Está en la página 1de 27

A Schematic-Theoretic View of Problem Solving and Development of Algebraic Thinking

Author(s): Diana F. Steele and Debra I. Johanning


Source: Educational Studies in Mathematics, Vol. 57, No. 1 (2004), pp. 65-90
Published by: Springer
Stable URL: http://www.jstor.org/stable/4150315
Accessed: 27-11-2017 21:07 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/4150315?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Educational
Studies in Mathematics

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
DIANA F. STEELE and DEBRA I. JOHANNING

A SCHEMATIC-THEORETIC VIEW OF PROBLEM SOLVING AND


DEVELOPMENT OF ALGEBRAIC THINKING

ABSTRACT. This study explored the problem-solving schemas developed b


pre-algebra students as they participated in a teaching experiment that wa
help students develop effective schemas for solving algebraic problem situatio
contexts of (1) growth and change and (2) size and shape. This article describes
and types of schemas that students developed through examples of problems u
schemas, examples of students' reasoning and writing, and excerpts from stude
Findings from the study indicated that there is a link between the type of g
students construct and the schemas they are forming. By using these sche
recognize, extend, and generalize patterns and quantitative relationships both
symbolically.

KEY WORDS: schemas, generalization, algebraic thinking

1. INTRODUCTION

A major concern of mathematics educators involved in efforts to improve


mathematics teaching and learning is that pre-algebra coursework focuses
too much on traditional algebra curriculum. Traditional algebra courses
typically emphasize manipulating variables, solving equations, and simpli-
fying expressions (Kieran, 1996). Because of this narrow focus, researchers
in mathematics education have called for research studies that focus on de-
signing teaching experiments, longitudinal studies of students' learning,
and instructional and curricular interventions to build knowledge about in-
struction in the domain of algebra (RAND, 2001). This study is both-a
teaching experiment and an instructional and curricular intervention. We
examined whether pre-algebra students effectively learn to express their al-
gebraic thinking in ways different from the traditional algebra curriculum.
What is algebraic thinking? Although it may include variables and ex-
pressions, algebraic thinking has a broader and different connotation than
the term algebra. Kieran (1996) defined the term algebraic thinking as "the
use of any of a variety of representations that handle quantitative situations
in a relational way" (p. 4, 5). Swafford and Langrall's (2000) definition of
algebraic thinking was "the ability to operate on an unknown quantity as if
the quantity was known, in contrast to arithmetic reasoning which involves
operations on known quantities" (p. 2). Driscoll (1999) said that algebraic

Educational Studies in Mathematics 57: 65-90, 2004.


@ 2004 Kluwer Academic Publishers. Printed in the Netherlands.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
66 DIANA F. STEELE AND DEBRA I. JOHANNING

thinking could be considered to be the "capacity to represent quantita


situations so that relations among variables become apparent" (p. 1).
definitions are all similar, and we use them to guide our own use of
term "algebraic thinking" in this paper.

2. A THEORETICAL FRAMEWORK TO SUPPORT DEVELOPMENT OF


AGEBRAIC THINKING

Cognitive development research provides a theoretical basis for


opment of students' ability to think algebraically by contributing t
of schema. Marshall (1995), a cognitive psychologist, stated th
is a mechanism in human memory that allows for the storage
generalization, and retrieval of similar experiences. A schema
individual to organize similar experiences in such a way that t
ual can easily recognize additional similar experiences. Jean Pia
1977) used the idea of schema as a basis for his research on the d
of scientific reasoning. For Piaget, schema development was co
goal-oriented action by the individual. Action, not as an observa
ior, but as an activity of the mind and tied to the experience of
as they interact with the physical world around them (Thomp
Piaget (1985) said that abstraction is also important for develo
schemas. Abstraction is the process by which the mind selects
and combines actions, and then records them in memory (von G
1995). Von Glasersfeld stated that reflection is the conscious pro
resenting experiences, actions, or mental processes and conside
results. Reflection on action, however, is the key to developing
fact, Sfard (1991) declared that mathematics itself is the product
ing on our actions. By combining the ideas of abstraction and
Piaget used the term "reflective abstraction." Reflective abstrac
general coordination of actions into new forms or structures (v
feld, 1995). Piaget (1985) declared that "The development o
structures [schemas] is due to reflective abstraction (p. 143)."
Experiences can be altered to fit an individual's schema, and a
can be incorporated into an existing schema. These alterations
to Piaget's (1952) notion of assimilation. Piaget used assimila
scribe the process by which the individual takes in new data. W
perience cannot be altered and assimilated into the existing sch
the schema itself must be reconceptualized and then accommoda
experience. Piaget called this reconceptualization an accommoda
schema. Individuals develop schemas only for those patterns in
or events that are repetitive experiences. Not repetitive as in pr
repetitive in the organization of situations or events which ar

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 67

for assimilation, accommodation, and reflection. O


generally create a schema but can potentially prov
which the memory sequence of a schema is built.
fine and further develop the schema and knowled
(Greeno et al., 1996). Schemas are triggered when
comprehend, understand, organize, or make sense o
Meaningful schema development means that a
derstand concepts to recognize and construct pa
emphasized that an individual does not memorize
way that he or she memorizes a formula or definit
done by rote with little or no understanding of wh
means. One may memorize isolated bits of knowled
to any other knowledge. Sfard (1991) characterized
structured, sequential cognitive schemata that are
meaningful understanding. She said that memorize
"must be processed in a piecemeal, cumbersome m
to a great cognitive strain.... " (p. 26). She pointed
sequential schemas are "hardly a place for assimilat
(p. 26). This type of schema is horizontal, shallow
unconnected information. Instead, for an individua
she must make the connections of the new experience
Sfard believed that this type of structured cognitiv
constructed by adding new layers that form a dee
nected hierarchy of knowledge. In this hierarchic
there is room for assimilating new information.

3. CONNECTION OF SCHEMAS TO GENERALIZATION IN PROBLEM SOLVING


AND ALGEBRAIC THINKING

The connection of schemas with mathematical problem sol


gebraic thinking can be provided by using generalization. Ma
declared that "[g]eneralization is the heartbeat of mathemati
Tall (1991) made the connection of assimilation to generalizat
ematics. He said that assimilation is an expansive generaliz
of generalization when one extends an existing cognitive struc
change in current ideas. He said that an accommodation is a ty
alization, a reconstructive generalization, one that requires rec
of an existing cognitive structure. Dubinsky (1991) believed t
individual learns to apply an existing schema to a wider rang
riences, then the schema has been generalized. This generaliz
occur because the individual becomes aware of the wider appl
the schema.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
68 DIANA F. STEELE AND DEBRA I. JOHANNING

Krutetskii (1976) described two different aspects of generalizab


One aspect is "[a person's] .... ability to see something general a
unknown to him in what is isolated and particular" (p. 237). In othe
the individual is able to form a generalization based on particular
seeing the general in the particular. A second aspect is "a person's a
see something general known to him in what is particular and conc
(p. 237). This aspect allows the individual to classify a particular c
an example of a generalization; in other words, seeing the particula
general. Mason (1996) suggested that it is important for students
experiences in both "seeing a generality through the particular and
the particular in the general" (p. 65). In this study we wanted stud
generalize in both ways. To develop a schema a student must first
problems. By looking for patterns and solving particular cases, st
generalize from particular cases to a general one. Then when studen
experience, they generalize by invoking their schemas. When stude
the general case to seek to incorporate a specific case, they then t
their schemas to either assimilate or accommodate new problems so
Sfard (2003) believed that understanding can be built simultaneou
both directions, from particular to general and from general to pa
because there is a circular nature to understanding. One must per
actions and reflect on them to develop a schema, however, one mus
schema to develop and refine it.
Silver (1981) found that good mathematical problem solvers
mathematical structure of problems to generalize solution strateg
solving related problems, students transfer what they learn from o
lem to another similarly structured problem. Confrey (1997) sugges
one approach to helping students examine problems related in math
structure and to develop algebraic thinking is a context-based appr
using the contextual approach teachers help students organize com
ties among situations, see relationships about how change in one qua
related to change in another, and make generalizations-all are imp
aspects of algebraic thinking. According to the Algebra Working
(1997), students' understanding of algebraic concepts should be de
through contexts such as growth and change, size and shape, and
ber patterns. Similar problems have also been recommended by th
University (Mason et al., 1985).

4. RESEARCH STUDIES WITH A FOCUS ON GENERALIZING


ALGEBRAIC PROBLEMS

Some researchers have conducted studies using the problem s


finding patterns in growth and change, size and shape, and n

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 69

and Wheeler (1987) called problems of this type g


A common quadratic generalizing problem is th
Cooper and Sakane (1986) investigated 8th-grade stu
eralizing quadratic problems, such as the Handshak
that students made generalizations too quickly and
important when a counterexample to their genera
Most students could not explicitly recognize that
examined for the general rule; some claimed that a
numbers was a sufficient rule in and of itself without
In an investigation, Mason et al. (1985) found that
demonstrated four stages of expressing generality i
Balacheff (1988) found similar results in a study h
14-year old students using a generalizing problem. H
dents used four stages or hierarchical types of proo
for the general case. Mason et al. and Balacheff's st
with Bell's (1976) discussion of stages of general
first stage involved making conjectures about gener
a few cases. This stage was followed by students tes
against particular examples. In the third stage stu
ness of the need to consider all possible cases. In th
made explicit generalizations. Balacheff and Mason
students only used the first two stages of generaliz
Lee and Wheeler (1987) conducted a study of G
lutions to generalizing linear and quadratic proble
gories of students who were successful in solving
"those who hit on a usable pattern perception and p
those there who were flexible in pattern perception
tern when one was unproductive" (p. 109). They fou
usually check their generalizations to see if they we
cases. Stacey (1989) also found in her study that stu
generalizing problems did not usually check their
they were correct for particular cases. Her findin
parisons of three groups of students' solution strat
9- to 11-year-old students in traditional mathemat
grade students in traditional mathematics classes,
and 8th-grade students who had received special in
solving.

5. PURPOSE

One of the questions raised by Mason and Pimm (1984) in the implications
of their research on algebraic representation was "How can you expose the

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
70 DIANA F. STEELE AND DEBRA I. JOHANNING

genericity of an example to someone who sees only its specificity"


With this study we chose to build on the past investigations of s
solutions to generalizing problems to address Mason and Pimm's
This study had two purposes: a teaching focus and a research focus
teaching focus, we intended to help students develop effective sche
they could use to generalize relationships between quantities in a
problem situations and to represent these generalizations in v
symbolic expressions. The purpose of the research focus was to in
problem-solving schemas developed by students as they solved lin
quadratic generalizing problems.
This study builds on other studies in the literature, by incorp
the notion of investigating problem-solving schemas in algebraic
ing and the dimension of purposely manipulating instruction by
students solve related algebraic problems. The questions that guid
research are the following. (1) In what ways do students construct
solving schemas for solving problems related in mathematical str
(2) What types of schemas do students develop? (3) In what ways
dents' schemas help them identify, extend, and generalize pattern
and among problems? (4) In what ways do students' schemas help
generalize both verbally and symbolically the patterns they see a
senting relationships between quantities in problems?

6. METHODOLOGY

6.1. The teaching experiment

Using qualitative research methods, we designed an investigation using the


teaching experiment method in which we could plan, adapt, and sequence
instruction to help pre-algebra students construct effective problem-solvin
schemas and develop their ability to think algebraically. As both teacher
and researchers, it was important that we seek to understand students'
schemas for solving algebraic problems independently of our own mathe
matical understandings. To do this we needed to build conceptual construct
of what we thought were their problem-solving schemas based on our ob-
servations as they interacted in their physical environment. We realized
that we could only make attributions to students of what we thought to
be their schemas; however, these attributions were grounded in the lan-
guage and actions students were able to make evident to us. Steffe an
Thompson (2000) declared that "[1]ooking behind what students say and
do in an attempt to understand their mathematical realities is an essen-
tial part of a teaching experiment" (p. 268). Von Glasersfeld (1995) af
firmed that the process of looking behind what students say and do is

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 71

conceptual analysis and is the strength of a teaching


research.
In keeping with traditions of the teaching experiment, the investiga-
tion involved both interview phases and teaching phases (Cobb, 2000;
Steffe and Thompson, 2000). In the interview phases, we explored stu-
dents' development and use of problem-solving schemas as well as their
development of algebraic thinking. In the teaching phases, in order for stu-
dents to construct well-developed schemas and bring forth these schemas to
solve problems, we manipulated particular problem assignments. We pro-
gressively sequenced problems to encourage students to see relationships
among algebraic situations and to either assimilate the problem solution
into their present schema or accommodate (modify) their schemas to solve
the problems. Their teacher was involved in the planning and execution
of this experiment. This involvement helped us verify students' thinking
(Lesh and Kelley, 2000), and she also acted as a witness to the teaching
experiment (Steffe and Thompson, 2000).

6.2. Participants

The participants for this investigation consisted of eight 7th-grade students


enrolled in a pre-algebra mathematics class in a small Mid-Western middle
school. The curriculum up to this point had been the traditional pre-algebra
curriculum from their text. Although we believe students had solved for
patterns throughout their K-6 schooling they had not to our knowledge
generalized verbally or symbolically. The teacher in this class, however,
taught in a manner consistent with an inquiry-oriented approach to instruc-
tion. The students had been writing in mathematics all year. They were
accustomed to being asked to explain, both verbally and in writing, their
strategies and mathematical procedures, and explain why they chose their
particular strategies and procedures. The teacher had encouraged them to
make lists and tables to explain their thinking. Five of the students were
boys and three were girls. The students were above average in mathematical
ability and thus, in this school system, they were placed into a pre-algebra
course rather than a general 7th-grade mathematics course.

6.3. Procedures

We did not first model how to solve the types of problems in this stud
instead we gave students the problems and asked them to solve them wi
any method they chose. We chose problem contexts that had potential
lead to pattern generalization to support students' understanding of qua
titative relationships. The geometric contexts were (1) size and shape a
(2) growth and change. The students in this study solved eight generalizi

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
72 DIANA F. STEELE AND DEBRA I. JOHANNING

problems-five linear and three quadratic problems. Each prob


struction sequence lasted 2 or 3 days; we met daily for 1 month. Fiv
problems are presented in this paper (see Steele, 2000, for the othe
lems). We emphasized to students the importance of the process th
their solutions. We, therefore, required them to discuss in writing firs
then later, in cooperative groups their problem-solving approache
wanted them to understand that not only could they use different repr
tations for solving problems, but that there were also alternate str
During the teaching portion of the investigation, we asked stude
first individually write out solutions, explanations, and generaliza
problems presented on paper distributed to them. We walked arou
observed the students' work, sometimes giving feedback and aski
tions; we collected students' work. After reading their work, the n
we asked them to meet in small cooperative groups of four to
their solutions. In these groups students read their descriptions, ex
their thinking, explained their generalizations, and displayed their s
expressions. They listened to each others' strategies and solutions,
pared their own strategies and solutions with those of others, and
whether they wanted to keep or change their approaches. As stude
cussed their reasoning in small groups, we observed, listened, and
questions. After each session, we kept a journal of field notes docu
our observations and keeping track of new questions.
During the interview phase of the teaching experiment we formally
viewed each student four different times using the same problems. We
questions such as: So you decided to try your expression on some
numbers to see if it would work? How did your diagram help you?
your table help you? Did you look for an easier way? Can you expla
the numbers represent in your expression? Do you see a relationshi
problems you have already solved? How did you reason about the
lems? Did you see other students' strategies you might try? We als
journal of field notes documenting our observations from each int
Using the field notes, our observations from cooperative group
interviews, and students' work, we reflected on the students' think
cesses and adapted instruction when choosing the next day's prob
next problem situation was slightly different but closely related in
matical structure to the previous one. This cycle continued until th
the 1-month teaching experiment.

6.4. Data collection and analysis

We audiotaped all teaching and interview sessions, transcribed th


tapes, and collected students' written work that contained their di

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 73

tables, and writing. We systematically and rigorous


classified, and consolidated the data using the Devel
cle (Spradley, 1980) to determine patterns and them
of data analysis we were able to "determine its parts, t
parts, and their relationship to the whole" (Spradl
simultaneous data collection and analysis, we began
about students preferred approaches to solving pr
build a model of students' schemas. We revised qu
repeatedly over the course of the study. As we ana
terns and themes that emerged concerned the qualit
that students developed.

6.5. Findings

We identified two qualities of problem-solving sche


veloped: a well-connected schema and a partially
use the term well-connected schema to identify a
has strong connections and relationships so tha
fully used to generalize a particular problem sit
across problems. We use the term partially form
schema that is weak in connections and does no
sary relationships to articulate generalizations acr
times within a problem. We found and labeled
most often used by students: a subtracting-out sc
schema. We will more fully describe both qualities
in the following discussion. To present these find
students' work. We chose these three students bec
resentative of the eight students-both in quality
developed.

7. EXAMPLES OF STUDENTS WITH WELL-CONNECTED SCHEMAS

7.1. Cathy's problem-solving schema: Subtracting out

A 7th-grader named Cathy was an example of a student who developed a


well-connected "subtracting-out" schema for solving algebraic problems.
Cathy first demonstrated this schema while solving the shaded squares
problem (Figure 1).

Figure 1. Given diagram of shaded squares problem.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
74 DIANA F. STEELE AND DEBRA I. JOHANNING

In the diagram a 3 x 3 grid of squares is colored so that only outside sq


are shaded. This leaves one square on the inside that is not shaded and 8
that are shaded. If you had a 25 x 25 grid of squares and only the outsi
of squares are shaded, how many squares would be shaded? If n represen
number of squares on a side and you have all the outside squares of an n
shaded, write an expression representing the total number of shaded squar
figure. (NCTM, 1998)

TABLE I

Cathy's table to represent the patterns for the


shaded squares problem

Measure Shaded D's

3x3 8

4x4 12

5x5 16

6x6 20

121110 8

2456 42
456 4 5

Figure 2. Cathy's diagra

Cathy began the pro


started with a 3 x 3
made a table (see Tabl
in each of her diagram
(Figure 2):

In a 25 x 25 grid, 96 s
squares to a side. But 4
them twice. I will use a

If you count the corner


this information, the f
because a square has 4 s
are shared....

Although Cathy used a table to organize her findings, she derived he


verbal and symbolic generalization by looking for patterns as she drew t
diagrams for smaller cases. She realized that even though there were 1
squares or lengths that created the perimeter of the 3 x 3 square, the
were only eight shaded squares. As she created and analyzed the 3 x
diagram she realized that although every side is three squares in lengt
when you multiply the number of squares on a side by four, the numb

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 75

is four higher than the number of shaded squares


problem she constructed the idea of shared comers
shared or common components. Her writing reflec
the structure of the grid to visualize what was ha
indicated that she analyzed several particular cases
interpret the general case.
Once Cathy constructed the approach of sharing a
continued to use it. In the example below she drew
the shaded squares problem to solve the triangle do

Figure 3. Given diagram for triangle dot problem.

In the diagram below is a 5-dot triangle. It is a triangl


each side. The 5-dot triangle is made using a total of 12
be used to make a 13-dot triangle? If n represents the nu
of an n-dot triangle, write an expression to represent h
the triangle? (NCTM, 1997).

In her written explanation of how she solved the


explained how she used the knowledge she had cons
this mathematically related problem. She wrote:

I looked at this problem and realized it was exactly like


time only with a triangle instead of a square. The reason
the triangle ... and 5 dots to a side is because you don't
You multiply the number of dots on a side times 3 bec
triangle, and you subtract 3 because there are 3 corners
shared. So for a 13-dot triangle there will be 36 dot
formula is (N - 3) - 3 = # of dots.

Cathy identified sharing and subtracting out as the


that related it to the previous problem. This ident
draw on her developing schema to subtract out the
Cathy elaborated on how the problems were relate
action of solving a three-sided figure by using the
izing her solution of a four-sided figure. She gene
of the shaded squares problem to solve any partic
lem. She drew several diagrams of smaller triangles
generalization.
Cathy also applied her developing subtracting-
mathematically related problem in which the struct
previous two problems-the square/vertices problem

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
1 2 3 4

Figure 4. Given diagram of

How many corners/vertices ar


are arranged in the following

Cathy first copied the giv


to display her information
her subtracting-out idea an
(Figure 5). She explained:

I noticed that the number of c


than the number of squares. .
only 7 comers because one is

If you had 3 squares, 2 corner


of squares minus 1 and take tha
number of squares times 4 to
(n -1) = # of corners. n = #
some of the comers are shared

TABLE II

Cathy's table to represent the patterns for the


squares/vertices problem

No. of O's No. of corners

1 4

2 7

3 10

4 13

5 16

shared corner

Figure 5. Cathy's diagram for shared vertices in the squares/vertices problem.

Cathy gave a detailed explanation for the relationships between the


quantities in this problem. Although she recorded her findings in a table
as a way to organize her final results, analyzing the diagrams helped her
understand the patterns and the relationships between quantities. She sys-
tematically chose her examples to apply a "sharing" relationship in the

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 77

diagrams of smaller problems and subtracted out t


While this problem took her longer to solve, she w
the slightly different structure into her subtracting-o
Cathy's schema had strong connections and relat
enabled Cathy to see quantitative relationships in
eralize both from the particular to the general an
the particular. If Cathy had only said that the nu
was 4 less than the number of squares per side mi
have been able to interpret the triangle dot problem
comers overlapped. Her schema might have been t
ber of sides by the number of dots on a side and
she might have altered her symbolic expression fr
problem to find the generalization for the triang
seems unlikely that she would have then been able
bolic manipulation to the squares/vertices problem
her generalizations by solving for several specific
lems in the sequence, Cathy continued to manipul
for each problem to find shared components, sub
generalized the patterns she observed into verbal s
expressions.

7.2. Mike's problem-solving schema: Building up

Another student, Mike, developed a well-connecte


related problems. Unlike Cathy, however, he did n
out when deciding how to count the shared compon
that we call "building up." Like Cathy, Mike also lo
drawing diagrams. To solve the shaded squares prob
ing a diagram of the 3 x 3 square and labeled its s
After drawing this diagram he wrote the following
explained his reasoning:

25 + 24 + 24 + 23 = 96 shaded squares
(s) + (s - 1) + (s - 1) + (s - 2)

s-2 s -1i
s-1

Figure 6. Mike's diagram for the shaded squares problem.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
78 DIANA F. STEELE AND DEBRA I. JOHANNING

First I thought about the number of shaded squares in the small fig
noticed that on the top you counted all 3 squares but on the [right] side y
counted 2 because 1 is shared. You do the same for the third [bottom] s
the final [left] side you only count 1 of the squares. So I came up with a
(s) + (s - 1) + (s - 1) + (s - 2). I also used this to get the number of
squares in the 25 x 25 grid, 25 + 24 + 24 + 23 = 96.

Mike then went back to his generalization from his smaller figu
provided a closed form for the symbolic generalization for any nu
shaded squares on the perimeter: (S . 4) - 4.
For this problem Mike used a smaller example (the given di
to make a generalization rather than the particular case that w
for in the question (the 25 x 25). He then found the number of
squares for the 25 x 25 before simplifying his symbolic expression
we asked him how he knew to subtract 4, he explained that h
1 + 1 + 2 (the numbers he had subtracted from the sides in Figure
subtracted that amount. He had not really collected like terms, but
had added like terms by thinking about each side of each diagram
though the final form of his equation utilized subtraction, and he
ticed the shared corners, he did not use a subtracting pattern to bu
generalization. Instead, he added on or built up the number of squ
each side to find the generalization. He partitioned figure into top,
bottom, and left, and then added these partitioned sections to ma
generalization.
The key difference between Cathy and Mike's approach was tha
did not associate minus 4 in the shaded square problem with the ove
corners. Rather he worked around the square analyzing each side
was in the process of building each side. This approach is differen
Cathy's; she described the structure as one whole in which each sid
square visually had the same number of squares, but when counti
side as 3 in length, she developed the idea of overlap or sharing.
Once Mike saw this building-up pattern based on the diagram,
tinued to use it and was able to express his thinking symbolically
like Cathy, applied his generalization from the shaded squares to so
triangle dot problem. He wrote:

First I noticed that this problem was the square problem with 3 sides inst
So, all I did was modify my formula to (n - 3) - 3. In a 13-dot triangle t
36 dots. I checked and the formula (n - 3) - 3 works for all the triangles.

Mike utilized his simplified symbolic expression ((S . 4) - 4) fro


shaded squares problem. He did not need to go through the pr
adding up the sides to find the generalization for the triangular dot pr

of (n ? 3) - 3. He generalized from the general to the particular.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 79

TABLE III

Mike's table for squares/vertices problem

D's Corners
1 4

2 7

3 10

4 13

100 301

N -> (n 2)4 + (n * 2)2 + 1

1 2 3 4 10 +1

JO~ E
Figure 7. Mike's

With the squa


table (see Table
beneficial way
Figure 7).
Mike wrote:

First I tried to find a relationship in the table between the # of squares and the # of
vertices. I found no noticeable pattern. But I noticed [in the diagram] that there are
half of the O's on the top and half on the bottom [if it is even]. So, if you multiply
the number on the top times 4 you get a number. Then if you multiply the number
of D's on the bottom by two you also get a number. [I]f you add the 2 numbers
and add 1 ... you get the number of comers. You ... ad [sic] 1 ... because there is
an extra comer sticking out at the end ....

His use of tables here seemed more of a way to verify or organize counts
rather than a springboard for the symbolic generalization. It was his use
of the diagram and the building-up schema as he partitioned the figure
into top, bottom, and left over sections that led to his generalization and
his ability to identify and generalize the relatioznship between the quan-
tities in the problem. He had accommodated his building-up schema to
include a problem with somewhat different but similar mathematical struc-
ture. Later in other problems in the teaching experiment, Mike used tables
more to help him identify the patterns and generalize to the nth case. For
example, the following is Mike's solution for the diagonals of a polygon
problem.

How many diagonals are in an 8-sided polygon? Write an expression to represent


how many diagonals are in an N-sided polygon? (NCTM, 1997).

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
80 DIANA F. STEELE AND DEBRA I. JOHANNING

TABLE IV

Mike's table for the diagonals of the polygon problem

Sides Diagonals How get total Diagonals per vertex

3 0 (s - 3) (s 2) 0
4 2 (s - 3) (s + 2) 1
5 5 (s - 3). (s - 2) 2
6 9 (s - 3) - (s 2) 3

Figure 8. Mike's diagrams for the dia


figures).

To solve this problem Mike again began with smaller problems. He drew
several polygons (see Figure 8). Then Mike made a table to help him look
for patterns between the number of diagonals and the number of sides on a
polygon (see Table IV). His table listed his data and also included columns
that explained how he used his data to obtain the number of diagonals of a
polygon.
Mike explained how he reasoned about the problem:

First I tried to find a pattern in the sides, diagonals, and diagonals per vertex that
I did not need to carry out to the target # [8 in this case]. The formula that I came
up with was

(sides - 3) . (sides +- 2)[see column 3].

The way I found this was the (sides - 3) equals lines [diagonals] per vertex. The
(sides +- 2) is the number of pairs that equal to the maximum number of lines
[diagonals] you can draw from each vertex. In a[n] octagon the sides minus three
is five. Then you multiply it by 4 because there are 4 groups of 5 and you get
twenty. [T]hat formula works for all polygons.

When we asked Mike how he first saw the number of diagonals, he said,
"I was drawing the diagonals for each shape, and I noticed the pattern."
He observed the pattern of increasing numbers of diagonals coming from
each vertex as the number of vertices increased. He then developed the
idea of pairs of diagonals as he counted. In the table he added and divided
to obtain the numbers in the diagonals column as a way to explain the
process of drawing the diagonals and keeping count as he went along. He
built up the diagonals as he went from each set of vertices, and then used the
table to help him complete the symbolic generalization. Mike had already
generalized the relationships from drawing the diagrams. But he adapted

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 81

and refined his original schema to adroitly accomm


kind of building-up situation. This interplay of bot
him get the symbolic generalization. Mike wrote t
generalization for diagonals on the polygon proble
checked his genrralization to solve for several spe
Mike again used a building-up schema to find
another problem-the roads problem. Mike recog
between the diagonals of the polygon problem and
again he elaborated on the problem by drawing a di
generalization as he applied the building-up schem
The roads problem is as follows:

In the Asphalt Kingdom each city is connected to the o


smooth blacktop. To make it simple for people to get ar
there is a road connecting each city with all of the othe
city in the Asphalt Kingdom. For example, 10 years ag
only had four cities with six roads to connect them so ther
any city to any other city. Today the Asphalt Kingdom h
How many roads does it have now? Write a general exp
king, King Steamroller, can use to determine how many
for any given number of N cities. (Johanning, 1999)

When Mike solved the roads problem he remembe


the diagonals of a polygon problem and wrote:

First I remembered how to get a formula on the number


[the previous (s - 3) - (s + 2)] . Since the roads were con
the formula (N +- 2) x (N - 3) + N gives you the total nu
formula by taking the number of pairs that equal the n
multiplying them by the number of diagonals take away
(n - 3) then you add the number of sides onto that bec
roads and you get the number of lines (roads). There are
[K]ingdom.

(N + 2) x (N - 3)+ N
7 x 11 + 14 = 91

We asked Mike how he remembered the formula for the diagonals of


a polygon. He said, " I remembered how to get it, but I did not remember
what is was.... I remembered the pairs, but I did not remember what the
formula was. I just remembered to add three pairs." He then explained how
he had drawn figures for each number of cities until he got to the six-city
diagram and then remembered this was how he had solved the diagonals of
a polygon problem. He pointed to the six-city diagram that he had drawn.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
82 DIANA F. STEELE AND DEBRA I. JOHANNING

Figure 9. Mike's six-city diagram for the roads problem.

I divided it in half because there's three pairs. Then you multiply it by


3, which is the number of diagonals that come out of each side. It's the
of diagonals out of each vertex. There's three coming out of here and th
points. And you have to add the perimeter. (Figure 9)

Mike modified and adapted the symbolic expression he had c


from the diagonals of a polygon problem, (n - 3) . (n + 2), to m
real-life situation of already having one road to each city on the pe
of the cities (six roads in this particular case). He identified the rela
of these diagrams to the diagrams he had drawn for the diagonals
polygon problem and used the same building-up schema that had
him to generate his old equation by adding n for the perimeter roads. H
constructed the relationships between the quantities and the proble
this time did not need a table at all. With these repeated opportun
work on problems that were related in mathematical structure, his
became more generalized, and he could apply it to a broader range
periences. Mike seemed to generalize the relationship between the
and the particular almost simultaneously.

8. EXAMPLE OF A STUDENT WITH A PARTIALLY-FORMED SCHEMA

8.1. Zeke's problem-solving schema: Building up

Zeke also used a building-up schema to solve the problems in the teaching
experiment. Like Mike, Zeke used diagrams and sometimes tables. In the
diagonals of the polygon problem Zeke made the same type of diagrams
as Mike for 3, 4, 5, 6, and 7-sided polygons. He also created a table (see
Table V) to summarize the information from the diagrams and to look
for patterns. He did not see the "pairs" relationship that Mike had seen in
which two vertices shared the same diagonal, but did systematically draw
and count the diagonals for each of the vertices. In his table, Zeke recorded
how he constructed the diagonals in the "how you get it" column after
drawing the following figures.
He wrote:

The formula I figured out is subtract 3 from the number of sides and write the
number down twice. Then count down from that number until you reach 0. Then

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 83

TABLE V

Zeke's table for the diagonals of the polygon problem

No. of sides How you get it No. of diagonals


3 0 0

4 1+1= 2
5 2+2+1= 5
6 3+3+2+1= 9
7 4+4+3+2+1= 14

add up all the numbers. Example: 4 + 4 +

N = # of sides - 3 + # of sides - 3 + count down to 0

After creating the table, Zeke then looked for a numeric pattern that
would lead to a way to generalize the numbers in the "how to get it"
column. He added up all the diagonals to get his generalization (a building-
up approach); however, he found the number of diagonals as an out-
come of his arithmetic operations, not from observing the way he had
drawn the diagonals from the vertices. His generalization was recursive
and worked only when all previous number of diagonals for each poly-
gon were known. He was not able to see past the numbers to understand
what was general in them, because he was not putting the numbers in a
context.

When Zeke later solved the roads problem he again used diagrams to
build up a systematic and organized way to find all the routes. He reco
nized the relationship to the earlier diagonals of the polygon problem aft
drawing only two diagrams for a smaller number of cities. He wrote:

I started by tring [sic] to visualize four cities so I made a model of it. Then I tr
to make a model of the 10 city kingdom and tried to connect the roads. When
made the roads I noticed that it was like the vertices [diagonals on the polygo
problem. I noticed each time was one less road th[a]n the time before. I count
45 roads for the 10 city kingdom because 9 + 8 + 7 + 6 + 5 + 4 + 3 + 2 + 1 = 45
Then I noticed there was a pattern. The formula is (N - 1) + (N - 2) + (N
3) + (N - 4)... - - - + (N - N) = answer. I next tried to find a[n] easier fo
mula. I tried a lot of things but they did not work. Then I tried distrubiting
[sic] with the original four-city kingdom and got N . (N - 1) - (N . 2). Th
I checked it on a five-city kingdom and got the same answers as ... using the
one.

Unfortunately, Zeke did not check his new generalization o


problem or he would have found it did not coincide with his
his original pattern of adding consecutive decreasing quanti
he clearly added on (built up) to find his sums of numbers g

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
84 DIANA F. STEELE AND DEBRA I. JOHANNING

he did not connect the numeric patterns with the context of how
drawn the diagonals for the four and ten-city problem even thou
recognized the relationship to the previous problems when drawin
diagonals.
Zeke recognized the need to find a closed form for a symbolic gen-
eralization. Perhaps this need came from cooperative group discussion
when other students constructed unclosed symbolic generalizations, i.e.,
(n - 1) + (n - 2) .. .). While observing students in groups, we sometimes
heard statements such as, "What if you had 1000 cities? It would take for-
ever." Because of these group discussions several students began to see the
need for simpler expressions (expressions in closed form) to express their
generalizations and began looking for "shortcuts" and evaluating them to
make sure they were equivalent.
In other problems similar in mathematical structure to the diagonals of
the polygon problem and the roads problem, Zeke did not see the structural
relationship because the diagrams looked differently from his diagrams to
connect vertices. Zeke also had not seen relationships between the earlier
shaded squares, triangle dot, or squares/vertices problems, even though he
had used a building-up schema similar to Mike's to solve those problems.
Because Zeke did not or could not relate how the numbers were generated,
as Mike and Cathy did by using how they drew their diagrams, he struggled
to find a closed-form generalization. Instead, he manipulated numbers to
find numeric patterns he could generalize. This lack of understanding did
not allow Zeke to apply his previous generalizations. Zeke often stayed
in the particulars. Because his schema for generalizing problem situations
was still only partially formed, he was not able to clearly articulate how
to get from the general case from previous problems to find solutions for
new related problems.

9. DISCUSSION OF FINDINGS

Piaget (1952) believed that for a schema to develop students need repetition,
recognition, and generalization. We adapted and progressively sequenced
problems that had similar, but slightly different mathematical structure,
to address these three components. Repetition helped students develop
schemas for solving the problems. Oftentimes, students were able to easily
recognize the new problem situation and assimilate their existing schemas
to solve them. However, many times they needed to accommodate their
existing schemas. Through the process of assimilation and accommoda-
tion, all students successfully generalized verbally, and students with well-
connected schemas also successfully generalized symbolically.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 85

TABLE VI

Students' problem-solving schemas

Well-connected schema Partially-formed schema

Building Subtracting Trial


up out and error Building up Subtracting out Trial and error

Cathy X
Mike X

Zeke X
Jane X

Steve X
Rebecca X

Tommy X
Josh X

9.1. Types of schema development

All of the students in the study developed similar


proaches (i.e., they drew diagrams, made tables, a
problems); however some used each or all of these
ten than others. For the most part, the approaches th
solve problems became their preferred ones, but so
changed over time as a result of experience and con
selves during cooperative groups. They expanded t
modate their new approaches. Although the appro
actual schemas students developed were different
of schemas).
Cathy found shared components and used a subtr
she first developed from solving the shaded squar
students predominantly used this type of schema
Mike and Zeke, used a building-up schema. They l
as pieces of a puzzle that needed to fit together, a
these pieces or built up to generalize. The eighth st
develop a particular schema; his approach to solv
of trial and error. He drew diagrams, but did not
generalize. He most often created tables to begin
numbers in the tables looking for numeric pattern
among problems, but he was not as successful gen

9.2. Qualities of schema development

One recurring theme concerned the connectedness


dents' schemas. Of the eight students in the tea

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
86 DIANA E STEELE AND DEBRA I. JOHANNING

developed well-connected schemas; three developed partially formed


(again see Table VI for quality of schema development). The student
well-connected schemas were more often successful in writing their
bolic generalizations in closed form, (i.e., 4n - 4). They were str
in their ability to generalize from their diagrams. The students who
successful in using both verbal and symbolic generalizations seemed t
tables more as a method of seeing how the diagrams worked. When th
relationships between problems students applied their past generaliza
to problems without first working several particular cases. They se
to achieve the conceptual knowledge that allowed for transfer of sch
knowledge when solving problems that had identifiable features in
mon (Greeno et al., 1996) and generalized their schemas (Dubinsky,
Greeno et al. (1996) said that transfer is a process of acquiring a sch
that relates across situations. These students' well-formed schemas also
related to the vertically constructed schemas discussed by Sfard (1991)-
schemas that are developed by adding new layers forming a deeper and
narrowly connected hierarchy of knowledge capable of assimilating new
information. As shown with the examples of Cathy and Mike, students
with well-connected schemas saw the need to check their generalizations
for specific cases to "prove" them. They checked for several cases usu-
ally by drawing diagrams and counting relevant parts. This need to check
specific cases to prove the general rule was in contrast to the findings of re-
searchers using similar generalizing problems (Cooper and Sakane, 1986;
Balacheff, 1988; Lee and Wheeler, 1987; Stacey, 1989).
Students whose schemas were partially formed could not consistently
or clearly articulate the generalizations and had more recursive unclosed
forms of symbolic generalizations, e.g., (n - 1) + (n - 2) + - -... . This type
of schema also coincided with findings of Sfard (1991). She characterized
this type of schema as one that is unstructured and inadequate for building
meaningful understanding. Students with partially formed schemas did not
see many relationships among problems and did not appear to achieve
much knowledge transfer. They also had difficulty identifying patterns
across particular cases, and therefore, could not raise their way of thinking
to no longer focus on the cases themselves but rather on the relationships
in structure within or among the problems. Like the students in Cooper and
Sakane, Balacheff, Lee and Wheeler, and Stacey studies these students did
not consistently check their generalizations by solving for specific cases.

9.3. Role of diagrams

All students used diagrams to recognize individual parts for solving par-
ticular problems. Diagrams became powerful tools to build successful

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 87

problem-solving schemas. Some students used diag


recognizing the similarities among problems and iden
problems they had stored in their schemas. Students
soning on the physical structure-the diagrams the
only reasoning about numeric patterns were better abl
lationship between quantities in the problem and repr
with symbolic algebraic generalizations. The three stu
formed schemas did not rely as heavily on diagram
terns as did the other five, and often abandoned the
recursive numeric patterns in their tables. Students w
to find the patterns were often unable to interpret the r
the quantities in the problems and were unable to ge
for n.

9.4. Symbolic notation

Students derived their own mathematical expressions from known quan-


tities for which they had solved. They knew that the symbols referred to
their problem quantities. Using their natural language to discuss the rela-
tionship between the quantities was a critical and necessary step in their
development of symbolic representation. As students solved problems in
the teaching experiment, they improved in their ability to express rela-
tionships between quantities and then to use symbolic notation rather than
natural language. This type of symbol sense (Arcavi, 1994) is the essence
of algebraic thinking. Arcavi characterized symbol sense as an individual's
ability to "understand how and when symbols can and should be used to
display relationship and generalizations ..." (p. 31).

10. CONCLUSIONS AND IMPLICATIONS FOR TEACHING AND LEARNING

10.1. Instructional design

In answering Mason and Pimm's (1984) question, "How can you expose
the genericity of an example to someone who sees only its specificity"? (p.
149), findings from this study suggest that you can do this by designing
instruction that gives students maximum opportunities to see the general
in the particular and to see the particular in the general. If teachers are
intentional about instructional design for algebraic thinking (the essence of
which is generalization) they will help students develop effective problem-
solving schemas. To do this teachers can design instruction around the use
of schema theory. The contexts of size and shape and growth and change
promote the use of diagrams and help students develop effective schemas

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
88 DIANA F. STEELE AND DEBRA I. JOHANNING

for algebraic thinking. Also, by ordering the structure of the probl


progressive development sequence, students identify relationships
problems and are able to symbolize relationships by connecting the
eralizations to the physical action of drawing the diagrams. There
teachers should create problem contexts that encourage students t
and analyze diagrams to develop approaches to solving problems th
mote algebraic thinking.

10.2. Generalization and schema development

Generalization is fundamental to schema development. Results fro


study indicate that there is a link between the type of generalizati
dents construct and the schemas they are forming. Students exten
schemas to assimilate the new particular case into an existing
schema or accommodate their existing general schema to incorpor
new particular case. Students often appear to generalize simultane
both directions, from particular to general and from general to pa
When students develop schemas that work for them and explain ho
schemas work in new situations, they learn to generalize. By findin
ulating, and generalizing patterns, students think algebraically. Stud
this study thought algebraically as they represented and related qua
relationships. This finding can be of practical use because it may fa
the progressive organization of learning activities to prepare stude
the formal study of algebra.

ACKNOWLEDGEMENTS

We thank Esther H. Billings for her insights while reading early drafts of
this paper. Her ideas helped us focus on key details in the data. We also
thank the anonymous referee who read our paper several times during the
review process and gave us valuable feedback.

REFERENCES

Algebra Working Group: 1997, 'A Framework for Constructing a Vision of Algebra: A
Discussion Document', Paper Prepared by National Council of Teachers of Mathematics.
Arcavi, A.: 1994, 'Symbol sense: Informal sense-making in formal mathematic', For the
Learning of Mathematics, 14, 24-35.
Balacheff, N.: 1988, 'Aspects of proof in pupils' practice of school mathematics', in D.
Pimm (ed.), Mathematics, Teachers and Children, Open University, Milton, Keynes, pp.
216-235.
Bell, A.: 1976, 'A study of pupils' proof-explanations in mathematical situations', Educa-
tional Studies in Mathematics, 7, 23-40.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
PROBLEM-SOLVING SCHEMAS 89

Cobb, P.: 2000, 'Conducting teaching experiments in collaborat


Kelley and R.A. Lesh (eds.), Handbook of Research Design in
Education, Erlbaum, Mahwah, pp. 307-334.
Confrey, J.: 1997, 'The Nature and Role ofAlgebra in the K- 14',
National Council of Teachers of Mathematics National Sympo
Cooper, M. and Sakane, H.: 1986, 'Comparative Experimental St
gies with Deriving a Mathematical Law',Proceedings of the Tent
ence for the Psychology of Mathematics Education,. University
Education, London, pp. 410-414.
Driscoll, M.: 1999, 'Fostering Algebraic Thinking: A Guide f
Heinemann, Portsmouth.
Dubinsky, E.: 1991, 'Reflective abstraction in advanced mathem
(ed.), Advanced Mathematical Thinking, Kluwer Academic Pu
95-123.
Greeno, J.G., Collins, A.M. and Resnick, L.: 1996, 'Cognition and learning', in D.C.
Berliner and R. Calfee (eds.), Handbook of Educational Psychology, Macmillan, New
York, pp. 14-46.
Kieran, C.: 1996, 'The Changing Face of School Algebra', Invited Lecture for ICME-8
Congress in Spain.
Krutetskii, V.A.: 1976, The Psychology of Mathematical Abilities in School Children, Uni-
versity of Chicago Press, Chicago.
Lee, L. and Wheeler, D.: 1987, 'Algebraic Thinking in High School Students: Their Con-
ceptions of Generalisation and Justification', Research Report, Concordia University,
Montreal, Canada.
Lesh, R.A. and Kelley, A.E.: 2000, 'Multi-tiered teaching experiments', in A.E. Kelley and
R.A. Lesh (eds.), Handbook of Research Design in Mathematics and Science Education,
Erlbaum, Mahwah, pp. 197-230.
Marshall, S.P.: 1995, Schemas in Problem Solving, Cambridge University Press, Cambridge.
Mason, J.: 1996, 'Expressing generality and roots of algebra', in N. Bednarz, C. Kieran and
L. Lee (eds.), Approaches to Algebra: Perspectives for Research and Teaching, Kluwer
Academic Publishers, Dordrecht, pp. 65-86.
Mason, J.A., Graham, A., Pimm, D. and Gower, N.: 1985, 'Routes to/Roots of Algebra',
Open University, Milton Keynes, UK.
Mason, J. and Pimm, D.: 1984, 'Generic examples: Seeing the general in the particular',
Educational Studies in Mathematics, 15, 277-289.
National Council of Teachers of Mathematics: 1997, 'Mathematics Teaching in the Middle
School', 2, p. 227, 244, 271.
National Council of Teachers of Mathematics: 1998, Mathematics Teaching in the Middle
School, 4.
Piaget, J.: 1952, 'The Origins oflntelligence in Children', International Universities Press,
New York.

Piaget. J.: 1971, 'Genetic Epistemology', W. W. Norton, New York.


Piaget, J.: 1977, 'Psychology and Epistemology: Toward a Theory of Knowledge', Penguin
Books, New York.
Piaget, J.: 1985, 'The Equilibrium of Cognitive Structures', Harvard University Press,
Cambridge.
RAND Mathematics Study Panel: 2001, 'What are the three areas of focus proposed by the
mathematics study panel?', Association of Mathematics Teacher Educators Newsletter,
June.

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms
90 DIANA F. STEELE AND DEBRA I. JOHANNING

Sfard, A.: 1991, 'On the dual nature of mathematical conceptions: Reflections on p
and objects as different sides of the same coin', Educational Studies in Mathema
1-36.
Sfard, A.: 2003, 'Balancing the unbalanceable: The NCTM standards in light of theories of
learning mathematics', in J. Kilpatrick, W.G. Martin and D. Schifter (eds.), A Research
Companion to Principles and Standards for School Mathematics, National Council of
Teachers of Mathematics, Reston, pp. 353-392.
Silver, E.A.: 198 1, 'Recall of mathematical problem information: Solving related problems',
Journal for Research in Mathematics Education, 12, 54-64.
Spradley, J.P.: 1980, Participant Observation, Holt, Rinehart, and Winston, Chicago.
Stacey, K.: 1989, 'Finding and using patterns in linear generalising problems', Education
Studies in Mathematics, 20, 147-164.
Steffe, L.P. and Thompson, P.W.: 2000, 'Teaching experiment methodology: Underlying
principles and essential elements', in A.E. Kelly and R.A. Lesh (eds.), Handbook of
Research Design in Mathematics and Science Education, Erlbaum, Mahwah, pp. 267-
306.
Steele, D.F.: 2000, 'Pre-algebra Students' Solution Strategies for Algebraic Problems',
Paper Presented at Annual Meeting of the Research Council on Mathematics Learning,
Las Vegas, NV.
Swafford, J.O. and Langrall, C.W.: 2000, 'Grade 6 students' preinstructional use of equa-
tions to describe and represent problem situations', Journal for Research in Mathematics
Education, 31, 89-112.
Tall, D.: 1991, 'The psychology of advanced mathematical thinking', in D. Tall (ed.),
Advanced Mathematical Thinking, Kluwer Academic Publishers, Dordrecht, pp. 3-21.
Thompson, P.W.: 1994, 'The development of the concept of speed and its relationship to
concepts of rate', in G. Harel and J. Confrey (eds.), The Development Of Multiplicative
Reasoning in the Learning of Mathematics, State University of New York Press, Albany,
pp. 181-236.
von Glasersfeld, E.: 1995, 'Radical Constructivism: A Way of Knowing and Learning',
Falmer Press, Washington DC.

DIANA F. STEELE

Department of Mathematical Sciences, Watson Hall 357


Northern Illinois University, DeKalb, Illinois 60115, U.S.A.
Fax: 815-753-6755; 815-753-1112
E-mail: dsteele @math.niu.edu

DEBRA I. JOHANNING

Michigan State University A-721 Wells Hall, East Lansing


Michigan 1 48824,
E-mail: johanni3 @pilot.msu.edu

This content downloaded from 148.225.8.2 on Mon, 27 Nov 2017 21:07:07 UTC
All use subject to http://about.jstor.org/terms

También podría gustarte