Journal of Mathematical Behavior 30 (2011) 166179

Contents lists available at ScienceDirect

The Journal of Mathematical Behavior

journal homepage: www.elsevier.com/locate/jmathb

Affect and graphing calculator use

Allison W. McCulloch
North Carolina State University, Department of Science, Technology, Engineering and Mathematics Education, 502L Poe Hall, 2310 Stinson Drive, Raleigh, NC
27695-7801, United States

a r t i c l e i n f o a b s t r a c t

This article reports on a qualitative study of six high school calculus students designed to
build an understanding about the affect associated with graphing calculator use in inde-
pendent situations. DeBellis and Goldins (2006) framework for affect as a representational
Keywords: system was used as a lens through which to understand the ways in which graphing cal-
Affect culator use impacted students affective pathways. It was found that using the graphing
Graphing calculators calculator helped students maintain productive affective pathways for problem solving as
Instrumentalization
long as they were using graphing calculator capabilities for which they had gone through a
Problem solving
process of instrumental genesis (Artigue, 2002) with respect to the mathematical task they
Calculus
were working on. Furthermore, graphing calculator use and the affect that is associated
with its use may be influenced by the perceived values of others, including parents and
teachers (past, present and future).
2011 Elsevier Inc. All rights reserved.

1. Introduction

Imagine watching a high school calculus class getting ready to take an exam. Before even looking at the exam paper that
has been handed to her you see one young lady pick up her graphing calculator to make sure it works. She sighs heavily,
as if she is relieved, and then sets it down right next to her exam paper. Later as she is working on a problem you see her
using her calculator and whisper a quiet Yes! when she sees something on her calculator screen that apparently makes her
happy. Another young man towards the back of the room looks perplexed as he works on a problem, and then he picks up
his calculator to do something, looks back and forth between his paper and his calculator and then actually reaches over his
shoulder and gives himself a pat on the back. It is clear that for these students mathematics is not an emotionless activity.
Furthermore, it is apparent that these students do not feel alone in their activity; each of them has a graphing calculator
with which they are interacting and that they are responding to with outwardly articulated feelings. What was it about this
tool that had these students expressing so much emotion? What are they actually doing with it and why?
More than twenty years ago Kaput (1989) predicted that in the near future technology tools in mathematics education
would have great impact on students affective experiences. He suggested that technology tools would be instrumental in
helping students avoid the discrepancy between an individuals expectations when beginning a problem solving activity
and the ongoing activity required to complete the activity that often leads to negative emotions about the mathematics. This
paper reports on a study that examines such a relationship between technology and problem solving, a qualitative study of
six high school calculus students and the affect they associate with their graphing calculator use. It aims to answer questions
like those motivated by the starting vignette. Specifically, in what ways does affect impact ones graphing calculator use? And,
in what ways does graphing calculator use impact ones affect? I begin with a review of the relevant literature and theories

Tel.: +1 919 513 2803; fax: +1 919 513 0505.

E-mail address: allison mcculloch@ncsu.edu

0732-3123/$ see front matter 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jmathb.2011.02.002
 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 167

that provide the background for this study. Next, I describe the research methods, followed by a report of the findings. Finally,
I conclude with a discussion of the implications the findings present for teacher education and future research.

2. Background

Until recently most of the research on graphing calculator use in mathematics education has either been quantitative
in nature focusing on student achievement and attitude or qualitative studies on the teaching and learning of a particular
mathematical topic (Burrill et al., 2003; Ellington, 2003). However, there is a growing body of research on how students are
adapting graphing calculator technology to their problem solving strategies (e.g. Ball & Stacey, 2004; Berry & Graham, 2005;
Graham, Headlam, Honey, Sharp, & Smith, 2003; Harskamp, Suhre, & van Streun, 2000; Sheryn, 2006). Over the past decade
researchers have been studying how students come to fully integrate graphing calculators as a tool for learning. Artigue
(2002) and her colleagues have proposed the term instrumental genesis to describe this process of coming to understand the
potentialities and constraints of an artifact such as the graphing calculator while at the same time developing mathematical
knowledge. Most of the studies focused on understanding instrumental genesis have looked specifically at graphing calcula-
tors with Computer Algebra System (CAS) capabilities. They have found that using graphing calculators with CAS capabilities
effectively is not easily learned and that students go through many phases before becoming proficient users (Artigue, 2002;
Drijvers, 2000; Guin & Trouche, 1999). As we learn more about the effects that the adoption of graphing calculator technology
seems to be having on the assessment and learning of particular topics it becomes apparent that we need to know more
about what students are actually doing with their calculators when they work independently. Furthermore, it is apparent
we need to learn more about the role of emotions, values and beliefs as they relate to graphing calculator use in problem
solving.
Few studies have looked specifically at how and why students use graphing calculators in particular ways (Burrill et al.,
2003); two exceptions are studies by Doerr and Zangor (2000) and Goos, Galbraith, Renshaw, and Geiger (2003). These
studies were both classroom based studies that aimed to understand the different roles that the graphing calculator takes on
within a classroom community. Doerr and Zangor (2000) conducted an observational case study of two precalculus classes
and their teacher. Within this case study they considered how the classroom as a community shaped the ways in which
technology was used. It was determined that within the context of the class the graphing calculator was used by the students
in five different modes: as a computational tool, transformational tool, data collection and analysis tool, visualizing tool, and
checking tool. These five modes of tool use emerged from the interactions between the teacher and the students. While this
study highlighted the ways that a teachers beliefs about the role of the graphing calculator can shape the ways that it is used
in the context of the classroom, the methodology employed did not allow for examination of how the classroom modes of
use translated to independent situations.
Goos et al. (2003) conducted a longitudinal study of 5 secondary classrooms and how graphing calculators and their
peripheral devices were used as a tool that was integral to the learning environment in the context of these classrooms.
Unlike Doerr and Zangor who focused on the actual actions taken with the tool, they focused on the students and teachers
relationship with the tool. They theorize that when technology like the graphing calculator is used in relation to teaching
and learning interactions there are four roles it may take on: master, servant, partner, and extension of self. Each of the
four categories is a metaphor for the interaction (or relationship) between the individual and the tool. These metaphors
emphasize the social nature of technology use in a classroom setting. Each of these studies has added to the knowledge of
the complex role that the graphing calculator plays in the mathematics that is produced and shared in the context of the
classroom. However, there is scant research on students graphing calculator use that has attended to the role of affect (i.e.
emotions, attitudes, values, beliefs) in students graphing calculator use.
One aspect of affect that has been studied as it relates to graphing calculator use is attitude. Such studies most often
focus on how graphing calculator use impacts ones attitude toward mathematics. In her 2003 meta-analysis of calculator
studies, Ellington identified 18 studies conducted between 1983 and 2002 that fell in this category. The studies all compared
two groups of students taught by equivalent methods, one having access to calculator technology and one not. Each of the
studies used Likert scale instruments for measuring attitude towards mathematics. Ellington grouped studies according to six
attitudinal factors identified in the Minnesota Research and Evaluation Project (Sandman, 1980) including attitude toward
mathematics, anxiety toward mathematics, self-concept in mathematics, motivation to increase mathematical knowledge,
perception of mathematics teachers and value of mathematics in society with only the first (attitude toward mathematics)
being measured by most studies. Ellingtons meta-analysis revealed that students who used calculators during instruction
reported a better attitude toward mathematics than those who did not (weighted effect size, g = .32). This is an important
finding, but it is just the tip of the iceberg. There are many affect related questions that remain unanswered. For example, why
do students who have access to calculators have better attitudes towards mathematics? Some researchers have suggested
that this phenomenon might be due to a feeling of confidence about engaging with mathematics when a tool is available
(Dunham, 2000). However, if it is simply confidence, what is it about having the graphing calculator available that makes
students more confident in mathematical situations?
In the Educational Studies in Mathematics special edition on affect in mathematics education, Opt Eynde and Hannula
(2006) introduced the case of Frank which served as the unifying feature for all of the articles in the issue. Frank was a junior
high school student who was participating in a project about the role of students beliefs and emotions during problem
solving in the mathematics classroom. Frank was asked to solve a problem about a group of Kosovan refugees trying to go
168 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179

to a hospital in Albania through the mountains. Part of the problem asked the time travel by car from the village to the
city where the hospital was located. Frank was video taped while solving this problem and shortly thereafter participated
in a video-stimulated recall interview in which he watched himself working on the problem and was asked about how he
felt and what he did during the problem solving session. Eynde and Hannula report that at one point in the session Frank
took out his calculator. When he was asked about this Frank replied, Actually, I did not really need the calculator there. I
wasnt thinking properly, and then I panic, and then I immediately want to go to my calculator, and then if I stop and think
for a moment, I probably know again what I have to do (p. 126). When he is further probed about his feelings regarding
this action Frank says, . . .I dont know how to put it, you dont feel well because you need to go to the calculator (p. 127).
Though this is a short episode, Frank provides further evidence that the decision to use technology tools like the graphing
calculator is laden with emotion and deserves to be studied.

3. Theoretical framework

This exploration of students graphing calculator use was guided by two theoretical constructs, affect as a representational
system (DeBellis & Goldin, 2006) and instrumental genesis (Artigue, 2002). Each is described in the sections that follow.

3.1. Affect as a representational system

The study of affect in mathematics education has its roots in the investigation of emotions and problem solving. Most
early work in this field focused on mathematics anxiety and attitude toward mathematics; there was little consensus of
what made up the affect domain or how it was best studied (McLeod, 1992; Zan, Brown, Evans, & Hannula, 2006). More
recently there has been a concerted effort to both define affect and develop frameworks to help coordinate discussion and
research within the mathematics education community (Zan, Brown, Evans, & Hannula, 2006). Mandlers (1989) cognitive-
constructivist model of the process of emotional experience has been the ground on which theoretical frameworks for the
study of affect in mathematics education have been built. Based on Mandlers theory McLeod (1992) offered up a framework
in which the affective domain was defined as the internal system that includes emotions, attitudes and beliefs. McLeod
suggested that these three constructs fall in line ranging in stability and intensity with emotions being the most intense
and the least stable, beliefs being the least intense and most stable, and with attitudes falling in between. Later DeBellis and
Goldin (1997) added a fourth dimension to McLeods framework for the affective system by including values as a part of the
affective domain. They represented these four dimensions using a tetrahedral model rather than by a continuum of stability
and intensity to show the interaction between them. They further operationalized this addition with the introduction of
their framework for affect as a representational system (DeBellis & Goldin, 2006).
The methodology in this study is grounded in DeBellis and Goldins research-based theoretical framework on affect as
a representational system. By stating that this system is representational, they mean that it exchanges information with
cognitive systems. DeBellis and Goldin describe two categories of affect, local affect and global affect. Local affect refers to
changing states of feeling during problem solving. Global affect refers to more stable, longer-term affective constructs such
as attitudes, values, and beliefs.
Implicit in the framework is the conjecture that over time local affective experiences that are similar and powerful
can come to influence the more stable constructs of global affect. For example, if a student has repeated experiences of
frustration when trying to create and use a table on a graphing calculator to solve a problem, that student may start to have a
negative attitude toward the table tool and even possibly develop the belief that the table is not a useful tool. In addition, an
individuals affect (both local and global) is also influenced by the affect of others, social and cultural conditions and external
contextual factors. This is referred to as external affect. For example, a student who repeatedly hears from a teacher that the
graphing calculator should only be used as a last resort when problem solving may possibly feel guilty when the decision is
made to use one.
Goldin (2000a) has pointed out that students use emotions to provide useful information, to facilitate monitoring and
to evoke heuristic processes. He suggests that affect is not inessential, but critical to the structure of competencies that
account for success or failure in problem solving. To better understand the role of local affect in problem solving Goldin
(2000a) introduced the notion of affective pathways. Affective pathways are sequences of states of feeling that interact with
cognitive representations in problem solving. An example of an affective pathway follows:
In an (idealized) model, the initial feelings are of curiosity. If the problem has significant depth for the solver, a sense
of puzzlement will follow, as it proves impossible to satisfy the curiosity quickly. Puzzlement does not in itself have
unpleasant overtones but bewilderment, the next state in the sequence, may. The latter can include disorientation,
a sense of having lost the thread of the argument of being at sea in the problem. . .If independent problem solving
continues, a lack of perceived progress may result in frustration, where the negative affect becomes more powerful
and more intrusive. This is associated with the occurrence of an impasse. However, there is still the possibility that
a new approach will move the solver back to the sequence of predominately positive affect. Encouragement can be
followed by pleasure as the problem begins to yield, by elation as major insights occur, and by satisfaction with the
sense of a problem well solved and with learning that has occurred. (p. 211)
 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 169

This idealized model illustrates how local affect might influence the heuristics employed by a problem solver. In the
context of this paper the focus is on how the availability of a tool like the graphing calculator might further influence an
affective pathway like the one described above. For example, if a student is facing feelings of bewilderment or disorientation
it is possible that the introduction of a useful tool might invoke feelings that are of a more positive sequence.
Meta-affect, or affect and/or cognition about affect, is the monitoring system that allows for people to recognize affective
representations with cognitive representations and produce other affective representations (Goldin, 2002). For example, a
student who has hit a road block in a problem is frustrated. However, if that student reflects on the feeling of frustration and
remembers that frustration is temporary when one tries another approach and is successful, then the student might become
curious about the problem and move forward. DeBellis and Goldin believe that meta-affect is one of the most important
aspects of affect. It is through this monitoring system that problem solvers learn to create productive affective pathways
for themselves. At this point, graphing calculator studies have addressed the more stable affective construct of attitude, but
few, if any, have addressed graphing calculator use as it relates to local affect. It is through the study of local affect that we
might move beyond classifying students attitudes and come to understand the influence that graphing calculator use has
on problem solving experiences.
The DeBellis and Goldin view of affect and problem solving is consistent with many frameworks that have been offered by
the research community in that it seriously considers the interactions between affective and cognitive systems (e.g. Evans,
Morgan, & Tsatsoroni, 2006; Malmivuori, 2006; Opt Eynde & Hannula, 2006). While the view is consistent, it studies affect
from a different perspective in that it was designed in the context of individual problem solving, not classroom interactions.
As a result, though social and cultural conditions are included, it does not regard the interactions between these conditions
and students emotions in the way that others do, by focusing on interactions in the classroom (e.g. Evans et al., 2006; Opt
Eynde, De Corte, & Verschaffel, 2006). Nevertheless, DeBellis and Goldins separation of local and global affect along with
the importance placed on meta-affect and recognition of external affect make it an appropriate and powerful framework for
this particular study.

3.2. The process of instrumental genesis

The premise that students understandings are shaped by the tools that they use and by their relationship with those
tools is consistent with socio-cultural theories of learning (e.g. Vygotsky, 1978). When studying how CAS tools are used
in mathematics, specifically how they mediate student learning, Artigue (2002) and her colleagues found it useful to turn
to Chevallards work in anthropology (e.g. Chevallard, 1992) to better understand the ways in which students develop a
relationship with a tool that takes into account the context of classroom learning. When considering the graphing calculator
it is necessary to point out that I consider the graphing calculator to be an artifact that is actually made up of many tools.
For example, there are both tools for visualization and tools for computation. However, I believe that these tools are useless
until they become instruments through a process that Artigue calls instrumental genesis.
In her explanation of this theory and its relationship to graphing calculator use, Artigue explains that an instrument
is a mixed entity, part artifact, part cognitive schemes which make it an instrument. For a given individual, the artifact
at the outset does not have an instrumental value (p. 250). The process of instrumental genesis involves one coming to
understand the potentialities and constraints of the tool while at the same time developing mathematical knowledge. In
other words, users shape the artifacts they use and the artifacts shape the users, and that yields instruments (Artigue &
Kilpatrick, 2008, p. 6). So, while a student may own a graphing calculator, that alone does not make it an instrument. It
is possible that particular modes (or tools) on the graphing calculator become instruments to a student before others. For
example, in the context of linear functions one student might have instrumented visualization tools of a graphing calculator
and meaningfully use graphical representations to understand and engage in solving a linear function problem. That same
student may not have developed the CAS capabilities as instruments with respect to linear functions and thus will not find
them helpful in this context. It is important to note that the mathematical context matters, as in the process of instrumental
genesis the tool and the mathematical knowledge are shaping each other. The same student just described above might have
developed the CAS capabilities in a different mathematical context, say function limits.
Kilpatrick (2009) noted that the theory of instrumental genesis appears to have considerable promise in research regard-
ing the ways in which technology is and is not being used. In the context of this study, whether or not students have developed
particular capabilities of the graphing calculator as instruments could have considerable impact an affective pathway when
attempts are made to incorporate its use.

4. Methodology

To build an understanding of how the availability of a graphing calculator impacts students local affect, or affective
pathway, during independent problem solving qualitative methods were used. The choice of design methods was influenced
by a series of pilot studies. Three lessons were learned from those experiences that influenced the design of this study. First,
students with different math backgrounds appear to have different perceptions of the usefulness of the graphing calculator.
Therefore the settings for this study were chosen carefully to increase the possibility of identifying unique individual graphing
calculator stories. Second, students have many different purposes for actually using the graphing calculator. Third, task
selection is very important to the development of meaningful conversations about graphing calculator use. The tasks selected
170 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179

Table 1
Study participants.

Gender Teacher rated ability School G.C. type

Aaron M Average C TI-89

Enoch M Strong A TI-83+
Maryanne F Strong B TI-83+
Melissa F Weak C TI-89
Rudy M Weak A TI-83+
Shemika F Strong A TI-83+

for this study were chosen based on the responses students provided in pilot studies. These tasks along with the details of
the methods are described in the following sections.

4.1. Setting and sample

Advanced Placement (AP) Calculus classes were chosen as the focus of this study because the curriculum and expectation
of calculator use is relatively consistent nationwide since it is set by The College Board. To ensure that the population of
students was as diverse as possible, students from four high schools in the northeastern United States were surveyed and
six students were chosen based on the results of the survey. These high schools were purposefully selected based on access,
presence of an AP Calculus program that uses graphing calculator technology, and their ethnic and socio-economic status
(SES) make-up.
High school A is located in a low-income urban community. It serves approximately 2000 students in grades 9 through
12. High school B serves approximately 2800 students in grades 9 through 12 from both suburban and rural communi-
ties. High school C serves approximately 1100 students in grades 9 through 12 in an affluent suburban community. High
school D serves approximately 1700 students in grades 9 through 12 in a middle class suburban community. All four of
these schools have provided the AP Calculus students a graphing calculator to use at home and at school. High schools
A, B, and D provided their students with a TI-83+, while high school C provided the TI-89 (which has Computer Algebra
System, or CAS, capabilities). Students may have used the calculator provided by the school or possibly their own personal
calculator. For the purposes of this study, the term graphing calculator refers to both calculators with and without CAS
capabilities.
Every student in every AP Calculus class at high schools A, B, C and D was asked to participate in the survey phase of
this study (n = 111). The survey provided data on student demographics, mathematical achievement, frequency of graphing
calculator use, modes of graphing calculator use, comfort with the graphing calculator, and reasons for graphing calculator
use. In addition, the students teachers provided a rating of weak, average or strong relative math ability for each student.
From this rich pool of data six students that were representative of the extremes in the survey responses while also being
representative of the types of schools participating in the study were selected to participate in the main study. The students
are introduced below as well as in Table 1.
Aaron (all names have been changed) attends high school C. His teacher considers him to be an average calculus student
compared to his peers. His mathematics grades throughout high school indicate that he has been very successful in his
course work. Aaron says that he is confident in his calculus abilities, but does not like when he is expected to use a graphing
calculator. Graphing calculators have been integrated into the curriculum of all of his high school math classes, but he admits
that he has not taken the time to learn how to use them.
Enoch attends high school A and is considered to be one of the strongest math students at his school. However, Enoch is
not very confident in his calculus abilities and says that he relies heavily on the graphing calculator. He explained that his
previous coursework did not prepare him for calculus and the graphing calculator helps him with skills that he should have
perfected in those courses.
Maryanne attends high school B. Her calculus teacher identified her as one of his strongest calculus students. She is the
only junior in this high school of over 2500 students who is taking calculus. Maryannes parents are both scientists and
have influenced her views on graphing calculator use significantly. She feels strongly that students should not rely on the
graphing calculator. Maryanne explained that she rarely uses her graphing calculator because it is very important for her to
be able to do everything without one if she is going to be successful in college.
Melissa attends high school C and was identified by her teacher as one of the weaker calculus students at her school.
Melissa is not confident in her calculus abilities and says that she relies heavily on her graphing calculator. It is important to
point out that though she is weak compared to the calculus students at her school; she is not a weak student. This was evident
by her AP exam score, which she shared with me through an informal conversation, and in her ability to conceptualize each
of the four tasks in the study. Melissa uses her calculator to graph and check quite often, but does not use the CAS capabilities
because she likes to do the manipulations on her own. Though she is not very confident in her abilities, she has shown that
she is a capable calculus student.
Rudy attends high school A and is one of only six students enrolled in calculus at his school. However, he is not a very
strong math student. Rudy is very aware that his less than proficient algebra skills are making learning calculus difficult. He
points out that his high school did not offer traditional algebra and geometry courses and he feels that the courses that were
 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 171

offered did not prepare him well for upper level mathematics. Rudy uses his graphing calculator often and in very creative
ways. He says that he would not have been able to do any calculus without it.
Shemika attends high school A. Her calculus teacher says that she is by far his strongest student. Shemika is very busy
both in school and after school and feels that the graphing calculator is a necessary tool given her other commitments.
Shemika is very knowledgeable about how the graphing calculator actually works, both its strengths and limitations. She
uses it very precisely, which is possibly because of her deep understanding of the mathematics.

4.2. Data collection

The opening vignette was constructed based on a classroom observation conducted during a visit to a calculus class at
high school C. In this vignette there were many instances in which students appeared to be experiencing affect related to
the graphing calculator that was influencing their mathematical engagement. However, simply observing students in the
classroom does not provide sufficient information about the nature of their affect or its impact (if any) on their mathematical
work. As a result, this study was designed to unveil the range of emotions the students related to graphing calculator use
through a series of interviews. Each student participated in three interviews. The first was a semi-structured interview
(Rubin & Rubin, 2005) followed immediately by a task-based interview (Davis, 1996; Goldin, 2000b). The third interview
was a video stimulated response (video-SR) interview (Lyle, 2003). A semi-structured interview was chosen for the initial
meeting to assure similar structure and content of the interviews with each participant, while still allowing for the additional
follow-up questions to provide elaboration when needed. A task-based interview is one in which participants are posed a
carefully selected task to solve and asked to explain their solution methods. Such interviews have been shown to be valuable
in studies of students mathematical thinking (Davis, 1996). Often task-based interviews are supplemented with a think
aloud protocol in which participants are asked to explain their thinking while they work. Video-SR is a procedure in which
videotaped behavior is replayed to an individual to stimulate recall of their activity (Lyle, 2003). Lyle (2003) suggests that
a video-SR methodology is appropriate for studies that benefit from minimal intervention in the activity (p. 862), such as
those in which asking students to think aloud might be considered disruptive.
In this study the semi-structured interview was utilized to collect background information. This interview provided initial
insight into the students perceptions of graphing calculator use in mathematics in addition to the ways they actually use
it when problem solving. In the subsequent task-based interview students were posed five tasks (described below) and
provided the graphing calculator they regularly used (i.e. TI-83+ or TI-89). This interview was both audio and video taped.
In addition, video of the students calculator screen was also collected. In order to address the need to ask questions about
strategy and emotion without influencing the problem solving process a video-SR design was used in the third interview. In
this case the purpose of the video-SR interview was to stimulate the recall of both cognitive and affective activity. Real-time
side-by-side video clips, of both written and calculator work, from each task were used to prompt the students to reflect
on the tasks, the role of the graphing calculator in their solution strategies, and the emotions they experienced during the
session. The students were asked to attend to not only their problem solving actions, but also to their facial expressions and
body language. The video-SR interviews took place the day after the semi-structured and task-based interviews.
All interview data were transcribed verbatim. The task-based interview data were also transcribed; meaning student
work was recreated using the video of the students working on paper along with video captured from their calculator
screens. Participants were invited to review and edit the transcriptions to ensure that the information was accurate. This
process, called member checking, added to the validity of the study (Creswell, 1998). No changes in the transcripts were
requested by the participants.
The tasks used in the task-based interviews were all designed so that they could be solved using mathematics that is
typically taught before the second semester of AP Calculus AB, when the students were taking part in this study. However,
they were neither necessarily familiar problem types, nor were they all graphing calculator friendly. The hope was that
these tasks would challenge the students and require them to use problem solving methods that were not necessarily
regularly used in their calculus courses. This work would then serve as a stimulus for a conversation about how the graphing
calculator impacted (or not) their mathematical experience. The four tasks are described in detail and possible non-calculator
solutions are presented in Table 2.

4.3. Analysis

The interview data were analyzed as it relates to DeBellis and Goldins framework for affect as a representational system.
The data were first coded for talk and actions related to modes of graphing calculator use (computation, use of graphs,
use of tables, and use of statistical tools). Next the data were coded for referents to emotions and values, those of self and
those perceived of others, with respect to graphing calculator use as they emerged from the data. For each student reported
affective pathways were constructed for each task. By reported affective pathway, I mean the path of emotions that the
student reports having felt with respect to the activity that has been recalled for a particular task. These pathways are
referred to as reported because the researcher does not have access to what students were actually feeling in the moment,
only what they reported that they remembered feeling. However, it must be noted that since the students were watching
video of themselves working, their reports were not simply a recall of memories, but were actually stimulated by the video
thus increasing the possibility that their responses were consistent with what they actually felt at the time. Once the coding
172 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179

Table 2
Description of the tasks.

The task Characteristics of the task Possible solution

Task #1: Find a rational function that
satisfies the given conditions:
a. It has a vertical asymptote at x = 3
b. It has a horizontal asymptote at
y = 2
Task #2: Find the maximum rate of
change of the graph of
y = x3 + 3x2 + 9x 27
Task #3: For what values of x is
2 < |k x| < 5
Task #4: Give an example of a function
for which |f(x)| = f(|x|)
The student must have a concept of what
horizontal and vertical asymptotes are
The student must first recognize that they are being
asked to find maximum slope of the first derivative
and then have a method for finding that solution
The student must make sense of the compound
inequality and consider the influence of different
values of k on the solution set for x
The student must decode the equality and consider
functions that would make the statement true
Given that the solution is a rational function
/ # and lim f (x) = 2, one
such that f(3)
x
possible solution is f(x) = 2x/(x 3)
The task asks for the maximum rate of change
of the derivative, y = 3x2 + 6x + 9. Using the
first and second derivative test, one could
determine that the maximum rate of change of
y = x3 + 3x2 + 9x 27 is 12 and occurs at x = 1
The absolute value will always be greater than
2, so that leaves 5 < k x < 5 to solve. The
solution is k 5 < x < k + 5
Possible solutions include f(x) = x2 and
f(x) = cos x + 1

was complete, the data were put back together and reread looking for emerging themes regarding affect and student decision
making as it relates to using their graphing calculators when solving problems. Throughout the analysis the interview data
were constantly compared between and among cases to determine whether or not the emerging themes were consistent
throughout.

5. Findings

Though the tasks designed for this study were not chosen because they were easily solved using a graphing calculator,
all six of the students used their graphing calculator at least once. A brief summary of each students graphing calculator
use across the four tasks is presented in Table 3. The construction of reported affective pathways for each of the graphing
calculator based solution methods revealed two consistencies: (1) the graphing calculator has the potential of helping
students maintain productive affective pathways and (2) students perceptions of external affect sometimes influences
decisions as to whether to employ graphing calculators. Since this was a qualitative study, my aim here is to richly describe
these findings; as such I have chosen to share instances that best exemplify the results that were consistent across all six
students. To that end, three of the six students will be highlighted in the following sections as exemplars of the types of
affective pathways that were identified, two to illustrate number 1 above and one to illustrate number 2. First Rudys work
and reported affective pathway on Task 3 and Aarons on Task 4 are presented to illustrate two very different ways in
which the availability of a graphing calculator can influence ones affective pathway. Next, Maryannes reported affective
pathway for Task 2 is presented as an illustration of the potential influence that perceived external affect could have on
ones mathematical experience.

5.1. Maintaining productive affective pathways

A comparison of the reported affective pathways constructed from the students work revealed that the availability of
the graphing calculator, and subsequent decision to use it, was often instrumental for maintaining a productive affective
pathway. As described above, a productive affective pathway is one that is not necessarily positive, but keeps the student
engaged in the mathematics. All six of the students noted anecdotally that having the graphing calculator available during
problem solving was important to them. They often shared examples of problems that they would not have tried and places
they would have stopped if they did not have a graphing calculator available. In this section two examples will be presented,
one in which a reported affective pathway remained productive and one in which it did not.
During the task-based interview Rudy was presented with Task 3: For what values of x is 2 < |k x| < 5. A description of
his work on this problem follows:

Rudy read this task for quite a while. He asked, Can k be anything? To which the researcher replied it could be any real
number. Rudy then wrote down 7 x = and picked up his calculator. He entered y1 = 7 x , graphed it, and quickly
went to the table. He scrolled up and down the table of values between x = 2 and x = 9 and paused for a moment. Next he
returned to the y = screen and changed the function from y1 = 7 x to y1 = 6 x and went back to the table. On the table
he scrolled between x = 1 and x = 8. Then he wrote on his paper, Depending on k there are 6 numbers that make x greater
than 2 and less than 5.

Rudys reported affective pathway for this task appears in Fig. 1. When asked how he felt when he first read the problem
Rudy replied, I was frustrated. At first I didnt understand it. I was trying to figure out what they meant. Once he picked up
his graphing calculator, he actually sat up straighter in his seat. Rudy noted that he was much more comfortable at that point.
 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 173

Table 3
Summary of student task solutions.

Task #1 Task #2 Task #3 Task #4

Aaron No GC No GC No GC GC
No solution Correct solution Incorrect solution Correct solution
Used trial and error to test
particular functions at single
points on the home screen
Enoch No GC GC No GC GC
Incorrect solution Correct solution Incorrect solution Correct solution
Started working by hand, got
stuck and turned to GC to view
the graph of the function and use
the calc tools
Maryanne GC GC No GC GC
Correct solution Correct solution Correct solution Correct solution
Solved all other tasks first, Solved by hand, noted an error Substituted functions into each
returned to this task and used and turned to GC to graph the absolute value expression and
trial and error to determine a function and its derivative then compared graphs to identify
numerator that resulted in a used calc tools to solve functions for which the equality
correct graph was true
Melissa GC No GC No GC No GC
Correct solution Incorrect solution Incorrect solution Correct solution
Turned on GC before even
reading the problem, sketched
coordinate plane with
asymptotes on paper then tested
functions on GC until identifying
one that met criteria
Rudy No GC GC GC GC
Incorrect solution Correct solution Incorrect solution Incorrect solution
Tried to solve by hand and did Rewrote the inequality without Chose functions to evaluate at
not remember how, then used the absolute value, chose values both 2 and 2 on the home
CG to graph the derivative and for k and then used the table to screen to determine whether or
used calc tools to solve determine how many integral not made the given statement
solutions there were for this new true
inequality
Shemika No GC GC No GC No GC
No solution Correct solution Incorrect solution Correct solution
Immediately graphed derivative
on GC and used calc tools to
solve

Activity SR-Associated Affect

Read task and thought

Frustration
about it for a while

Substituted 7 for k and

put into GC, scrolled Curiosity and comfort
up and down table

Substituted 6 for k,
scrolled up and down Curiosity
table

Noticed a pattern and

Contentment
recorded solution

Fig. 1. Rudys reported affective pathway.

174 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179

Fig. 2. Aarons written work.

He explained that he often uses his graphing calculator to test ideas. As he started noticing a pattern, he became curious and
tried a few more numbers to test the pattern further. When he wrote down his answer he said he felt good because he was
able to write something down, especially since it was a problem that he did not know how to do when he first read it. (Note:
Rudy was unaware that he had not solved the problem as it was stated.) When Rudy was asked how he would have done
this problem if he had not had his graphing calculator he said I wouldnt have tried it. For Rudy, his graphing calculator use
on this task was central to his maintaining a productive affective pathway, even if it was productive toward a goal different
than that stated in the given task.
Due to the context of this study (i.e. a graphing calculator was available to students at all times) there are not any examples
for which the absence of a graphing calculator resulted in an unproductive affective pathway. However, in the examples there
were situations in which students had not developed the graphing calculator capabilities they wished to use as instruments
(i.e. the process of instrumental genesis was incomplete), which ultimately resulted in an unproductive affective pathway.
Aarons work on Task 4 (Give an example of a function for which |f(x)| = f(|x|)), presented below, is an excellent example. His
work follows:

After reading through the task Aaron immediately wrote down x2 as a solution. The researcher then asked him if he
could come up with a few more examples, which prompted him to write down sin x. Next he picked up his calculator, a
TI-89,
and entered sin(!/3) on the home screen and got 3/2 as a solution. Then he entered sin( !/3) and got
3/2. He stared at this for a long while. After a few moments he drew a 306090 triangle on his paper and labeled
the lengths of the sides (see Fig. 2 below). Once again, he paused. He then picked up his calculator and calculated both
cos(!/3) and cos( !/3) and got 1/2 as an answer both times. He then wrote down sin x and cos x as solutions.
Finally, he calculated both (3)3 and (3)3 on his calculator and wrote x3 as a solution as well.

Aarons reported affective pathway for this task appears in Fig. 3. When asked about how he felt when he first read the
problem Aaron replied that he was fine because he immediately knew a function that would satisfy the conditions stated.
However, once he was asked if he could find any other solutions he reported feeling both curious and a little nervous. He
was curious to see if he could find other solutions, but also nervous because he was being put on the spot to do so. Aaron said
that he figured either the sine or cosine function should work, so he decided to use his graphing calculator to test them out.
However, when he picked up his graphing calculator he could not figure out how to input an absolute value sign to which
he said he was kind of annoyed. He went on to explain, Its frustrating when I cant figure out how to use my calculator to
help me. When he couldnt figure out how to use his graphing calculator to support his work he instead drew a 306090
triangle because I knew it has something to do with trigonometry and it might be helpful. . .but it wasnt. Finally, Aaron
gave up on the sine and cosine functions and tested two values for the function x3 (without using the absolute value function)
and wrote that down as a solution. He put his pencil down and said, I give up. As he watched himself working during the
video-SR interview he kept repeating that I was so frustrated that I couldnt make my calculator do what I needed and
finally that he was embarrassed to have to give up. Though Aaron had a graphing calculator available to him, he did not
know how to use it in such a way that it could be of help to him. In the context of this problem, he clearly had not developed
the instruments he needed to make the graphing calculator useful. Though there is no evidence to support such a claim, it is
plausible that if Aaron had developed the absolute value capability as an instrument with respect to absolute value functions
this episode might have had a very different outcome.
Though these are only two examples they serve to illustrate the possible impact that technology tools, like the graphing
calculator, can have on a students affective pathway and ultimate engagement in problem solving. In fact, all six of the
students in this study reported multiple situations in which having a graphing calculator available provided them with a
means to maintain a productive affective pathway, including Aaron who reported having had many instances of contexts in
which graphing calculator use did result in productive affective pathways. The only exceptions were those in which graphing
calculator capabilities being used were not developed as instruments with respect to the task that was posed, like Aarons
work on the absolute value task above.
 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 175

Activity SR-Associated Affect

Read task and

immediately wrote down Comfort
a solution

Prompted to find
Curious and nervous
more solutions

Tried to use abs function

on GC but did not know Discouraged
how

Drew 30-60-90 triangle Helpless and annoyed

Tested function for single

Uncomfortable
value on GC

Said, I give up. Embarrassed

Fig. 3. Aarons reported affective pathway.

5.2. Perceptions of external affect

It is evident in the examples presented above that both Rudy and Aarons local affect (i.e. affective pathway) while solving
their respective tasks was impacted by the availability (or lack there of) of a technology instrument. However, the availability
and use of the graphing calculator was not the only thing that impacted students local affect during problem solving, their
local affect was also affected by their perceptions of what others might think of their problem solving activities. For example,
as the students took part in the video-SR interviews they reported emotions related to their decisions regarding the graphing
calculator that they tied to what they perceived their parents, teachers, and even their future professors might think of those
decisions. Maryannes work on Task 2 (Find the maximum rate of change of the graph of y = x3 + 3x2 + 9x 2) follows as an
example of the impact of external affect, in the sense of DeBellis and Goldin, on a students reported affective pathway.

Maryanne began this task by working on paper. She first wrote down y = 27 + and then crossed it out (see Fig. 4). Next
she wrote the first derivative, set it equal to zero, and factored it (forgetting to square the x). Directly below she sketched a
number line on which she denoted x = 1 and x =3 and the sign of all values on either side of these points. She looked at
this for a moment and then crossed out the number line. Next she wrote down the second derivative and a second number
line denoting x = 1. She then evaluated the first derivative at 1, looked at it for awhile, and left it and went to work on
another task.
When Maryanne returned to the task she immediately crossed out her previous work. She re-read the problem, and above
the phrase rate of change she wrote slope. Next, she wrote the first and second derivative and sketched another number
line with x = 1 denoted on it. Then she picked up her graphing calculator, a TI-83+, and graphed the original function,
y1 = x3 + 3x2 + 9x 27. After changing the window using the zoom standard command she looked at the graph for a
moment. Next she returned to the y = screen and inserted nderiv in front of the original function she had entered, changed
her mind and instead entered nderiv(y1, x, x) in y2 which commanded the calculator to sketch the graph of the original
function, y1, and the first derivative of y1 on the same screen. After looking at the graph of both functions together, she
zoomed out so that she could see the functions in a larger window. She followed this inspection of the graph with the use
of the max tool on the calculator to find the maximum of y2, the derivative. The calculator determined the maximum to be
12 at x = 1. Maryanne considered this for awhile, returned to her written work where she crossed out the number line she
had drawn with x = 1. She wrote down her solution, maximum slope at x = 1 is 12. Finally, she finished by sketching a
number line one more time with x = 1 denoted.
176 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179

Fig. 4. Maryannes written work.

Maryannes reported affective pathway for this task appears in Fig. 5 below. As Maryanne watched herself look at the
graph she created on her graphing calculator she recalled being very comfortable with the problem until she evaluated the
first derivative at 1 and found it equal to 0. She recalled that at that point she knew she had made a mistake, but was unsure
what it was. I was really confused. As she watched herself pick up the graphing calculator and construct the graphs of both

Activity SR-Associated Affect

Read task, immediately

did 1st derivative Comfort
and zeros on paper

2nd derivative and zero

(incorrect) on paper

Evaluated 1st derivative

at -1 and found equal Confused
to 0

Graphed functions
Comfort and guilt
on GC

Recognized error Relief

Found and recorded

Happiness
solution

Fig. 5. Maryannes reported affective pathway.

 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 177

the first and second derivative she groaned. When asked about this reaction she replied that she shouldnt have to rely on
the graphing calculator. . .ever. However, at the same time she admitted that she felt relief when she recognized her error,
Whew, I knew that wasnt right (smiles), but I knew I knew what I was doing. As a matter of fact she pointed out exactly
when she recognized what her error was. She said, Well, I knew that the slope is clearly the steepest in this area so I could
tell that to begin with. . . I found the maximum of the slope just using the maximum function. Then I looked back and it said
it was at one and I was like thats where I made my simple algebra mistake!
Though the graphing calculator obviously impacted Maryannes ability to maintain a productive affective pathway, her
response that she should never have to rely on the graphing calculator suggests that she associated a certain amount of
guilt for choosing to use it. When asked to provide further explanation Maryanne clarified that these feelings of guilt were
prompted by what she perceived her mother would think. She said, My mom doesnt, really does not like teaching to the
calculator. She knows my teachers havent done that, but she talks about how she doesnt like seeing kids using calculators
at all. Thus Maryannes affective pathway was not only impacted by her graphing calculator use, but also by what she
perceives others, like her mother, might think of that use.
Other students in the study also reported their graphing calculator use was sometimes influenced by their perceptions of
external affect. For example, as Rudy reflected on his work on the absolute value inequality task presented above, in addition
to being content with his solution, he mentioned that although he could not have solved the problem without the graphing
calculator he did not feel good about having to use it. This was because he did not think his teacher would have approved of
the way that he used it. Enoch mentioned that he felt good when he used his graphing calculator because he was under the
impression that it was expected that he would be able to use it when he got to college, and Ill look better to my professors
next year if I am good at using it. It appears that for these students when technology tools are involved, perceived external
affect impacts their affective pathways during problem solving and sometimes even their decisions regarding whether or
not to use the tool at all.

6. Conclusion, limitations, and continuing research

Goldin (2000a) has argued that it is important to understand students affective pathways when problem solving so that
we can figure out how to help students on unproductive affective pathways deal with frustration and impasse and turn those
into more productive emotions like curiosity, bewilderment, motivation and maybe even elation. The results of this study
suggest that one way to help students maintain productive affective pathways is by providing an instrument with which
they can act in response to feelings of frustration or impasse. It is notable that Rudy indicated that he most likely would
have given up on the absolute value inequality task if he had not had his graphing calculator. Using the graphing calculator
changed the situation from one of defeat and frustration to one in which he was curious because he had a tool that he could
use to easily explore. It seems as if one of the most important roles of graphing calculator use, no matter what actions are
being taken, is to help maintain a productive affective pathway that supports mathematical success.
While students look to their graphing calculators to help maintain a productive affective pathway, Aarons experience
with the absolute value task is an important one. We see in this example, that having a tool available is not enough. If the
particular graphing calculator capability that the student wishes to employ has not been developed as an instrument with
respect to the concepts at hand, it is useless and can even have adverse affects on ones affective pathway, possibly resulting
in feelings of defeat as we saw with Aaron. Researchers (e.g. Artigue, 2002; Guin & Trouche, 1999) have documented the
importance and complexity of mathematics instruction that aims toward instrumentation of graphing calculator and CAS
tools. Aarons experience here is further evidence of the importance of such work. We cannot expect that simply providing
tools will necessarily result in the more frequent occurrence of productive affective pathways; instead we need to think
carefully about how those tools are incorporated into the classroom so that students are supported in the instrumentation
process.
Through the use of reported affective pathways it was possible to examine the role of meta-affect with respect to graphing
calculator use. The data show that these students did employ meta-affect and that their meta-affect was greatly influenced
by perceived external affect of parents and teachers. The influence of external affect on graphing calculator use is not really
surprising given the literature on the promotion of technology in the classroom. Previous work has considered how pro-
motion by teachers might impact student technology use, but has not considered other external influences on its use (e.g.
parents, future teachers) or even its impact on the emotional experience of problem solving. The students in this study were
influenced not only by their current and past teachers, but also by parents and future teachers. This suggests that this is
an area in need of further study. As Goldin (2002) points out, Powerful affective representation that fosters mathematical
success inheres not so much in the surface-level affect as it does in the meta-affect (p. 63). As we think about meta-affect
and its potential role in problem solving, it is necessary to understand the constructs that might possibly be impacting it.
Admittedly this study has major limitations. It was a very small exploratory study. With only six students, I can only
speculate that the consistencies found in their experiences might be true for a larger group. This can only be determined with
further research. However, the consistency seen here does suggest that further research should be undertaken. Additionally,
it is important to recognize the possible limitations of a video-SR methodology for unveiling in-the-moment emotions.
Some might argue that these emotions may not actually represent what the student felt at the origin of the mathematical
experience. However, they do represent what the student recalled feeling at the time. Therefore, at the very least this recall
reveals the emotional memories that the student is taking with them from the mathematical experience. Finally, given the
178 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179

role that perceived external affect ended up playing for these students in their graphing calculator use it is important to note
that this study was not done in the context of the classroom. As such, the only information regarding the values and beliefs
related to graphing calculator use that was available was from the students perspective.
The purpose of this study was to look at independent graphing calculator use through the lens of affect as a represen-
tational system. It was found that using the graphing calculator helped students maintain productive affective pathways
for problem solving as long as they were using graphing calculator capabilities that they had developed as instruments
with respect to the mathematical task they were working on. Furthermore, the results of this study indicate that perceived
external affective representations of others (e.g. teachers past, present and future; parents) does influence meta-affect and
has the potential to either help maintain or interrupt a productive affective pathway. In light of this, future studies should
be designed to understand the complexity of how graphing calculators are promoted by external forces (i.e. teachers and
parents), how students interpret such promotion and finally how this promotion influences their independent use. Addi-
tionally, further studies like the one described are needed to determine if the nature of the reported affective pathways
here are common, or if there are other ways in which the availability of graphing calculators might influence one. Finally,
as we learn more about the influence of the availability of a graphing calculator on local affect it is important that studies
are designed that compare these results to the use of other technology tools that are commonly used in the teaching and
learning of mathematics.

Acknowledgments

Many thanks to Aaron, Enoch, Melissa, Maryann, Rudy and Shemika without whose time and effort this study would not
have been possible. Thank you also to Manya Raman, Valerie DeBellis and Karen Keene for feedback on early versions of this
article.

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 Journal of Computers for Mathematical Learning, 7, 245274.
Artigue, M., & Kilpatrick, J. (2008). What do we know? And how do we know it? In Paper presented in plenary 1 at the 11th international congress on
mathematical education Monterrey, Mexico.
Ball, L., & Stacey, K. (2004). A new practice evolving in learning mathematics: Differences in students written records with CAS. In M. J. Hines, & A. B.
Fuglestad (Eds.), Proceedings of the 28th conference of the international group for the psychology of mathematics education. Bergen, Norway: PME.
Berry, J., & Graham, T. (2005). On high-school students use of graphic calculators in mathematics. ZDM The International Journal on Mathematics Education,
37(3), 140148.
Burrill, G., Allison, J., Breaux, G., Kastberg, S., Leatham, K., & Sanchez, W. (2003). Handheld graphing technology in secondary mathematics: Research findings
and implications for classroom practice. Michigan State University.
Chevallard, Y. (1992). Concepts fondamentaux de la didactique: Perspectives apportes par une approche anthropologique. Recherches en Didactique des
Mathmatiques, 12(1), 77111.
Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, CA: Sage Publications.
Davis, R. B. (1996). Cognition, mathematics, and education. In L. P. Steffe, P. Nesher, P. Cobb, G. A. Goldin, & B. Greer (Eds.), Theories of mathematical learning
(pp. 285299). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
DeBellis, V. A., & Goldin, G. A. (1997). The affective domain in mathematical problem solving. In E. Pehkonen (Ed.), Proceedings of the twenty-first annual
meeting of PME. Lahti, Finland: PME.
DeBellis, V. A., & Goldin, G. A. (2006). Affect and meta-affect in mathematical problem solving: A representational perspective. Educational Studies in
Mathematics, 63, 131147.
Doerr, H. M., & Zangor, R. (2000). Creating meaning for and with the graphing calculator. Educational Studies in Mathematics, 41, 143163.
Drijvers, P. (2000). Students encountering obstacles using a CAS. International Journal of Computers for Mathematical Learning, 5, 189209.
Dunham, P. H. (2000). Hand-held calculators in mathematics education: A research perspective. In E. Laughbaum (Ed.), Hand-held technology in mathematics
and science education: A collection of Papers (pp. 3947). Columbus, OH: The Ohio State University.
Ellington, A. (2003). A meta-analysis of the effects of calculators on students achievement and attitude levels in precollege mathematics classes. Journal
for Research in Mathematics Education, 34(5), 433463.
Evans, J., Morgan, C., & Tsatsaroni, A. (2006). Discursive positioning and emotion in school mathematics practices. Educational Studies in Mathematics, 63(2),
209226.
Goldin, G. A. (2000a). Affective pathways and representation in mathematical problem solving. Mathematical Thinking & Learning, 2(3), 209220.
Goldin, G. A. (2000b). A scientific perspective on structured, task-based interviews in mathematics education research. In A. E. Kelly, & R. A. Lesh (Eds.),
Research design in mathematics and science education. Mahwah, NJ: Lawrence Erlbaum Associates.
Goldin, G. A. (2002). Affect, meta-affect, and mathematical belief structures. In G. C. Leder, E. Pehkonen, & G. Torner (Eds.), Beliefs: A hidden variable in
mathematics education? The Netherlands: Kluwer Academic Publishers.
Goos, M., Galbraith, P., Renshaw, P., & Geiger, V. (2003). Perspectives on technology mediated learning in secondary school mathematics classrooms. Journal
of Mathematical Behavior, 22, 7389.
Graham, T., Headlam, C., Honey, S., Sharp, J., & Smith, A. (2003). The use of graphics calculators by students in an examination: What do they really do?
International Journal of Mathematical Education in Science and Technology, 34(3), 319334.
Guin, D., & Trouche, L. (1999). The complex process of converting tools into mathematical instruments: The case of calculators. International Journal of
Computers for Mathematical Learning, 3, 195227.
Harskamp, E. G., Suhre, C. J. M., & van Streun, A. (2000). The graphics calculators and students solution strategies. Mathematics Education Research Journal,
12(1), 3752.
Kaput, J. J. (1989). Information technologies and affect in mathematical experiences. In D. B. McLeod, & V. M. Adams (Eds.), Affect and mathematical problem
solving: A new perspective. New York, NY: Springer-Verlag.
Kilpatrick, J. (2009). Algebra is symbolic. 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 (pp. 1121). Atlanta, GA: Georgia State University.
Lyle, J. (2003). Stimulated recall: A report on its use in naturalistic research. British Educational Research Journal, 29(6), 861878.
Malmivuori, M. (2006). Affect and self-regulation. Educational Studies in Mathematics, 63(2), 149164.
 A.W. McCulloch / Journal of Mathematical Behavior 30 (2011) 166179 179

Mandler, G. (1989). Affect and learning: Causes and consequences of emotional interactions. In D. B. McLeod, & V. M. Adams (Eds.), Affect and mathematical
problem solving: A new perspective. New York, NY: Springer-Verlag.
McLeod, D. B. (1992). Research on affect in mathematics education: A reconceptualization. In D. A. Grouws (Ed.), Handbook of research on mathematics
teaching and learning. New York: Macmillan.
Opt Eynde, P., & Hannula, M. S. (2006). The case study of Frank. Educational Studies in Mathematics, 63(2), 123129.
Opt Eynde, P., De Corte, E., & Verschaffel, L. (2006). Accepting emotional complexity: A socio-constructivist perspective on the role of emotions in the
mathematics classroom. Educational Studies in Mathematics, 63(2), 193207.
Rubin, H. J., & Rubin, I. S. (2005). Qualitative interviewing: The art of hearing data (2nd ed.). Thousand Oaks, CA: Sage Publications.
Sandman, R. S. (1980). The mathematics attitude inventory: Instrument and users manual. Journal for Research in Mathematics Education, 11, 148149.
Sheryn, L. (2006). What do students do with personal technology and how do we know? How one student uses her graphical calculator. International Journal
of Technology in Mathematics Education, 13(3), 18.
Vygotsky, L. (1978). Mind in society. Cambridge, MA: Harvard University Press.
Zan, R., Brown, L., Evans, J., & Hannula, M. S. (2006). Affect in mathematics education: An introduction. Educational Studies in Mathematics, 63(2), 113121.