual parameter matching
strategy when they were faced with the task of finding an equation that
describes a given data set. The approach taken is illustrated by one student
in his attempts to find an equation for the harmonic motion of a spring,
gathered from a motion detector. This student, basing his strategy on the re-
cognition of the class of trigonometric functions and his knowledge about
the symbolic representations of transformations, started with an equation
+ +
of the type y = A sin(B(x C)) D. This student systematically varied
the parameters A, B, C, and D in his equation in order to obtain a visual
fit between the plot of the data set and the graph of the function.
The students became very skilled at determining appropriate viewing
windows for data sets and for symbolically defined functions. Sometimes
they used trial and error approaches to find better viewing windows, but
most of the time they set the window by using specific entries from the
table values of the relationship to be graphed. They developed the realiza-
tion that a quick judgment based on the shape of a graph was not enough
for deciding the nature of the function. The students had seen situations
in which an exponential shape can be visually mistaken for a quadratic
shape and when a close zoom in on a function made it look linear. As
a consequence of this and of the teacher's belief that the calculator itself
cannot provide the authority for a mathematical argument, the students
found reasons other than appearance to justify the underlying structure of
the function.
The graphing calculator was used as a visualizing tool when the stu-
dents examined the discrete data values describing a phenomenon and had
to relate the data back to the continuous physical process and the func-
tion which potentially describes that process. For example, in the Ferris
Wheel investigation, a group of students constructed a table for the rate
of change using their 'rate function' and observed that what appeared
to be the highest value of the function repeated consecutively. Knowing,
from their experience of problem situation, that the motion of the wheel
was continuous, they decided that 'there was something in between'. The
students then shifted from the table to the algebraic expression for the con-
tinuous graph and traced it, thereby confirming their conjecture and finding
the single highest value. The graphing calculator provided a visual repres-
entation of the continuous function that described the rate of change in
the vertical motion of the Ferris Wheel. This became the link between the
156 H.M. DOERR AND R. ZANGOR

discrete data table, which was missing a critical data point (the maximum
value), and the physical phenomena, which, from their experience with the
wheel, the students conjectured must reach some unique maximum value.
The graphing calculator's use as a visualizing tool was also reflected
in how the students solved equations or inequalities. In the class discus-
sions, the teacher had discussed all of the available methods for solving
an equation and encouraged the students to choose among these methods.
The methods included paper and pencil solutions, use of the calculator's
'Solve' command, and graphical solutions. We found that many of the
students used a graphical approach, which involved less computation and
appeared to support a more meaningful interpretation of solutions. For ex-
ample, in one problem, the students were asked to determine when a bank
account attained a given amount of money. Many students approached this
by examining the intersection of the graph of the function describing the
amount of money at any point in time with the graph of the horizontal line
representing the given amount. This visual approach seemed to support
their description of the amount of money in the bank reaching a given
value.

4.7. The calculator as a checking tool

In this study, we considered the use of the calculator as a checlung tool
when it was used to check conjectures made by students as they engaged
with the problem investigations. This was followed by rejecting and re-
conjecturing, trying to prove the conjecture, or simply accepting the con-
jecture. In many tasks, the students posed a conjecture about a possible
function as fitting a data set and then they used their graphing calculator
to check how well it fit. The students' strategies depended in part on how
they had found an equation in the first place. For example, in the case of the
student using visual parameter matching for find an equation for the data
from harmonic motion (described above), the visual mis-match between
his conjectured equation and the data was crucial for his decision to reject
his equation and try a new one. In this way, the graphing calculator was an
essential check of the match between his equation and the data. In other
cases, we found that the calculator was used as a checking tool only in the
most trivial sense. For example, when one student found an equation using
the regression features of the calculator, he checked the match between
the regression equation and the data. Since the students generally chose
an appropriate regression model, the graphical mismatches were generally
minor. Those few students who preferred regression equations were in-
clined to 'go with what the calculator says' and rarely questioned the fit of
the equation to the data.
 When the students had determined the function through a meaning-
ful mathematical process, they would check to see how well their graph
matched the data. Occasionally, a graphical mismatch revealed a mistake
in entering the form of the equation or in a computation. For example, in
finding an equation for the harmonic motion described earlier, one student
determined the amplitude, both the horizontal and vertical shift, and the
frequency by tracing the coordinates of consecutive maximum points in
the data set. This approach enabled her to find a meaningful equation based
on her knowledge of trigonometric functions and their transformations. In
this case, the check between her conjectured equation and the data set was
confirmed as she observed that the graph and the data were visually aligned
along critical features such as local maxima and minima.
The calculator's role as a checking tool was evident in the context of
activities on transformations on functions, which were designed both in go-
ing from the equation to the graphical representation and from the graph to
finding an equation (non-unique) to match the given graph. The calculator
was initially used as a visualizing or graphing tool, but as students grasped
the ideas of transformations, the calculator was used almost exclusively as
a checking tool. The graphing calculator helped the students understand
the idea of the non-uniqueness of algebraic representations for the same
graph in the case of exponential and trigonometric functions. For example,
the students were asked to compare the graph of y = 8 * 2" with the
graph of J = 2'+3 and found that the graphs appeared identical. Using the
overhead display unit, the teacher switched to the table of values, revealing
that the numerical values were also identical. This then led to an algebraic
argument that supported the claim that these two functions were identical.
In this way, the graphing calculator was used to confirm conjectures, but at
the same time, the teacher led the mathematical discussion to the need for
justification by algebraic reasoning.
The role of the graphing calculator as a checlung tool by both the
students and the teacher was especially interesting in the case of the trigo-
nometric functions. The periodicity of the trigonometric functions results
in a particularly salient characteristic, namely the 'look alike' feature of the
graphs of these functions. One can argue that the 'look alike' appearance
on the screen of the calculator is common to almost any class of functions,
certainly to the exponential class that was studied by these students. How-
ever, for periodic functions, it is the case that simple horizontal shifts of
the viewing window can lead to no visible change in the appearance of
the graph and this characteristic is not true for other functions. Difficulties
in interpreting a periodic graph were increased by the limitations of the
graphing calculator's screen, where the units on the coordinate axes are
158 H.M. DOERR AND R. ZANGOR

not labeled. The investigation of trigonometric functions with the calcu-

lator became limited to the use of the tables and occasionally the 'Trace'
command. This pattern of use as a checking tool for the trigonometric
function was limited by a mismatch between the decimal representations
of the calculator and many of the tasks where the values of the independent
variable were often given as rational multiples of n.The calculator's table
and the 'Trace' command showed these values in decimal form. Hence,
there was no easy direct comparison of the closeness of numerical val-
ues without an extra tedious conversion step of the value of the rational
multiple of rr to a decimal value.

4.8. Constraints and limitations vvith the graphing calculator

The graphing calculator emerged as a constraint and limitation in two
ways: (1) students' attempted uses of the device as a 'black box' without
attending to meaningful interpretations of the problem situation; and (2)
the personal (or private) use of the tool. The use of the calculator as a
'black box' represents the class of situations in which the students did not
have a meaningful strategy for the use of the calculator. In one particular
task, the students were asked to design simulation of the spread of a ru-
mor. The students were given the calculator syntax for generating single
random numbers and lists of random numbers. Some groups of students
decided that they had to use the random number generator and tried to
do so before they actually read the content of the activity and understood
what the random number syntax might be useful for. These students did
not understand that the calculator's random number generator was simply
an alternative tool, not the only tool, and certainly not 'the solution' to the
problem. At least one group of students became considerably sidetracked
on the quantitative details of how the random number worked. They spent
considerable time discussing if they needed to correct for whether or not a
zero could be generated by the random number generator. This discussion
occurred well before they had found a use for random numbers in the
design of their simulation.
Another limitation of the graphing calculator was in its use as a private
device. Unlike the case of group work with a computer, where the screen
is a visible and hence potentially shared artifact, the calculator's screen
was not visible in a shared way and was not necessarily common among a
group of students. The tendency of the students to use their calculators as
private devices regularly led to the breakdown of group interactions. We
found a pattern that began when two or more students in a group tested
or checked a possible conjecture or computation on their own calculators
and then continued to use the tool to explore possible solutions, interpret-
ations or refinements of their own thinking. The results of the graph, table,
or computation were used to further their own individual thinking, rather
than shared back with the group. This closed the communication network
among the students and made the re-opening of discussion quite difficult.
As time progressed, the students had different interpretations, represent-
ations and problems rather than a common or shared view of the task at
hand. Once this closing of communication had happened within a group,
the students tended to continue to pursue their investigations as individuals,
with little talking and no sharing of ideas, representations, or results. If the
students had questions, they would ask them of the teacher, rather than
each other.
This use of the calculator as a private, personal tool was in sharp con-
trast to its use during the class discussion, as a shared device (via the
overhead screen). This shared, public display was supportive of the com-
munication among students and was used to facilitate the comparison and
unification of ideas. Since the teacher let the students control the calculator
in most instances, this shared device helped encourage student initiative
and often resulted in the students' leading of the discussion. The decision
to use the calculator as a shared device in the classroom came from the
development of the discussion and was often student-requested.

The role, knowledge and beliefs of the teacher and the actions of the stu-
dents as they interacted with each other and with the graphing calculator
created meaning for the tool and meaning with the tool. The teacher's
confidence in her own knowledge and skills and her own flexible use of
the calculator led to a classroom environment where students were free
to use their calculators as they wanted and were actively encouraged to
use them to calculate, explore, confirm, or check mathematical ideas. The
teacher's knowledge of the limitations of the calculator led her to encour-
age the students to question their calculator-based results. In this way, the
calculator became a tool that itself needed to be checked on the basis
of mathematical reasoning. Contrary to the concerns raised by Williams
(1993) and Wilson and Krapfl (1994), the calculator did not become a
source of mathematical authority in this classroom. We see this as a con-
sequence of the teacher's knowledge of the limitations of the calculator
and her belief that conjectures are proven on the basis of mathematical
reasoning or argument.
The teacher's belief that the graphing calculator would be a helpful tool
for the students to use in finding meaningful responses to mathematical
160 H.M. DOERR AND R. ZANGOR

tasks is evidenced by the kinds of questions she asked the students. She
regularly asked the students about the meaning of the coefficients that they
found and how those coefficients related back to the experienced phenom-
ena. We found that in finding equations to fit graphs or numerical data sets,
the most common strategy used by the students was a meaningful mathem-
atical approach. The visual-parameter matching strategy, while based on
the mathematical recognition of the general form of the function, was used
extensively, but by only a few students. These findings confirm the results
found by Ruthven (1990). Furthermore, for the trigonometric functions,
we found that the use of the graphing calculator as a checking tool to con-
firm their conjectures was not used by most students, but rather they relied
solely on their algebraic and geometric knowledge of the transformations.
The teacher's belief in meaning making led to the de-valuing of regres-
sion equations as solutions and limited the 'black box' use of the graphing
calculator. This was particularly visible in the teacher's strategic choice to
use a rate function that was built upon the students understanding of the
familiar slope formula rather than a use of built-in derivative functions.
The rate function served to illuminate a powerful role for the graphing
calculator as transforming a local, computational task (the computation of
average rate of change) to a global, interpretative task (the relationship
between the numerically determined rate of change function and the given
function over the entire domain).
The role of the graphing calculator as a device that can both monitor
and control data collection appears particularly powerful and has been
little explored in the research literature to date. In this study, we found that
students engaged in two distinct actions: one where they repeated the data
collection to match their expectations and the other where they controlled
their actions to match their expectations. As a visualizing tool, the students
used the graphing calculator to find equations that matched data sets, to
find appropriate views of the graph, to link the visual representation of the
graph to physical phenomena in explanatory ways, and to solve equations.
The most significant limitation of the calculator was as its use as a
private device. While we did observe, as did Farrell (1996), that students
frequently used their calculators while the teacher or other students were
talking in lecture or whole class discussion, we also observed that this per-
sonal use of the technology served to breakdown group communications.
Once the students began to work individually on a task, it was very difficult
for them to resume functioning as a group, since their thinking about the
problem had often progressed in different directions. In contrast, the use
of the overhead projection unit appeared powerful in generating shared
representations, alternative interpretations, and contrasting conjectures.
The graphing calculator has been the focus of numerous research studies
that have sought to answer the question of whether or not the use of the
graphing calculator is effective in achieving learning goals for students.
However, such studies do little to illuminate the patterns and modes of
graphing calculator use by students or the roles, knowledge and beliefs
of teachers or the influence of the mathematical tasks on the students'
activities. In this study, we have analyzed and described how students and
their teacher interpret and make use of the graphing calculator as a tool
that is part of their mathematical practice.
We found that five patterns and modes of graphing calculator tool use
emerged in this practice: computational tool, transformational tool, data
collection and analysis tool, visualizing tool, and checking tool. This sug-
gests that the graphing calculator is a rich, multi-dimensioned tool and that
the continued study of its use in classroom practice will need to carefully
delineate the patterns and modes of use that occur in any given context as
conclusions are drawn related to student learning. The results of this study
suggest that the role, knowledge and beliefs of the teacher influenced the
emergence of such rich usage of the graphing calculator. The teacher's role
in encouraging interpretation and explanation led to a valuing of mean-
ingful mathematical constructions for equations, to the transformation of
rate from a computational task to an interpretive task, to the valuing of
algebraic arguments to support graphically and numerically generated con-
jectures, and to a de-valuing of regression equations or appeals to the
calculator as an authority in a mathematical argument. We found that the
calculator as a private device inhibited mathematical communication in a
small group setting, while the shared screen of the graphing calculator ap-
peared to be a powerful tool for supporting the comparison and unification
of mathematical ideas.

Adams, T.: 1997, 'Addressing students' difficulties with the concept of function: Applying
graphing calculators and a model of conceptual change', Focus on Learning Problems in
Mathematics 19(2), 43-57.
Cobb, P. and Yackel, E.: 1996 'Constructivist, emergent, and sociocultural perspectives in
the context of developmental research', Educational Psychologist 31(3/4), 175-190.
Confrey, J. and Doerr, H.M.: 1996, 'Changing the cuniculum to improve student un-
derstandings of function', in D.F. Treagust, R. Duit and B.J. Fraser (eds.), Improving
Teaching and Learning in Science and Mathematics, Teachers College Press, New York,
pp. 162-171.
162 H.M. DOERR AND R. ZANGOR

Doerr, H.M.: 1998, 'Student thinking about models of growth and decay', in A. Olivier
and K. Newstead (eds.). Proceedings of the 22nd Co~lferenceof the Internatio~~nl
Group for the Psychology ofMathemcrtics Education, Vol. 2, University of Stellenbosch.
Stellenbosch, Africa, pp. 256-263.
Doerr, H.M. and Tripp, J.: 1999, 'Understanding how students develop mathematical
models'. Mathematical Thinking and Learning 1(3), 23 1-254.
Drijvers, P. and Doorman. M.: 1996, 'The graphics calculator in mathematics education',
Jounicrl of Marhernatical Behavior 1 3 4 ) . 425440.
Dunham. P. and Dick, T.: 1994. 'Research on graphing calculators'. The Mathenintics
Teacher 87(6). 4 4 0 4 4 5 .
Farrell. A.M.: 1996, 'Roles and behaviors in technology-integrated precalculus
classrooms'. Jo~rrnalofMathe171aticalBehatior 15(1). 35-53.
Hiebert. J.. Carpenter, T.P.. Fennema, E., Fuson. K., Weame. D.. Murray, H.. Olivier,
A. and Human. P.: 1997, Making Sense: Teaching and Learning Mathematics ~ . i t h
Understanding, Heinen~ann,Portsmouth. NH.
Hollar. J.C. and Norwood, K.: 1999, 'The effects of a graphing-approach intermediate
algebra curriculum on students' understanding of function', Journal ,for Research iri
Mathernotics Education 30(2). 220-226.
Lave, J. and Wenger. E.: 1991. Situated learning: Legitilnrrte peripheral participntion,
Cambridge University Press, New York.
Leinhardt. G., Zaslavsky, 0 . and Stein. M.K.: 1990. 'Functions, graphs. and graphing:
Tasks. learning and teaching'. Rel,ie\r c!f'Educatiorlul Resecirch 60(1), 1-64.
Meira. L.: 1995, 'Mediation by tools in the mathematics classroom', in L. Meira and
D. Carraher (eds.). Proceeditig.~cf the 19th Annual Corference of the International
Groups ,for the Psxchology of Mathematics Education, Vol. 1. ERIC Clearinghouse.
Columbus. OH. pp. 102-1 11.
Meira. L.: 1998. 'Making sense of instructional devices: The emergence of transparency in
mathematical activity '. Journal for Research in Mathematics Education 29(2), 12 1-1 42.
Moschkovich. J., Schoenfeld, A. and Arcavi. A,: 1993, 'Aspects of understanding: On mul-
tiple perspectives and representations of linear relations and connections among them'.
in T.A. Romberg. T.P. Carpenter and E. Fennema (eds.). Integrc~tingResenrc,h on the
Graphical Representations of Functions. Lawrence Erlbau~nAssociates. Hillsdale. NJ.
pp. 69-100.
National Council of Teachers of Mathematics.: 1989. C ~ ~ r r i c u l ~and
l m E~~aluatiorl
Stcind-

anls,for School Mathematics, Author. Reston, VA.

Penglase, M.and Arnold, S.: 1996, 'The graphics calculator in mathematics education:
A critical review of recent research', Mathematics Education Research Journal 8(1).
58-90.
Porzio. D.T.: 1997. March. Examining Effects of Graphing Calculator Use on Students'
Understanding o f Numerical, Graphical and Synbolic Represeritations of Calculus Con-
cepts, Paper presented at the Annual Meeting of the American Educational Research
Association, Chicago, IL. (ERIC Document Reproduction Service No. ED 410 1 12.)
Quesada, A.R. and Maxwell, M.E.: 1994, 'The effects of using graphing calculat-
ors to enhance college students' performance in precalculus', Educational Studies in
Mathematics 27(2), 205-215.
Roth, W.-M. and McGinn, M.K.: 1997, 'Graphing: Cognitive ability or practice?', Science
Education 81(1), 91-106.
Ruthven, K.: 1990. 'The influence of graphing calculator use on translation from graphic
to symbolic forms', Educational Studies in Mathematics 21(5), 431-450.
Shoaf-Grubbs, M.M.: 1994, 'The effect of the graphing calculator on female students'
spatial visualization skills and level of understanding in elementary graphing and algebra
concepts', in E. Dubinsky. A. Schoenfeld and J. Kaput (eds.), Research iri Collegiate
Mathematics Education. Vol. 1 , American Mathematical Society in cooperation with the
Mathematical Association of America, Providence, RI, pp. 169-194.
Simmt. E.: 1997. 'Graphing calculators in high school mathematics', Jozcr-rial of Computers
in Mathernatics anti Scierzce Teachirig 16(2/3). 269-289.
Tharp, M.L.. Fitzsimmons, J.A. and Ayers, R.L.: 1997, 'Negotiating a technological
shift: Teacher perception of the implementation of graphing calculators'. Jo~rrnal rv
Computers in Matheniarics and Science Teaching 16(4), 55 1-75.
Williams. C.G.: 1993, 'Looking over their shoulders: Some difficulties students have with
graphing calculators'. Marlzernatics and Complrter Education 27(3), 198-202.
Wilson. M.R. and Krapfl, C.M.: 1994. 'The impact of graphics calculators on students'
understanding of function'. Journal of Coniputers in Mathernatics and Science Teaching
13(3). 252-264,
Yerushalmy, M.: 1991. 'Student perceptions of aspects of algebraic function using multiple

HELEN M. DOERR and ROXANA ZANGOR

Department of Mathematics
Syracuse Universih
Syracuse, New York
U.S.A.

E-mail: hmcloerr@sj~edu

http://www.jstor.org

LINKED CITATIONS
- Page 1 of 1 -

You have printed the following article:

Creating Meaning for and with the Graphing Calculator
Helen M. Doerr; Roxana Zangor
Educational Studies in Mathematics, Vol. 41, No. 2. (Feb., 2000), pp. 143-163.
Stable URL:
http://links.jstor.org/sici?sici=0013-1954%28200002%2941%3A2%3C143%3ACMFAWT%3E2.0.CO%3B2-3

This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library's website or contact a librarian to learn about options for remote access to JSTOR.

References

The Effects of a Graphing-Approach Intermediate Algebra Curriculum on Students'

Understanding of Function
Jeannie C. Hollar; Karen Norwood
Journal for Research in Mathematics Education, Vol. 30, No. 2. (Mar., 1999), pp. 220-226.
Stable URL:
http://links.jstor.org/sici?sici=0021-8251%28199903%2930%3A2%3C220%3ATEOAGI%3E2.0.CO%3B2-S

Functions, Graphs, and Graphing: Tasks, Learning, and Teaching

Gaea Leinhardt; Orit Zaslavsky; Mary Kay Stein
Review of Educational Research, Vol. 60, No. 1. (Spring, 1990), pp. 1-64.
Stable URL:
http://links.jstor.org/sici?sici=0034-6543%28199021%2960%3A1%3C1%3AFGAGTL%3E2.0.CO%3B2-Y

Making Sense of Instructional Devices: The Emergence of Transparency in Mathematical

Activity
Luciano Meira
Journal for Research in Mathematics Education, Vol. 29, No. 2. (Mar., 1998), pp. 121-142.
Stable URL:
http://links.jstor.org/sici?sici=0021-8251%28199803%2929%3A2%3C121%3AMSOIDT%3E2.0.CO%3B2-A