Академический Документы
Профессиональный Документы
Культура Документы
to offer tasks that promote inquiry in algebra. The Visual Math curriculum
attempts to offer such opportunities by its choice of sequence, the structure
of the tasks and the setting for the inquiry. Each task promotes more than
just a single way of solution, often demands a new aspect that has not been
taught before, and can be extended to support a long problem-solving pro-
cess. The effect of such a complex change is hard to measure and requires
different types of studies (Huntley et al., 2000, describe the dilemma of a
development team who seeks ways to study a full curricular sequence). In
this study we sought ways to follow students’ use of functions in problem
solving along three years, while trying to offer similar tasks that can be
approached differently but always meaningfully at different phases of the
learning. While in other studies we, among others, often try to observe
how students solve non traditional and more challenging tasks, here we
tried to relax the impact of the type of task. This was done in order to
concentrate on the development of the concept of function and the impact
of this development on strategies of problem solving at various phases.
2. T HE RESEARCH DESIGN
The major sources of data for this analysis were three task-based clinical
interviews. The first took place towards the end of the first learning phase,
after students completed a central unit of qualitative modeling and were
already familiar with all the representations of function. The second inter-
view was done when students were more familiar with manipulations on
expressions, but students could still solve problems and equations using
comparisons of functions (rather than by manipulating equations). The
third interview took place towards the end of the sequence, when solv-
ing procedures and manipulations were already implemented. The study
was planned as a longitudinal study of pairs of students. In the interviews,
the students were presented with written situation problems to be solved
in any way they chose, using paper and pencil, graphing software, or a
number calculator. Our goal was to learn about problem-solving processes
by exploring the strategic and mathematical actions and thoughts of the
students. We sampled students of various levels of exhibited mathematics
achievement. From each of four classrooms we picked a more able pair, a
less able pair, and an ‘average’ pair.2 We chose to interview pairs, since
students were working in pairs in class. We hoped that these students
would have common habits of work and communication and would be
able to develop an interesting discussion.
In this analysis we will focus on only one of the twelve pairs: Gal and
Roni. This pair was considered to be among the low achievers in their
128 MICHAL YERUSHALMY
algebra class. Gal and Roni worked as a pair in class as well, and were the
most suitable interviewees for this analysis because they presented serious
work throughout the three years. Their learning patterns – serious and per-
sistent work, a tendency to present and share their thinking throughout a
conversation, and being relatively low achievers that would spend a long
time on most problems – made them an interesting pair for this analysis.
However, in order to allow the reader gain an appreciation for the ways in
which Gal and Roni are and are not representative of the larger sample,
we will often juxtapose their work with another pair from the study, Ben
and Eyal, who were among the high achievers. Ben and Eyal also worked
together during the three years and were extremely cooperative with both
adults and their classmates. Though this analysis method results in a non-
standard case study – one in which data represents more than the two
students being most closely examined – this method is our attempt to cope
with some of the complexities posed by qualitative analysis methods of
longitudinal study.
We chose a linear break-even situated task (Chazan, 1993; Fey, 1989)
as the repeated task for the three interviews analyzed here. The generic
structure of each of the three problems was a simple everyday payment
along three years, was, that during such a long period that involved so
many changes, using ‘the same task’ might not be a reliable procedure
(Newman et al. 1989.) However, we are convinced that studies attempting
to follow, among other aspects, the use of technology throughout an innov-
ative sequence should be longitudinal, so that the tool would be a natural
artifact of the environment. Beyond the repeated task interviews, we tried
to minimize the external effects stemming from the length of the study by
having the same teacher teaching for the three years and the same groups
of students learning together.
In each of the next three sections we will summarize and analyze an inter-
view with Gal and Roni on one break-even linear problem. The problem,
its solution, the grade level, the time of the interview, and the experience
students brought with them to the specific interview are presented first. We
then bring in examples of the students’ actions while solving the problem.
Each section also provides a short summary of the related work of Ben and
Eyal. Each section closes with a subsequent summary intended to support
observations of the stages of the solution process and the attributes used in
this process.
Computing the price for these many examples, Gal argued that although
“. . . there are infinite examples to be considered, there is no need to com-
pute so many cases” because there is an ‘order’:
Gal: If it (the discount amount) increases constantly it indicates that there is a
constant order.
MS: And then what?
Gal: Then you don’t need to compute.
They considered their solving attempts complete once they designed
the categorical table and identified its behavior. This heavy reliance on
the presentation of numerical data was found in the work of other lower
achievers but absent in the work of Ben and Eyal. Ben and Eyal began by
assigning letters to quantities in the story: “Eyal: Let’s say x, no actually
f(x) is the total sum of purchase and . . . f(x) is probably the sum of dis-
count?” but further attempts to manipulate and compute using symbolic
language failed. At the end they reached the correct solution by guessing
the answer and evaluating it using arithmetic. Symbols were used in the
PROBLEM SOLVING STRATEGIES AND MATHEMATICAL RESOURCES 131
conversation as a way to communicate about the process, but they did not
yield any more powerful representation than the verbal descriptions.
Half a year through the learning of qualitative modeling, involving
function graphs and verbal descriptions of the rate of change, the stu-
dents were faced with an unfamiliar problem in this interview. The only
useful resource to them at that time were relationships between independ-
ent and dependent quantities. Both pairs tried to introduce order into the
dependency: for Gal and Roni, the table represented order and behavior,
and they argued that this order represented the solution. As Chazan (2000)
observes, there is a difference between a relationship between quantities
and a relationship between numbers, as the first is grounded in the world
around us and can be expressed using graphs, devices, or natural language
and also described more formally. We will follow now how the informal
investigation of Gal and Roni was further developed.
Gal and Roni’s work on the problem started almost immediately with or-
ganizing a few examples into two tables.
132 MICHAL YERUSHALMY
They did not have enough space on the sheet to continue the graphs,
and so they decided to form another table of values. Unlike the first two
tables, this time they formed two tables using variations (increase of 2.7
on one and 2 on the other) with the explicit purpose of comparing the two
outputs for equal inputs.
Roni: Two, four, six, eight, ten, twelve, fourteen . . .
Gal: Here we see that it increases more.
Roni: Let’s just see when it switches. Which one is faster?
Gal: Here, about the ten. Let’s see . . . in one it is better here, in two here, in three
here.
The boys marked the ‘switch’ with a horizontal line connecting the two
tables and suggested that this is the solution. They did not yet write any ex-
pressions. The interviewer (MS), who was anxious to understand why they
preferred repeated computations over the use of the graphing software,
suggested they use the computer. Their reason for their unwillingness to
use the software became clear immediately – they did try to avoid the use
of expressions that are the ultimate way to use the software!
134 MICHAL YERUSHALMY
Gal: Ha, why should we look at a correspondence rule? It would have probably
been simpler to do every time from 8 to 9 steps in 10 Agorot. (10 Agorot are
1/10th of 1 IS)
MS: Ha?
Gal: Because if the question is simply “where [is it better to park]?” . . .. relating
the number of hours you need to park then this is simpler. Why should we do a
rule?
They were looking at the graphs and the table of values, but Gal insisted
that zooming into the neighborhood of the input 8, where they identified a
‘switch’ in the table, could provide the solution without any use of sym-
bols. However, responding to the interviewer’s prompt, they completed
their work by looking at the graphs and mainly the tables generated by
PROBLEM SOLVING STRATEGIES AND MATHEMATICAL RESOURCES 135
the software, and they read the solution in the same way they did with the
paper and pencil.
For Ben and Eyal, symbolic representation, which had been used some-
what artificially in their first interview, was the trivial choice here. They
started by formulating a symbolic comparison model and immediately
asked to use the software. Although they were very fluent and quick with
the expressions, Ben hesitated whether the expression he inserted, 2x+6, is
really the correct one3 , and abandoned the computer in favor of pencil and
paper work for creating and checking tables and graphs. They formed a
table of values directly from the story, without using the expressions they
had already written in the software. Using reverse arithmetic operations,
they found the number of hours that balanced the difference of 0.7 IS per
hour with the extra entrance fee of 6 IS.
Both pairs reached the solution using non-algebraic methods. How-
ever, their problem solving processes were different. Gal and Roni initially
formed the problem model non-symbolically. They constructed a table of
values and used it as a model of the situation and as a means to represent,
to compare and to compute the common price of two increasing processes.
This time, unlike during the first interview, they were explicitly aware of
the role of each representation: They used correspondence tables for the
initial exploration and co-variations for graphing and ordering the tables.
They formed two tables for two graphs, and compared the two processes
represented in the table as if they were two continuous processes repres-
ented by graphs. It is often argued that multiple representations create a
cognitive load that eliminates meaningful uses of representations (Pimm,
1995), but Gal and Roni kept all components connected. They kept in mind
the problem story as an anchor guiding their repeated attempts, repeated
analysis of specific cases, and exploration of patterns of behavior. But as
in the first interview, algebraic symbols were still an artifact here although
they had already been using them for a few months in their algebra class
to describe processes with and without software. They managed to use
functions to describe the processes in the problem and to suggest an ap-
propriate numerical solution without moving from ‘quantities’ to algebra,
and they were extremely suspicious as to whether using a symbolic rule
could produce any further benefit.
cise, since the comparison model and the procedures for solving equations
were already practiced for almost a year. And indeed, for Ben and Eyal,
as well as for other higher achievers, this problem was solved quickly by
using a comparison of two functions, forming a linear equation, and solv-
ing the equation using standard techniques4 . We will describe now how this
exercise situation turned into a problem-solving situation for Roni and Gal.
The Bonus problem story Background
A factory owner wishes to give a spe- 9th grade
cial bonus for the holidays. His account
suggests two methods for the special Current work in class consists of
payment. Method A: The January pay- linear and quadratic motion problems of
ment will include additional 450 IS per various types; usually more complicated
person. Method B: In the January pay- than the interview problem.
ment each employee will receive an ad- Most word problems solved in class
ditional amount equal to 1/5 of his regu- cannot be solved by simply reading the
lar monthly salary. solution from the graph and therefore
Find a way to present the cost of both graphs are mostly used as sketches of
methods in order to help the owner to situated models.
choose the better bonus strategy. Graphing software is being used less
The problem equation: frequently, and more attention is
x + 450 = x + x5 given to equivalency of expressions
and equations throughout symbolic
manipulations.
As in the previous two interviews, Gal and Roni differed as to their initial
approach to the solution: Gal tried immediately to decide what would be
the appropriate tool and strategy to use. Roni concentrated on the specific
process needed for the solution: defining the dependency so they can find
out ‘what each worker gets.’
The sketch Roni made represented the slopes inaccurately (the lines
should intersect in the first quadrant). But it appropriately served the pur-
pose of demonstrating the fact that the processes are linear but not ‘paral-
lel’, and therefore the equation must have a single solution. The following
script suggests that at this stage, the sketch was a tool for describing the
situation and a plan for working out a solution. In the previous interview,
the graphs were formed accurately by plotting points and used for con-
structing the processes and evaluating the numeric result. The justification
of the sketch and the plan this time were connected mostly to the structure
of the symbolic rules (e.g. same or different coefficients of x), rather than
to the problem story and its numerical values as in the previous interview:
138 MICHAL YERUSHALMY
The study examined students that studied for three years an algebra course
that was structured along the variation function approach supported by
graphic technology. We frequently interviewed students during these three
years on different tasks. Although we were involved in the interviewees’
classroom activities almost on a daily basis, the data presented here is pur-
posely limited to three interviews on a similar task and mainly with only
one pair. This is done in an attempt to capture the transitions, development
of knowledge, and the subtle changes that occurred along the relatively
long period of learning this alternative algebra course. In order to present
this long-term process of learning and follow the changes that took place
along the three interviews, we have chosen a timeline representation of the
solution process of Gal and Roni.
The charts (in Figure 1) on the left-hand side describe Gal and Roni’s
problem solving strategies and on the right the mathematical resources
used at the implementation phase. The light rectangles represent the use
of graphing software that was available along all interviews. An over-
view of the six charts helps to capture similarities and differences and to
concentrate on the trends of the learning and development of skills.
The length of work is the first noticeable difference between the three
interviewees. In an algebra course that focuses on algebraic solutions,
solving a linear break-even problem should be a short process. None of
the processes here suggest such short work. Although the first solution
was the shortest, it was not short at all: Gal and Roni made a few attempts
to solve the problem, analyzed the dependency between two quantities,
and offered a way to chart an answer. The process was the longest in the
second interview. The students seemed to have developed a facility with all
the representations and they needed time to maneuver. In the third inter-
view, Gal and Roni were still spending a relatively long time on choosing
alternative representations.
Another noticeable difference along the three interviews is the use of
software. The software available along all interviews and used in two of
them was integrated in two different manners. The chart of the 2nd in-
terview suggests two cycles of solution – one with paper and pencil and
the second with the software, prompted and almost forced on them by
the interviewer (starting at the 13th minute). In the third interview, again,
140 MICHAL YERUSHALMY
Figure 1. Time charts of Gal & Roni three interviews (Left – Problem Solving strategies,
Right – Mathematical Resources). Light rectangles represent the use of software.
there were two cycles, the choice to use the software was the students’
this time, and the work with the software continued rather than repeated
the paper and pencil work and the mental attempts at solving. They were
ready to use the software once they got an idea of what they expected to see
and how it would be different from what they had already found. In both
interviews the use of the computer was delayed until the last part of the
solution process (the last third on the second interview and the second half
on the third interview). Higher achievers often started their solution using
software (e.g. Ben and Eyal in the second interview) mainly because the
representation they learned to value most – expressions – was their only
channel of communication with the function graphing software. We will
turn now to the analysis of the problem solving strategies along the three
charts.
PROBLEM SOLVING STRATEGIES AND MATHEMATICAL RESOURCES 141
don’t do that, simply add two for each hour.” A comparison between the
resultant construction of each method served as feedback and a checking
strategy.
However, Gal and Roni advanced only slowly in their use of expres-
sions, and were at a premature stage for using expressions even at the
third year. They had a harder time when asked to treat the equation and
the situation simultaneously – when they had to concentrate on the ma-
nipulations, they concentrated on the graphs and expressions and not on
the given phenomena. Only when they returned to work numerically and
graphically using the software, the narrative returned to be the familiar
parallel discussion of the story and the mathematics. Thus, like others
who take a function approach to algebra with graphing technology, we
find that modeling tasks are successfully worked on and manipulations are
less frequently used and with less success.
The caution often suggested, that students who are doing algebra for
a long period using graphing software will develop a dependency on the
software, seems irrelevant here. For high achievers such as Ben and Eyal,
the software was useful immediately when available, but when it wasn’t
they performed successfully as well. Those who, like Gal and Roni, pre-
ferred to delay the use of expressions, did not use the software until they
gained an appreciation of the expressions rules. Worries that graphing
144 MICHAL YERUSHALMY
ACKNOWLEDGEMENTS
The author would like to thank Michal Sion and Suheir Nassar for their
contribution to the collection and analysis of the data and to earlier drafts
of the paper.
146 MICHAL YERUSHALMY
N OTES
1. In Israel algebra is being studied at junior high school in grades 7,8 and 9. In parallel
(but not necessarily integrated) they learn geometry and introductory probability and
statistics. The traditional algebra curriculum starts with algebraic reasoning of expres-
sions, equations, and continues with the study of linear and quadratic funcitons. The
Visual Math alternative algebra curriculum was deisgned to fit into this structure.
2. The criteria to determine the interviewees’ level included the impression got by the
interviewer (who followed the classes weekly in the 7th grade) from students particip-
ation, verbal ability and tendency, their cooperation with the teacher and classmates
and their work in the computer lab. In principle, those who were categorized as ‘less
able’ students were those who at that time required extra explanations, more demon-
stration and directions in computer tasks, and often, during the whole class discussion,
presented dissatisfaction.
3. Ben argued that 2x+6 means that 6 has to be constantly added to the parking price for
any hour x while the situaiton was decsribed differently in the given problem.
4. For Ben and Eyal, this problem did function as an exercise. Once they formed an equa-
tion they asked for a computer in order to read the intersection point. The interviewer,
in an attempt to create an opportunity to view more problem solving, interrupted and
asked them to reach a solution before using the software. As a result, they returned to
a strategy they had used in the previous two problems: computations of the difference
between the two processes. However, this time, the arithmetic undoing operation was
replaced by viewing the difference function 450 − x5 and they reached the result by
equating it to zero.
5. In the Function Graphing software they were using in class there were two different
entries: one for graphing functions and another for graphing a relation as a comparison
of two functions.
R EFERENCES
Center for Educational Technology: 1995, Visual Mathematics: Algebra. Written materials
and software, Published by the author, Tel-Aviv.
Chazan, D.: 2000, Beyond Formulas in Mathematics and Tecahing: Dynamics of the
Highschool Algebra Classroom, Teachers’ College, NY.
Chazan, D.: 1993, ‘F(x)=G(x)? An approach to modelling with Algebra’, For the Learning
of Mathematics 13(3), 22–26.
Confrey, J. and Smith, E.: 1994, ‘Exponential functions, rates of change and the multiplic-
ative unit’, Educational Studies in Mathematics (26), 135–164.
Fey, J.T.: 1989, ‘School algebra for the year 2000’, in S. Wagner and C. Kieran (eds.)
Research Issues in the Learning and Teaching of Algebra, Erlbaum, Hillsdale, NJ, pp.
199–213.
Heid, K.: 1995, ‘Algebra in a technological world. National Council of Teachers of
Mathematics (NCTM)’, Curriculum and Evaluation Standards for School Mathematics.
Janvier, C.: 1996, ‘Modeling and the initiation to algebra’, in N. Bednarz, C. Kieran and L.
Lee (eds.), Approaches to Algebra: Perspectives for Research and Teaching, Dordrecht,
The Netherlands, Kluwer, pp. 225–238.
PROBLEM SOLVING STRATEGIES AND MATHEMATICAL RESOURCES 147
Kieran, C., Boileau, A. and Garancon, M.: 1996, ‘Introducing algebra by means of
a technology-supported functional approach’, in N. Bednarz, C. Kieran and L. Lee
(eds.), Approaches to Algebra: Perspectives for Research and Teaching, Dordrecht, The
Netherlands, Kluwer, pp. 257–294.
Kintch, W. and Greeno, G.: 1985, ‘Understanding and solving word arithmetic problems’,
Psychological Review 92, 109–129.
Nassar, S.: 2000, The Characteristics of Solution Processes of Low-Achieving Pupils who
Study Algebra in the Function with Technology Approach, Thesis submitted in the
Department of Education, University of Haifa, IL.
Nemirovsky, R.: 1996, ‘A functional approach to algebra: Two issues that emerge’, in N.
Bednarz, C. Kieran and L. Lee (eds.), Approaches to Algebra: Perspectives for Research
and Teaching, Dordrecht, The Netherlands, Kluwer, pp. 295–313.
Newman, D., Griffin, P. and Cole, M.: 1989, The Construction Zone: Working for Cognitive
Change in School, Cambridge University Press.
Perkins, D.N.: 1986, Knowledge as Design, Hillsdale, NJ, Lawrence Erlbaum Publishers.
Pimm, D.: 1995, Symbols and Meanings in School Mathematics, London, Routledge.
Huntley, M.A., Rasmussen, C.L., Villarubi, R.S., Sangtong, J. and Fey, J.T.: 2000, ‘Effects
of standards-based mathematics education: A study of the core-plus mathematics pro-
ject algebra and functions strand’, Journal for Research in Mathematics Education vol.
31(3), 328–361.
Schoenfeld, A.: 1985, Mathematical Problem Solving, Academic Press, Inc.
Schoenfeld, A.: 1987, ‘What’s all the Fuss about Metacognition?’ in A. Schoenfeld (ed.),
Cognitive Science and Mathematics Education, Erlbaum.
Yerushalmy, M.: 1997, ‘Designing representations: Reasoning about functions of two
variables’, Journal of Research in Mathematics Education 27(4).
Yerushalmy, M. and Shternberg, B.: 2001, ‘A visual course to the concept of function’,
in A. Cuoco (ed), The Roles of Representations in School Mathematics, 2001 National
Council of Teachers of Mathematics Yearbook (2001), pp. 251–268.
University of Haifa,
Faculty of Education
michalyr@construct.haifa.ac.il