
































 Índice

Aims i

Methodologies for investigating relationships between concept development and 1

the development of problem solving abilities
Richard Lesh, Indiana University; Lyn English, Queensland University of Technology;
Thomas Fennewald, Indiana University

A Technology-Based Investigation of United States High School Student 16

Mathematical Problem Solving
Pamela L. Paek, Charles A. Dana Center, The University of Texas at Austin, Austin,
Texas, United States

Formulating mathematical conjectures in learning activities, assisted with 26

technology
Fernando Barrera Mora and Aarón Reyes Rodríguez, Universidad Autónoma del
Estado de Hidalgo

Problem Posing Performance of Grade 9 Students in Singapore on an open- 36

ended stimulus
Chua, Puay Huat and Yeap, Ban Har, National Institute of Education, Nanyang
Technological University, Singapore

Future directions and perspectives for problem solving research and curriculum 46
development
Lyn English, Queensland University of Technology; Richard Lesh, Indiana University
and Thomas Fennewald, Indiana University

Teaching mathematics in the classroom through problem solving 59

Norma S. G. Allevato, Universidade Cruzeiro do Sul – UNICSUL/ São Paulo/Brazil;
Lourdes R. Onuchic, Universidade Estadual Paulista Julio de Mesquita Filho –
UNESP/Rio Claro/Brazil

The method of problem solving based on the Japanese and Polya´s models. A 71
classroom experience in Chilean schools
Aravena D. Maria and Caamaño E. Carlos, Mathematics Department, Basic Sciences
Institute, Catholic University of Talca – Chile

An ICT environment to assess and support students’ mathematical problem – 81

solving performance in non-routine puzzle-like word problems
Angeliki Kolovou, FIsme, Utrecht University, the Netherlands; Marja van den Heuvel-
Panhuizen, FIsme, Utrecht University, the Netherlands and IQB, Humboldt University
Berlin, Germany; Arthur Bakker, FIsme, Utrecht University, the Netherlands; Iliada
Elia, Department of Education, University of Cyprus
The computer tool for verification hypotheses in parametrical problems solving 91
D. Mantserov, D. Petrichenko, S. Pozdnyakov

The decision-making as a school activity 100

María Candelaria Espinel Febles, University of La Laguna; Ana Teresa Antequera
Guerra, IES Luis Cobiella Cuevas

Cognitive and metacognitive processes of pre-service mathematics teachers 109

while solving mathematical problems
Omar Hernández Rodríguez, Ed.D. and Wanda Villafañe Cepeda, Ed.D. University of
Puerto Rico

Teachers’ beliefs about mathematical problem solving, their problem solving 122
competence and the impact on instruction: A case study of three Cypriot primary
teachers
Constantinos Xenofontos and Paul Andrews, University of Cambridge

Strategies for Solving Word Problems on Speed: A Comparative Study between 132
Chinese and Singapore Students
Jiang Chunlian, Department of Mathematics and Statistics, Central China Normal
University, Wuhan, 430079, People’s Republic of China

Beyond Show and Tell: Neriage for Teaching through Problem-Solving - Ideas 145
from Japanese Problem-Solving Approaches for Teaching Mathematics
Akihiko Takahashi, Ph.D.
DePaul University

Instructional Practices to Facilitate Prospective Mathematics Teachers’ Learning 158

of Problem Solving for Teaching
Olive Chapman, University of Calgary
 TSG-19: Research and Development in Problem Solving in
Mathematics Education
ICME-11, JULY 6-13, 2008, Monterrey, Mexico

Aims
This Topic Study Group aims to provide a forum for those who are interested in aspects
of problem solving research and development at any educational level, to share recent
findings or to exchange ideas. It will also provide an opportunity for the general
participants to become acquainted with the progress and current issues of the field, as
well as its foreseen future directions. A further goal of the organizing team is that the
meetings of this TSG should promote communication and instigate collaborations among
the participants. That is, the themes and ideas addressed during the development of the
sessions will provide the basis to structure a proposal to write a book on the field.

Themes
Problem solving is the heart of mathematics. The teaching and learning of problem
solving has a long history in mathematics education. Problem solving is an activity, which
provides students with opportunities to construct and experience the power of
mathematics. It is also an instructional approach, which provides a consistent context
for students and teachers to learn and apply mathematics.

The primary focus of this TSG is to identify and discuss the current status of research
and development in problem solving in mathematics education around the world. More
specifically the following areas will be explored:

(1) !"#$%&'()*+%&#*,'#-"./0'1#/("-'))')#2%3"03'&#2%#.+*,'.+*2-+0#/("40'.#)"032%5;

(2) To explore the process in which students learn and make sense of mathematics via
problem solving activities, and how can the teacher facilitate this process;
(3) To discuss ways to evaluate problem solving competencies;
(4) To discuss the role of using computational tools in problem solving approaches;

i
(5) To identify and discuss future directions of problem-solving research and
development.

The program includes round-table discussions, individual presentations, small group

presentations and plenary discussions during the development of the sessions. Thus,
the sessions are based on discussing the contributions approved by the organizing
committee and address issues related to:

(a) Foundations of problem solving. Here, we are interested in discussing themes

around the principles or tenets that are important in problem solving activities. Questions
that can guide the discussion include: What are the main principles or tenets that
distinguish a problem solving approach in research and practice? What does it mean to
learn mathematics in terms of problem solving activities? What is the role of routine or
nonroutine tasks in problem solving approaches? To what extent the practice of
development mathematical knowledge is consistent with problem solving approaches?
How perspectives like models and modelling or those that emphasize models in general
relate to problem solving? To what extent the principles and tenets associated with
problem solving have evolved in accordance with the development of computational
tools?

(b) Studies in students' behavior during mathematical problem solving. These will
mostly concern with cognitive, metacognitive, social, and affective aspects of problem
solving; accessing knowledge effectively and the interaction in collaborative work are of
particular interest. Some relevant questions around these themes involve: How
students’ problem solving approaches can be characterized? How have problem-solving
approaches evolved in terms of research questions and methods? What are the current
trends? What theoretical frameworks have been developed in mathematical problem
solving? What is a suitable methodology for studying problem solving processes? How
should students’ problem solving competencies be evaluated?

(c) Instructional approaches: Learning and teaching in problem solving. What

exactly does the student learns from problem solving experiences? Do we wish to be

ii
able to teach mathematical facts or even theories through problem solving? Are present
teaching practices effective for all purposes? How to tackle some practical problems, e.g.
training teachers, extra time that the problem solving seems to involve.
Questions related to the issues include; what makes a task a "good" problem? How can
a problem be used for teaching mathematical topics? To what extent should we expect
students to pose and solve their own problems? What types of assessment are
consistent with problem solving approaches? To what extent, international assessments
like PISA or TIMSS actually evaluate problem-solving competencies?

(d) Research and development in problem-solving with ICT technology. The

development and availability of computational tools have influenced not only the
development of the discipline; but also the way students construct their mathematical
knowledge. In this context, we are interested in discussing questions that involve: How
can we effectively use technologies (e.g., internet, calculators, computers etc) to
facilitate problem solving? How can we effectively use technology to advance
problem-solving research? Do computer representations take away some of the initiative
from the solver? What types of mathematical reasoning, including mathematical
arguments, do students or problem solvers develop as a result of using various
computational tools? What types of strategies and problem representations become
important in problem solving environments that promote the use of computational tools?

(e) Curriculum proposals and problem solving. Some curriculum frameworks (NCTM,
2000) recognize the relevance of problem solving activities. However, there is still a need
to discuss ways in which fundamental tenets associated with problem solving need to be
organized to support a particular curriculum. How should a curriculum proposal, that
enhances a problem solving approach, be organized or structured? What fundamental
mathematical ideas and processes should be central in a proposal that promotes
problem-solving approaches?

iii
 METHODOLOGIES FOR INVESTIGATING RELATIONSHIPS BETWEEN CONCEPT DEVELOPMENT
AND THE DEVELOPMENT OF PROBLEM SOLVING ABILITIES

Richard Lesh Lyn English Thomas Fennewald

Indiana University Queensland University of Technology Indiana University
ralesh@indiana.edu l.english@qut.edu.au tfennewa@indiana.edu
This paper is the second in a pair that Lesh, English, and Fennewald will be presenting at ICME
TSG 19 on Problem Solving in Mathematics Education. The first paper describes three
shortcomings of past research on mathematical problem solving. The first shortcoming can be
seen in the fact that knowledge has not accumulated – in fact it has atrophied significantly during
the past decade. Unsuccessful theories continue to be recycled and embellished. One reason for
this is that researchers generally have failed to develop research tools needed to reliably observe,
document, and assess the development of concepts and abilities that they claim to be important.
The second shortcoming is that existing theories and research have failed to make it clear how
concept development (or the development of basic skills) is related to the development of
problem solving abilities – especially when attention is shifted beyond word problems found in
school to the kind of problems found outside of school, where the requisite skills and even the
questions to be asked might not be known in advance. The third shortcoming has to do with
inherent weaknesses in observational studies and teaching experiments – and the assumption that
a single grand theory should be able to describe all of the conceptual systems, instructional
systems, and assessment systems that strongly molded and shaped by the same theoretical
perspectives that are being used to develop them. Therefore, this paper will describe theoretical
perspectives and methodological tools that are proving to be effective to combat the preceding
kinds or shortcomings. We refer to our theoretical framework as models & modeling
perspectives (MMP) on problem solving (Lesh & Doerr, 2003), learning, and teaching. One of
the main methodologies of MMP is called multi-tier design studies (MTD).

Three Shortcomings of Past Problem Solving Research

1. Relationships are Unclear between Concept Development and the Development of Problem
Solving Competencies. One shortcoming of past problem solving research is that it has not been
clear how concept development is expected to interact with the development of relevant problem
solving heuristics, beliefs, dispositions, or processes. In fact, in most curriculum standards
documents (NCTM, 2000; http://www.doe.state.in.us/standards/), problem solving tends to be
listed as the name of a chapter-like topic similar to algebra, geometry, or calculus. In other
words, the implicit assumption is conveyed that problem solving ability is expected to increase
by: (a) first, mastering relevant concepts, (b) second, mastering relevant problem solving
heuristics, strategies, beliefs, dispositions, or processes, and (c) third, learning to put these
concepts and processes together to solve problems. Consequently, when such assumptions are
coupled with the flawed belief that students must first learn concepts and processes as
abstractions before they can put them together and use them in “real life” problem solving
situations, problem solving tends to end up never getting taught at all in many classrooms. So,
one of the most critical challenges for future problem solving research is to clarify the nature of
relationships that should exist between concept development and the development of problem
solving competencies.

1
2. Relationships are Unclear between Competencies on Textbook Word Problems and the Levels
and Types of Understandings needed to use Mathematics Concepts and Abilities Beyond School.
A second shortcoming of past problem solving research is that, because of its almost exclusive
emphasis on textbook word problems, mathematics educators have given relatively little
attention to the kinds of problem solving competencies that are needed when mathematical
thinking is required outside of school. The main exception to this rule has been in the area of
elementary arithmetic concepts – where the emphasis has been on skill development rather than
problem solving. This neglect of problem solving beyond school is especially significant
because, in research about problem solving in fields (such as engineering) that are heavy users of
mathematics and technology (Lesh, Hamilton & Kaput, 2007; Zawojewski, Diefes-dux &
Bowman, 2008), it has become clear that, in a technology-based age of information, problem
solving outside of school tends to be significantly different than problem solving in the context
of word problems of the type emphasized in school textbooks and tests (Lesh & Caylor, 2008).
Therefore, even if past research on mathematical problem solving would have been successful at
explaining students’ behaviors in the context of traditional kinds of textbook word problems, it is
unlikely that such explanations would be useful without modification in the kinds of future-
oriented problem solving situations that are emerging beyond school. Of specific relevance is
that, in the world beyond school, specific questions being answered might not be known until
long after the problem (dilemma, decision, discomfort) begins to be addressed. (Knorr-Cetina &
Mulkay, 1982; Latour, 1987; Sawyer, 2006)

3. Research on Mathematical Problem Solving has not Accumulated. Failed or flawed concepts
or conjectures have continued to be recycled or embellished – with no significant changes being
made in the underlying theoretical perspectives. This is exacerbated by the fact that mathematics
education researchers have generally avoided tasks that involve developing critical tools for their
own use. Unlike their counterparts in more mature sciences (physics, chemistry, biology), where
some of the most significant kinds of research often involve the development of tools to reliably
observe, document, or measure the most important constructs that are hypothesized to be
important, mathematics educators have developed very few tools for observing, documenting, or
measuring most of the understandings and abilities that are believed to contribute to problem
solving expertise. Furthermore, partly because operational definitions and tools have not been
developed to observe, document, and assess the development of most constructs that have been
claimed to be important, apparently failed or flawed concepts tend to be continually embellished
or recycled. For example, the first paper in this pair describes the continuous embellishment of
a theory that focuses on explicitly-learned rules. It also describes instances where “new”
theoretical constructs clearly consist of nothing more than new names for previously discredited
constructs. The basic problem has been that short lists of rules (e.g., heuristics, beliefs,
metacognitive processes) tend to have descriptive but not prescriptive power; longer lists of
prescriptive rules become so long that knowing when to use them becomes as important as
knowing how to use them. Yet, introducing meta-rules (i.e., higher-level rules that operate on
lower order rules) simply transfers the same basic shortcoming to a higher level.
Using new names to recycle old constructs is not the same as producing durable constructs, and
continuous embellishment is not the same as theory development. As Popper (1963) emphasized,
one of the most important characteristics that distinguishes a scientific theory from an ideology
rests on the potential falsifiability of its assumptions and claims. In fact, according to Popper,
one of the most important ways that theories develop is by rejecting hypotheses. Yet, rejected

2
hypotheses rarely occur in the kind of research that has dominated inquiry about mathematical
problem solving. So, for mathematics educators who are interested in research on problem
solving, what we believe to be most needed are: (a) tools for observing, documenting, and
measuring important constructs, (b) theoretical perspectives which do not encourage orthodoxy,
and (c) research methodologies which encourage the consideration of diverse ways of thinking –
but which also encourage selection among alternatives.
Shortcomings Associated with Theories & Research Methodologies
1. Concerning the development of useful theories. The theoretical perspective that we will
emphasize in this paper traces its roots primarily to Piaget (Piaget and Beth, 1966) and to
American Pragmatists such as Pierce, James, Holmes, Meade, and Dewey (Lesh and Doerr,
2003) who were originators of many of the most important ideas underlying modern views of
situated cognition and socio-cultural factors that mold and shape knowledge development. Like
pragmatism, the theoretical perspective that we emphasize is not so much a theory as it is a
framework for developing theories. Like the pragmatists, we believe that no single theory is
likely to provide solutions to the complex kind of problems that mathematics educators most
need to understand and solve. One reason why this is true is because mathematics education
tends to be more like engineering than physics – in the sense that the systems that we need to
understand are largely products of human design or human guided development. So, the same
conceptual systems that are used to make sense of these products of human development are also
used to change them.
How does this challenge the idea that there can be a grand theory of education? Consider the
fact that, in fields such as aerospace engineering, we will never have a fixed and final theory of
things like space shuttles. One reason why this is true is because as soon as we understand them
better, we will change them. So, every conceptual system that is developed for thinking about
them is really the nth in a continuing series. Furthermore, the design of such engineered artifacts
usually involves trade-offs among competing perspectives and interests - because we usually
want products that are low in cost but high in quality, or low in risk but high in possible gain, and
so on. In fields like engineering, such observations have prompted the well known quip that:
Engineering is the science of understanding and designing things when there is not enough time,
money, or other resources - and when trade-offs need to be considered which involve conflicting
conceptions of success. Education is much like this – and we will not be able to reach a final
grand theory because of it.
In general, in research based on MMP, we adopt the pragmatists’ point of view that what
research on problem solving most needs is not another grand theory which claims to explain
everything from cooking, to carpentry, to students’ behaviors on textbook word problems
(Schoenfeld, 2007). So, we do not expect realistic solutions to realistically complex problems to
be solved by single research studies, nor even by single theories. Instead, what are most needed
are models which are embodied in artifacts and tools that are designed to be powerful, sharable
and reusable. Such models also need to integrate ways of thinking drawn from a variety of
practical and theoretical perspectives. Once this is done, model development can lead to theory
development; no single theory should be expected to provide guidance for most important
problems or decision-making issues related to mathematical problem solving, rather models and
theories should guide decision-making.

3
2. Concerning inadequacies of observational studies and teaching experiments. Two research
methodologies which have dominated research of mathematical problem solving are: (a)
observational studies in which researchers observe (and often analyze videotapes of) students
solving problems, and (b) teaching experiments in which the researcher attempts to validate the
importance of some heuristic, belief, disposition, or process by demonstrating that it can be
taught.
One shortcoming of observational studies tends to be that constructs that are useful for
describing past problem solving behaviors are not necessarily useful for prescribing future
problem solving behaviors. A second shortcoming of observational studies is that the things that
observers see are always strongly influenced by the researchers’ preconceived notions. This
shortcoming is especially powerful during a time when “heavy users” of mathematics claim that
beyond school, significant changes have been occurring in the nature of situations where some
type of mathematical thinking is needed beyond school.
One shortcoming of teaching experiments is that they simply have not been successful. Small
treatments have produced small effects, and large treatments either do not get implemented
sufficiently or their complexity tends to makes it impossible to draw causal inferences. So, in
either case, such studies usually end up concluding that the researcher didn’t try hard enough –
or in the right way. But, beyond these practical considerations, studies designed to show that “it”
works confront the same problems that led to the demise of aptitude-treatment interaction
theories (ATIs). ATIs assumed that if students can be classified as having one set of profiles or
traits as variables, then it should be possible to match the students with suitable treatments, so
that particular learning goals will be reliably achieved. However, as Cronbach and Snow (1977)
showed, ATI treatments were not effective assessments of student learning because complex
interactions amongst the variables often mattered more than the variables themselves; to test all
of the combinatorial combinations of the variables was not only computationally impractical task
but also theoretically impossible.
Not only did such “treatments” prove to be combinatorically impractical, but they also proved to
be theoretically impossible because of feedback loops, second-order effects, and other emergent
properties associated with complex systems which have chaotic (unpredictable) outcomes. In
fact, many of the most powerful actions that determine success tend to involve two-way
interactions (rather than one-way actions) among students, teachers, and other relevant agents or
resources (e.g, parents, programs, policy makers, curriculum materials, assessment systems,
etc.).
3. Concerning the usefulness of design research methodologies. A recent Handbook of Design
Research in SMET Education (Kelly, Lesh & Baek, 2008) describe a powerful new class of
research methodologies that should be especially useful in research on mathematical problem
solving. Although design research methodologies are new to learning scientists and mathematics
educators, they have been used for years in many design sciences such as engineering
(Zawojewski, Diefes-dux, & Bowman, 2008) – where the “thing” that needs to be understood
and explained are also being designed by the relevant researchers.
Learning scientists tend to believe that design research methodologies were developed by
theoreticians who were attempting to make results of their lab-based research more practical
(Brown, 1992) or by software developers or educational program developers who were
attempting to provide better theoretical grounding for best practices (Collins, 1992). But, in

4
mathematics education, and in particular in our own research investigating the nature of students’
developing mathematical concepts and processes, design research methodologies developed
mainly out of attempts to minimize the amount of researcher guidance (Lesh, 2002). This work
was modeled on Soviet-style teaching experiments (Krutetskii, 1976).
One dilemma that influenced our research was that regardless whether we focus on the
development of students’ conceptual systems, or the design of curriculum materials or programs,
the “things” that we want to study are “things” that we ourselves are helping to develop or
design. So, how can hypotheses be tested when they involve phenomena that are continually
changing – and when the changes are partly driven by the conceptual systems that researchers
are developing? How can we be certain that principles which are rejected or accepted today will
not need to be revisited tomorrow?
A second way to describe the preceding dilemma is to recognize that when we investigate the
nature of students’ interpretation abilities, students’ interpretations clearly are influenced by both
their own structuring abilities and also by the structure of the tasks that researchers or teachers
present. So, when researchers examine students interacting with activities designed by
researchers, the interactions and the explanations of the students are not likely to be reducible to
simple input-output rules, because researchers are a part of the system they are studying.
To deal with the preceding kinds of issues, it is useful to notice that mathematics educators are
not alone! Similar dilemmas confront engineers or other design scientists when they are
attempting to understand worlds that they themselves are designing – and worlds filled with
feedback loops and second-order effects (where A impacts B, B impacts C, and C impacts A).
The following observations about the ways that engineers confront these sorts of issues are
useful to consider:
• Regardless whether engineers are designing software or space shuttles, the underlying
designs tend to be important parts of the products that are designed. So, when trial products
are tested, the underlying design principles also are tested.
• The design “specs” that engineers are given enable them to test products and to choose
among strengths and weaknesses associated with alternative designs. This allows engineers
to move in directions that are increasingly better without basing decisions on preconceived
notions of what is “best.”
On the one hand, such procedures cannot overcome the fact that any current model can only be
the nth iteration in a continually evolving series. On the other hand, sequences of such models
often provide auditable trails of documentation which reveal important trends or patterns that
otherwise would not be apparent. These patterns often enable generalizations to be made.
Models & Modeling Perspectives on Mathematics Problem Solving, Learning & Teaching

Foundations of MMP have been described in a number or recent publications (Lesh & Doerr,
2003; Lesh & Lehrer, 2002; Lesh & English, 2005; Lesh & Sriraman, 2005). MMP evolved
primarily out of Piagetian and American Pragmatist perspectives - which also presaged many
modern situated and socio-cultural views of problem solving, learning, and teaching (Lesh &
Doerr, 2003, p. 519-556). Compared with other theoretical perspectives that have been used to
investigate mathematical problem solving: (a) MMP emphasizes interpretation and
communication aspects of understanding as much as it emphasizes procedural capabilities, (b)

5
MMP investigates problem solving processes developmentally – using techniques similar to
those that others have used to investigate what it means to understand the development of
concepts ranging from early number concepts (Steffe et al.,1983; Clements & Bright, 2003;
Fuson, 1992), to rational numbers and proportional reasoning (Lesh, Post & Behr, 1985;
Middleton et al., 2001) to the foundations of algebra (Driscoll et al., 2001), statistics (Konold &
Lehrer, in press), or calculus (Kaput, 1997), and (c) MMP emphasizes the fact that, as we enter
the 21st century, significant changes have been occurring in both the kinds of situations where
some type of mathematical thinking is needed for success beyond school, and the levels and
types of understandings and abilities that are needed for success in these situations (Lesh,
Hamilton & Kaput, 2007). So, even if past theories of problem solving would have proven to be
adequate for describing students’ thinking in the context of traditional textbook word problems,
MMP research entertains the notion that these theories may need to be modified significantly to
describe the kind of mathematical thinking that is needed beyond school in a technology-based
age of information (Lesh, Hamilton & Kaput, 2007).
Unlike most past theories that have been used to investigate mathematical problem solving,
MMP was not developed primarily to explain problem solving per se. Instead, MMP was
designed to investigate the development of mathematics concepts. Nonetheless, MMP also has
used developmental studies to investigate what it means to “understand” problem solving
processes. These studies have shown that concept development and the development of problem
solving processes are closely and synergistically related. For example, one of the earliest
questions that MMP researchers investigated was: What is it, beyond the kind of understandings
emphasized in most textbooks, tests, and classroom teaching, that enables students to use the
things they have learned in real life situations beyond school? (Lesh, Landau, & Hamilton,
1983) Results of these studies made it clear that the following two questions are significantly
different.

• What should students do when they are stuck (i.e., when they are not aware of any relevant
concepts or processes)?

• What additional levels or types of understanding do students need to develop in order to be

able to use concepts and abilities which: (a) are recognized as being relevant to a given
problem solving situation, (b) are at intermediate stages of development, and (c) need to be
adapted significantly to be useful in current circumstances?

MMP investigates questions such as: (a) What does it mean for students to “understand”
relevant heuristics, strategies, beliefs, dispositions, or metacognitive processes? (b) What is the
nature of students’ early understandings of relevant heuristics, strategies, beliefs, dispositions,
or metacognitive processes? (c) How (or in what ways) do students’ early understandings
develop?
In the most general terms, models can be thought of as being systems for describing or designing
other systems. As such, they are conceptual systems or interpretation systems, and because they
are developed for a purpose, they are purposeful conceptual systems. Then, a distinctive

6
characteristic of mathematical and scientific models is that they focus on systemic (or emergent1)
properties of systems-as-a-whole.
One reason why MMP focuses on interpretation abilities is because, outside of school, in
virtually every area where researchers have investigated similarities and differences between
experts and novices (or between gifted versus average ability students, or between successful
versus relatively unsuccessful problem solvers), results have shown that experts not only do
things differently, but they also see (or interpret) things differently.
A second reason why MMP focuses on interpretation abilities is because Piaget-inspired
researchers have shown that the development of most mathematics concepts depend on the
development of students’ abilities to make sense of situations using operational/relational
systems-as-a-whole. That is, relevant concepts do not take on their appropriate mathematical
meanings until students are able to think systemically. Examples of systemic properties include
invariance with respect to a system of operations, transitivity with respect to a system of
relations, or properties that involve minimization, optimization, or stabilization of operational-
relational systems. So, what Piagetians showed is that, if the conceptual systems that students
use to interpret their experiences are not yet functioning as systems-as-a-whole, then students’
thinking tends to be unstable (e.g., they lose the metaphorical “forest” when their attention
focuses on “trees” - or vice versa). Their thinking also tends to be characterized by: (a) centering
– losing cognizance of one attribute when others are noticed, or (b) conceptual egocentrism –
lacking the ability to be self-critical, or to consider alternative ways of thinking.
When we say that modeling is about interpretation and expression, this includes the fact that
modeling is about the description and explanation of existing systems, and it also is about the
design and development of new systems. According to MMP, the development of mathematical
competence is about the development of powerful mathematical models and modeling abilities at
least as much as it is about the acquisition of mathematical facts, skills, or processes. Yet, when
we focus on the mathematics of description and explanation, this does not mean that the
mathematics of computation and derivation are neglected. Such neglect would be as foolish in
mathematics as it would be to ignore basic skills in athletics or performing arts – where equal
attention also is given to scrimmages, competitions, and performance in other complex decision-
making situations where the emphasis is on much more than isolated basic skills – and where
knowing when to do things is as important as knowing how to do them.
Because of the preceding perspectives, MMP defines problem solving activities to be goal-
oriented activities in which problem solvers need to make significant adaptations to their current
ways of thinking in order to achieve the desired goal. Consequently, MMP focuses on problem
solving situations in which model development is an important part of the product that problem
solvers produce – or the underlying design is an important part of the conceptual tools that are

1
According to MMP, mathematics is the study of structure (Lesh and English, 2005). If we look at the undefined
terms that occur in the formal axiomatic systems that are used to define mathematical concept, then every
“undefined term” in these axiomatic systems is an emergent property of the systems. That is, all of its mathematical
meaning comes from the systems that are used to define them. Similarly, underlying every statement of value,
MMP expects there to be a system of values. Underlying every heuristic or metacognitive process, MMP expects
there to be a conceptual system. And, underlying every fact or skill, MMP expects there to be a conceptual system in
which the fact or skill becomes meaningful.

7
designed. Then, when solution development involves conceptual adaptation, at least as much as
information processing, it is misleading to characterize problem solving as getting from givens to
goals when the path is not obvious. In fact, model development tends to involve several express-
test-revise cycles in which significant changes generally need to be made to initial conceptions of
givens, goals, and possible solution processes. So, the development of solutions involves the
adaptation of existing conceptual systems much more than it involves the search for ideas and
procedures which have been misplaced. So, the kinds of heuristics that are most useful are those
that help students’ ways of thinking evolve beyond current conceptions – all of which tend to be
at intermediate stages of development.
MMP-based Design Research Studies & Tools to Support Research Collaborations
According to MMP, researchers, teachers, and students – all are considered to be model
developers. Students develop models in response to problems that are simulations of important
new kinds of situations where important types of mathematical thinking are needed beyond
school in the 21st century. Teachers develop models (and conceptual tools) for making sense of
students’ modeling activities. Researchers develop models of interactions between students and
teachers. At all three levels, model developers express their current ways of thinking in the form
of artifacts or tools which are designed explicitly to be useful for some specifically targeted
purpose. Because the “design specs” make it clear that the underlying design (or conceptual
system) is an important part of the artifact that is designed, when the artifact or tool is tested the
important aspects of the conceptual system that it embodies is also tested. Furthermore, because
the artifacts or tools that are produced need to be powerful (in the specific situations in which
they were created), sharable (with other people), and reusable (in the future, and in other
situations beyond the one in which they were created), they also contribute to community
building and to the accumulation of knowledge. So, at all three interacting levels of model
development, the products that model developers produce are expected to go beyond simply
being tested for usefulness, sharability, and generalizability; they also are designed to have these
attributes.
The preceding perspectives lend themselves to multi-tier design studies (Kelly, Lesh & Baek,
2008) which are aimed at investigating interactions among the model development activities of
students, teachers, and researchers. Furthermore, multi-tier design studies also are specifically
designed to coordinate the work of multiple researchers who are working at multiple sites and
who represent a variety of practical or theoretical perspectives which may range from student
development, to teacher development, to curriculum development, to theory development. At all
levels of multi-tier design studies, many of the most important products that problem solvers
produce are powerful, sharable, and reusable tools for their own use. So again, the result is to
promote community building and the accumulation of knowledge.
MMP Alternatives to Past Problem Solving Research Methodologies
1. Model-Eliciting Activities - Alternatives (or Supplements) to Clinical Interviews or
Videotape Analyses: MMP research now uses model-eliciting activities in many situations where
we once used clinical interviews or videotape analyses – both of which tend to be very labor
intensive and difficult to replicate. (Lesh & Lehrer, 2000). Unlike most of the problem solving
situations that have been used in research on mathematical problem solving, MEA’s were
designed, first and foremost, for research purposes (Lesh & Caylor, 2008). The fact that they also
have proven to be useful to support student learning (Lesh & Doerr, 2003), assessment (Lesh &

8
Lamon, 1992), and teacher development (Schorr & Lesh, 2003) is mostly due to inherent
synergies between our views of mathematics education research and practice. Principles for
designing MEA’s have been explained in several resent publications (Lesh, et. al. 2000;
Hjalmarson & Lesh, 2008).
As their name suggests, MEA’s are activities in which students’ develop a model - or an artifact
or a tool which explicitly embodies an important conceptual system (explanation, interpretation,
design) that the researcher wants to investigate. Therefore, because the underlying conceptual
system is an important part of the designed artifact or tool, testing these products also involves
testing the underlying design principles that they embody. Furthermore, because underlying
conceptual systems are expressed in forms that can be examined and assessed by students,
teachers, and researchers, solutions to MEA’s tend to involve sequences of iterative express-test-
revise cycles similar to the kind that are involved in the first-, second-, and nth-drafts that are
involved in the development of other kinds of written or drawn descriptions of situations.
Therefore, auditable trails of documentation tend to be produced automatically, and important
aspects of the evolving models can be inspected by both students and teachers (or researchers).
These documentation trails often supplement the kind of information that can be obtained with
time-consuming videotape analyses. Furthermore, because MEA’s are designed so that
significant conceptual adaptations occur during relatively brief periods of time (e.g., 60-90
minute problem solving episodes), MEA’s often function something like little Petrie dishes in
science laboratories. That is, important developments occur in easily observable forms during
sufficiently brief periods of time so that researchers can go beyond observing successive states of
knowledge to also observe processes that lead from one state to another. Furthermore, compared
with the kind of information that can be gained from clinical interviews or videotape analyses, it
often is possible to involve far more students using MEA’s – and the results tend to be far more
sharable.
When we compare information that can be gained from MEAs versus clinical interviews, it is
noteworthy that a primary goal of clinical interviews is to follow students thinking – rather than
simply investigating how close students can come to the researchers preconceived notions about
how students should think about important mathematical concepts or processes. But, especially
when the kind of thinking that is being investigated focuses on students’ interpretation abilities,
every interpretation that students produce is influenced by both the structure of the task and the
students’ structuring abilities. Therefore, each time a student develops a new interpretation,
interpretation abilities tend to be impacted. And, if the interpretation involves a powerful
mathematical construct, these impacts tend to be significant. So, in MMP research, we address
this fundamental difficulty by trying to be as explicit as possible about how both students and
teachers (or researchers) structure the tasks at hand – and about how they interact.
Whereas, in clinical interviews, researchers adapt their questioning to the thinking revealed by
individual students, in MEAs, the students are able to interpret a single problem in a variety of
ways and at a variety of levels of sophistication. So, MEAs tend to be self adapting.
2. Local Conceptual Development Studies - Alternatives to Expert-Novice Studies: MMP
research often compares problem-solvers-who-are-isolated-individuals to problem-solvers-who-
are-groups – in somewhat the same way that other researchers from other theoretical
perspectives have compared experts and novices, or gifted problem solvers and average-ability
problem solvers. Using MEA’s, one result of this approach is that it has become clear that
problem solvers’ early interpretations of model-eliciting activities usually involve a collection of

9
partly-overlapping yet undifferentiated partial interpretations of different aspects of the relevant
situations. So, regardless whether the problem solving is an individual or a group, model
development tends to involve gradually sorting out, clarifying, revising, refining, and integrating
the preceding kinds of gradually evolving ways of thinking. Furthermore:
(a) The evolution of these initially-unstable communities of constructs tends to resemble the
evolution of complex and diverse ecological systems – far more than they resemble the
movement of a point along a path (i.e., getting from givens to goals when the path is unclear).
(b) Heuristics that are intended to help problem solvers make productive adaptations to existing
ways of thinking often are significantly different than heuristics that are intended to help problem
solvers figure out what to do when they are stuck (with no apparent concepts available).
(c) Heuristics and metacognitive processes evolve in ways that are often quite similar to the
dimensions of development that apply to other types of concepts or abilities that mathematics
educators have studied. In particular, Vygotsky’s (1978) concept of internalizing external
functions often results in early understandings of heuristics that are distinctly social in character.
So, instead of “looking at a similar problem” it often is useful to “look at the same problem from
another point of view” (and to be aware of the fact that one’s current point of view is not the only
possible point of view.
(d) In the context of MEA’s, heuristics and metacognitive processes generally function tacitly
rather than as explicitly executed rules; in MEAs, heuristics and metacognitive processes have
far less to do with helping students know what to do next, and have far more to do with helping
them interpret the situation (including alternative ways of thinking about givens, goals, personal
competencies, and “where they are” in solution processes). For example, when athletes or
performing artists analyze videotapes of their own performances (or those of others), it is useful
for them to develop a language for describing these past performances. But, this language
usually is not intended to result in prescriptive rules about what to do at some given point in
future performances. Instead, the language and imagery that they develop tends to be aimed
mainly at helping them make sense of things during future performances. In other words, they
are aimed mainly at the development of more powerful models.
3. Multi-Tier Design Research - Alternatives to Japanese Lesson Plan Studies: MMP research
often investigates the development of teacher knowledge by engaging teachers in the
development of tools for facilitating, documenting, analyzing, or assessing the development of
students’ models during MEA’s. For example, one class of teacher-level tools has been referred
to as ways of thinking sheets (Berry, 2006; Carmona, 2004; Hjalmarson, 2004; Lesh, Doerr,
Carmona, & Hjalmarson, 2003). Ways of thinking sheets are sharable and reusable tools that
teachers often find it useful to develop making sense of students’ work – and for recording
alternative ways of thinking that students adopt (a) while they are working on solutions to
MEA’s, (b) when they are giving oral reports of their results for MEA’s, or (c) when they submit
written reports of their results for MEA’s. For teachers, the purposes of these ways of thinking
sheets usually is to help them give students feedback about strengths and weaknesses of their
work, or to help them identify appropriate follow-up activities that harvest students’ insights and
addresses their needs. Or, for teacher developers, the development of ways of thinking sheets
often functions similarly to Japanese Lesson Plan Studies (Driscoll et al., 2001). Teachers often
work in teams of three, and tool development goes through a series of iterative express-test-
revise cycles. However, because research has shown that one of the most powerful ways to

10
positively influence teachers’ teaching practices is to help them become more insightful about
the nature of their students’ thinking (Zawojewski, Diefes-dux & Bowman, 2008), and because
the development of ways of thinking sheets minimize the amount of time that teachers are taken
away from teaching, they lend themselves to effective on-the-job teacher development activities.
• In real classrooms (those that are not ongoing research laboratories), the learning that has
occurred through the use of MEA’s has been more impressive when teacher-development
and student-development go hand-in-hand.
• In a sense, ways of thinking sheets are MEA’s for teachers. Just like MEA’s for students,
they tend to generate auditable trails of documentation about the development of teachers’
knowledge and abilities.
4. Evolving Expert Studies - Alternatives to Ethnographic Observations or Questionnaires:
MMP research often involves evolving expert studies in which researchers, and other relevant
experts, are engaged in the development of models, artifacts or tools in which the underlying
conceptual system is an important part of the product. For example, as discussed in the book,
Foundations for the Future in Mathematics Education (Lesh, Hamilton & Kaput, 2007), in the
book, Models and modeling in Engineering Education: Designing experiences for all students
(Zawojewski, Diefes-dux & Bowman, 2008), and in a series of semester-long follow-up studies
with professional engineers, business managers, and others who are heavy users of mathematics,
we engaged three-person teams of these experts to work with teachers and learning science
researchers to co-design MEA’s which they believed would help clarify insightful and future-
oriented responses to the following questions. What is the nature of typical problem-solving
situations where elementary-but-powerful mathematical constructs and conceptual systems are
needed for success in a technology-based age of information? What kind of “mathematical
thinking” is emphasized in these situations? What does it mean to “understand” the most
important of these ideas and abilities? How do these competencies develop? What can be done
to facilitate development? How can we document and assess the most important (deeper, higher-
order, more powerful) achievements that are needed: (i) for informed citizenship, or (ii) for
successful participation in wide ranges of professions that are becoming increasingly heavy
users of mathematics, science, and technology? How can we identify students who have
exceptional potential which are not measured on standardized tests?
Unlike studies in which researchers observe or interview experts, the preceding studies
recognized that the opinions of researchers, teachers, and other experts would be certain to
evolve significantly if they participated in activities in which they repeatedly expressed their
current ways of thinking in forms that were tested and revised iteratively – based on peer review,
and based on trials in which their draft activities were tried out with students. In the beginning,
these experts’ opinions focused on traditional kinds of skill building, but by the end, there was a
consistent and overwhelming consensus that future-oriented problem solving will involve: (a)
designing and making sense of complex systems, (b) working in teams of diverse specialists each
of whom use continually evolving tools, and (c) participating in multiple-stage projects in which
relevant abilities emphasize the multiple-media communicating, collaborating, planning,
monitoring, and assessing. Also, computational and multi-media modeling often replaces
models that were based on single, solvable, differentiable functions.

References

11
Berry, S. (2006) The evolution of teacher-designed “ways of thinking” sheets for giving feedback
to students. Unpublished doctoral dissertation, Purdue University, West Lafayette, IN.
Brown, A. (1992). Design experiments: Theoretical and methodological challenges in creating
complex interventions in classroom settings. The Journal of the Learning Sciences, 2(2),
141-178.
Carmona, G. (2004). Designing an assessment tool to describe students’ mathematical
knowledge. Unpublished doctoral dissertation, Purdue University, West Lafayette, IN.
Clements, D. & Bright G. (2003), Learning and teaching measurement. Reston, VA: National
Council of Teachers of Mathematics
Collins, A. (1992). Toward a design science of education. In E. Scanlon & T. O’Shea (Eds.),
New directions in educational technology (pp. 15-22). Berlin: Springer Verlag.
Cronbach, L. & Snow, R. (1977). Aptitudes and Instructional Methods: A Handbook for
Research on Interactions. New York: Irvington.
Driscoll, M., J., Zawojewski, J. S., Humez, A., Nikula, J., Goldsmith, L., & Hammerman, J.
(2001). Algebraic Thinking Tool Kit. Portsmouth, NH: Heinemann.
Fuson, K. (1992). Research on whole number addition and subtraction. In D. A. Grouws (Ed.),
Handbook of research on mathematics teaching and learning (pp. 243-275). NewYork:
Macmillan.
Hjalmarson M. & Lesh, R. (2008). Engineering & Design Research. In Kelly, A. E., Lesh, R.
A., & Baek, J. Y. (2008). Handbook of innovative design research in science, technology,
engineering, mathematics (STEM) education. NY: NY. Taylor & Francis.
Hjalmarson, M. A. (2004). Designing presentation tools: A window into mathematics teacher
practice. Unpublished doctoral dissertation, Purdue University.
Kaput, J., Hegedus, S., & Lesh, R. (2006). Technology becoming infrastructural in mathematics
education. In R. Lesh, E. Hamilton & J. Kaput (Eds.), Models & modeling as foundations for
the future in mathematics education. Mahwah, NJ: Lawrence Erlbaum Associates.
Kaput, James J. (1997) Rethinking Calculus: Learning and Thinking, American Mathematical
Monthly, 104(8).

Kelly, A. E., Lesh, R. A., & Baek, J. Y. (2008). Handbook of innovative design research in
science, technology, engineering, mathematics (STEM) education. NY: NY. Taylor &
Francis.
Kilpatrick, J. (1985). A retrospective account of the past twenty-five years of research on
teaching mathematical problem solving. In E. A. Silver (Ed.), Teaching and learning
mathematical problem solving: Multiple research perspectives (pp. 1-16). Hillsdale, NJ:
Lawrence Erlbaum.
Knorr-Cetina, K. and Mulkay, M. (1992). Science in Context. London: Sage.
Konold, C., & Lehrer, R. (in press). Technology and mathematics education: An essay in honor
of Jim Kaput. In L. English (Ed.), Handbook of International Research in Mathematics
Education, (2nd edition). New York: Routledge.

12
Konold, C., Harradine, A. and Kazak, S. (2007) Understanding Distributions by Modeling Them.
International Journal of Computers for Mathematical Learning. 12(3).
Krutetskii, V. (1976). The psychology of mathematical abilities in school children. Chicago:
University of Chicago Press.
Latour, B. (1987). Science in Action: How to Follow Scientists and Engineers through Society.
Cambridge, MA, Harvard University Press.
Lesh R., Hamilton E., & Kaput J. (Eds.), (2007). Models & modeling as foundations for the
future in mathematics education. Mahwah, NJ: Lawrence Erlbaum Associates.

Lesh, R. & Doerr, H. (2003) Beyond Constructivist: A Models & Modeling Perspective on
Mathematics Teaching, Learning, and Problems Solving. In Doerr and Lesh (Eds.) Hillsdale,
NJ: Lawrence Erlbaum Associate
Lesh, R. & Caylor, E. (2007) Modeling as Application VS Modeling as a Way to Create
Mathematics. International Journal of Computers for Mathematical Learning. Springer.
Lesh, R. & Caylor, E. (2008) Differing Conceptions of Problem Solving in Mathematics
Education, In B. Greer & L. Verschaffel Science Education & Professional Schools.
Mathematics World Problems. NY, NY. Springer Publishing.
Lesh, R. & English, L. D. (2005). Trends in the evolution of models & modeling perspectives on
mathematical learning and problem solving. ZDM: The International Journal on
Mathematics Education, 37 (6), 487-489.
Lesh, R. & Lehrer, R. (2002) Philosophical Roots of Models & Modeling Perspectives.
Philosophical Roots of Models & Modeling Perspectives. In R. Lesh, Research Design in
Mathematics Education: Focusing on Design Experiments. In L. English (Ed.) International
Handbook of Research Design in Mathematics Education. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Lesh, R. & Sriraman, B. (2005). John Dewey Revisited- Pragmatism and the models-modeling
perspective on mathematical learning. In A. Beckmann, C. Michelsen & B. Sriraman [Eds.],
Proceedings of the 1st International Symposium on Mathematics and its Connections to the
Arts and Sciences. May 18-21, 2005, University of Schwaebisch Gmuend:
Germany.Franzbecker Verlag, pp. 32-51.
Lesh, R. & Zawojewski, J. S. (2007). Problem solving and modeling. The second handbook of
research on mathematics teaching and learning. Reston, VA: National Council of Teachers of
Mathematics.
Lesh, R., & Lamon, S. (Eds.). (1992). Assessment of authentic performance in school
mathematics. Washington, DC: American Association for the Advancement of Science.
Lesh, R., & Lehrer, R. (2000). Iterative refinement cycles for videotape analyses of conceptual
change. In A. Kelly & R. Lesh (Eds.), Research design in mathematics and science
education. (pp. 665-708). Dordrecht, Netherlands: Kleuwer Academic Press.
Lesh, R., Caylor, E. & Gupta, S. (2007) Data Modeling & the Infrastructural Nature of
Conceptual Tools. International Journal of Computers for Mathematical Learning. Springer

13
Lesh, R., Doerr, H. M., Carmona, G., & Hjalmarson, M. (2003). Beyond constructivism.
Mathematical Thinking and Learning, 5(2,3), 211-234
Lesh, R., Hoover, M., & Kelly, E. (1993). Equity, assessment, and thinking mathematically:
Principles for the design of model-eliciting activities. In I. Wirszup & R. Streit (Eds.),
Developments in school mathematics education around the world: Vol. 3 (pp. 104-130).
Teachers of Mathematics.
Lesh, R., Hoover, M., Hole. B., Kelly, E., & Post, T. (2000). Principles for developing thought-
revealing activities for students and teachers. In A. Kelly & R. Lesh (Eds.), Handbook of
research design in mathematics and science education. Mahwah, NJ: Erlbaum.
Lesh, R., Landau, M., & Hamilton, E. (1983). Conceptual models and applied mathematical
problem solving research. In R. Lesh & M. Landau (Eds.), Acquisition of mathematics
concepts and processes (pp. 263–343). NY: Academic Press.
Lesh, R., Post, T., & Behr, M. (1985). Representations and translations among representations, in
mathematics learning and problem solving. In C. Janvier (Ed.), Toward a theory of
mathematical representations. Hillsdale, NJ: Erlbaum
Lesh, R.A. & Doerr, H. (2003). Foundation of a Models and Modeling Perspective on
Mathematics teaching and Learning. In R.A. Lesh & H. Doerr (Eds.), Beyond
Constructivism: A models and modeling perspective on mathematics teaching, learning, and
problem solving. 9-34. Mahwah, NJ: Erlbaum.
Lesh. R. (2003) Models & Modeling in Mathematics Education. Monograph for International
Journal for Mathematical Thinking & Learning. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Lester, F. K., & Kehle, P. E. (2003). From problem solving to modeling: The evolution of
thinking about research on complex mathematical activity. In R. Lesh & H. Doerr, (Eds.),
Beyond constructivism: Models and modeling perspectives on mathematics problem solving,
learning and teaching (pp. 501-518). Mahwah, NJ: Erlbaum.
Mead, G. H. (1962). Mind, self, and society: From the standpoint of a social behaviorist. (C. W.
Morris, Ed.). Chicago: University of Chicago Press. (Original work published 1934).
Mead, G. H. (1977). The social psychology of George Herbert Mead; selected papers. In A.
Strauss (Ed.) Heritage of Sociology Series. Chicago, IL: University of Chicago Press.
(Original work published 1956).
Middleton, J. A., Silva, T. D., Toluk, Z. & Mitchell, W. (2001). The Emergence of Quotient
Understandings in a Fifth-Grade Classroom: A Classroom Teaching Experiment,
Proceedings of the Annual Meeting of the North American Chapter of the International
Group for the Psychology of Mathematics Education held at Snowbird, Utah 18-21 October
2001 (pp. 263-271)
National Council of Teachers of Mathematics. (2000). Principles and standards for school
mathematics. Reston, VA: Author.
Piaget, J., & Beth, E. (1966). Mathematical epistemology and psychology. Dordrecht, The
Netherlands: D. Reidel.

14
Popper, C. (1963) Conjectures and refutations: The growth of scientific knowledge. Cambridge
University Press.
Richardson, S. (2004). A design of useful implementation principles for the development,
diffusion, and appropriation of knowledge in mathematics classrooms. Unpublished doctoral
dissertation, Purdue University, West Lafayette, IN.
Sawyer, K. (2006). Explaining Creativity. New York: Oxford University Press.
Schoenfeld, A. (2007). Problem Solving Reconsidered: Toward a Theory of Goal-Directed
Behavior. Presentation given at the 2007 NCTM national conference pre-session. Address
delivered in Atlanta, GA, USA. 19 March 2007.
Schoenfeld, A. H. (1985). Mathematical problem solving. New York: Academic Press.
Schoenfeld, A. H. (1987). What’s all this fuss about metacognition? In A. Schoenfeld (Ed.),
Cognitive science and mathematics education (pp. 189-215). Hillsdale, NJ: Erlbaum.
Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition,
and sense making in mathematics. In D. Grouws (Ed.), Handbook of research on
mathematics teaching and learning (p. 334-370). New York: McMillan.
Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition,
and sense making in mathematics. In D. Grouws (Ed.), Handbook of research on
mathematics teaching and learning (p. 334-370). New York: McMillan.
Schorr, R. Y. & Lesh, (2003). A modeling approach for providing teacher development. In R.
Lesh & H. Doerr (Eds.) Beyond Constructivism: Models and Modeling Perspectives on
Mathematics Problem Solving, Learning and Teaching (pp. 141-158). Mahwah, NJ:
Erlbaum.
Silver, E. A. (Ed.). (1985). Teaching and learning mathematical problem solving: Multiple
research perspectives. Hillsdale, NJ: Erlbaum.
Steffe, L., Von Glasersfeld, E., Richards, J., Cobb, P. (1983). Children's Counting Types:
Philosophy, Theory, and Application. Greenwood Pub Group
Tan, J. (2002). Education in the twenty-first century: Challenges and dilemmas. In D. da Cunha
(Ed.), Singapore in the new millennium: Challenges facing the city-state. (pp. 154-186).
Singapore: The Institute of Southeast Asian Studies.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes.
Cambridge, MA: Harvard University Press.
Wilson, J., & Clark, D. (2004). Towards the modeling of mathematical metacognition.
Mathematics Education Research Journal, 16(2), 25-48.
Zawojewski, J. Diefes-Dux, H. & Bowman, K. (Eds.) (2008) Models and modeling in
Engineering Education: Designing experiences for all students. Rotterdam: Sense
Publications.

15
 A TECHNOLOGY-BASED INVESTIGATION OF UNITED STATES HIGH SCHOOL
STUDENT MATHEMATICAL PROBLEM SOLVING
Pamela L. Paek

Charles A. Dana Center, The University of Texas at Austin, Austin, Texas, United States

Summary

The purpose of this paper is to discuss an investigation of students’ behavior during

mathematical problem solving that may provide new ways for teachers to help students think
about mathematics. This study explored 122 high school students’ problem solving strategies,
using technology that enabled the researcher to track and model in detail the steps students used
to answer mathematical problems. This paper reports findings explaining the different problem
solving strategies and metacognitive approaches that students used to solve various types of
mathematics items. Based on study findings, the paper also recommends strategies teachers can
use to help students become more metacognitively aware of their mathematical thinking. Finally,
the paper describes the technology used to document and analyze student problem solving
behaviors.

Introduction describing the study and research questions

A major challenge for teachers of mathematics is understanding what students know and
what misconceptions deter them from solving problems correctly. Teachers can infer that
students who are higher achieving (as measured by grades and test scores) understand more than
do students who are lower achieving, but that inference merely allows us to stratify them, not to
deeply understand how they are engaging with the mathematics.

The purposes of this study, then, are 1) to identify the strategies that high school students
use when solving mathematics problems so as to better understand the processes they use and 2)
to uncover some potential reasons females underperform in mathematics compared to males.
This study enables a more detailed understanding of student test-taking behavior by providing a
more authentic look at what students do before they choose a final answer. Ultimately,
determining what enables higher-performing students to respond correctly can inform new ways
of conceptualizing instruction.
Theoretical framework

The need to improve students’ mathematical problem solving capacity in the United
States can be seen in the performance of U.S. 15-year-olds on the Program for International
Student Assessment [PISA] test, in which the U.S. ranked 24th out of 29 developed nations in
mathematics literacy and problem solving (Augustine, 2007, OECD, 2004). One reason for the
poor performance of U.S. students may be that these students are not provided with instruction
that successfully integrates content learning with experiences that foster their understanding of
problem solving processes (NRC, 2006).

16
Problem Solving

To improve mathematics instruction in problem solving, educators must first better

understand specifically what it is that students do and understand when they problem solve.
Strengthening students’ mathematical problem solving requires that teachers actively engage
students in a complex mathematical task in which students construct their understanding of
mathematics in nonroutine ways. Performing such a task, however, does not always imply a
thorough understanding of that task (Stevens, et al., 2004; 2005; 2006). Students can perform
mathematical steps without making the connections necessary to transfer or apply that
knowledge in a different context (Lajoie, 2003). Rather than simply performing steps, to solve a
problem successfully and understand what they are doing, students need to have knowledge of
the content and a strategy to solve the problem. They must monitor their progress through to the
solution and must be motivated to tackle the problem and solve it (Marshall, 1995). Instructional
approaches that emphasize that students take responsibility for their own learning—that is,
approaches that provide them with opportunities to choose, guide, and assess their own task
activities and progress—are key to student success in problem solving and in making the
connections necessary to apply their problem solving skills in new contexts (Inquiry Synthesis
Project, 2006; NRC, 2006).

When approaching mathematics problems, students rely on various resources and types
of information (Chi & Glaser, 1985; Ericsson & Simon, 1993; Schoenfeld, 1988). These
resources and types of information form the frameworks that students use to interpret and solve
different items. Ideally, students are able to identify and interpret a problem sufficiently well to
choose the correct framework for solving it, resulting in a correct response. In reality, however,
many students sometimes do not know which framework to choose, or choose inappropriate
frameworks, which results in inconsistent patterns of correct and incorrect responses (Marshall,
1995; Tatsuoka, 1993). Furthermore, choosing an appropriate framework does not necessarily
always lead to a correct answer, because computational errors can also lead to an incorrect
response (Tatsuoka, 1990). Thus, trying to explicate students’ problem-solving processes
requires that teachers and researchers look beyond their correct and incorrect answers and
undertake instead a detailed, empirical investigation of how students organize information.

Every mathematics problem contains a host of concepts that can be linked to a group of
specific steps that must be followed to successfully solve the problem. When students choose to
follow certain steps in a particular order, they are demonstrating a pathway for organizing
information to solve that problem. By assembling detailed information on students’ responses to
multiple problems, researchers can trace the steps that students took to solve each problem and
evaluate how well their approaches worked on different types of items—at the individual student
level and across various groups of students (e.g., by classrooms). Most students refine their
problem-solving strategies over time—which is consistent with models of skill acquisition
(Ericsson, 2004)—gradually using fewer steps and eventually settling on a preferred approach
(Stevens, Soller, Cooper, & Sprang, 2004; Stevens & Soller, 2005). Researchers can analyze
student self-regulation and self-monitoring of strategies by investigating these steps, which will
provide a better understanding of the complexity of student problem-solving performance
(Hartley & Bendixen, 2001; Song & Hill, 2007).

17
IMMEX: Using Technology to Study Problem Solving

Given the myriad ways that students can organize information and regulate the steps they
take to solve problems, the project of truly understanding what students are thinking and doing as
they work problems can seem insurmountable. Advances in technology, however, have provided
one way to get inside students’ heads, as it now enables us to track in detail the actual steps
students take to solve problems and the time they spend on each step (Hartley & Bendixen,
2001).

One pioneering technology that enables better understandings of student thinking is the
Integrated (now Interactive) Multi-Media Exercises (IMMEX) program, which draws on both
case-based (Schank, 1990) and production system (Newell & Simon, 1972) models of problem
solving. Such a tool gives a more qualitative look at how students solve problems, since it
captures in intricate detail the variety of approaches they can take. This captured information
opens the door to a deeper understanding of the comprehensive nature of student thinking
because rather than analyzing only a student’s final answer, this tool allows researchers to look at
each step that led to that answer.

IMMEX presents a problem to be solved in a problem space—that is, a space with a

finite set of concepts, numbers, and equations that students must combine in order to create a
solution path. Within the IMMEX problem space, various drop-down menus provide pathways
the students can choose from. A problem space typically will not encompass all the combinations
that students could use to solve problems, but it does provide an essential preliminary view for
better estimating what they are able to do (Tatsuoka, 1990, 1995). Further, although IMMEX’s
simulated problem spaces are finite, they do provide enough different types of information that
students with diverse math backgrounds could successfully solve the problems.

Working in IMMEX, students can assess the problem structure—the information needed
to solve the problem—and then organize a mathematical representation—an arrangement of that
information into a series of steps that solves the problem (Bennett, Morely, & Quardt, 1998).
Most students want to arrive at an answer and will follow some process to produce one. If their
chosen process leads to a wrong answer, they will probably try a different process if given a
chance to try again. IMMEX software allows researchers to track all the steps—forward,
backward, and sideways—that students take as they attempt to solve problems. IMMEX also
records and displays the sequence of steps and the time spent on each one (Ericsson & Simon,
1993; Stevens, 1991).

Design and Methodology

This study analyzed—and compared across gender—the problem-solving strategies of

122 high school students in their junior and senior years (72 females and 50 males) on a set of
SAT mathematics problems administered both on paper and in IMMEX. I also interviewed a
subset of students for their feedback and impressions about solving problems online, a mode that
forced them to show their steps.

18
Methods

To understand the specific steps that students take when tackling different kinds of
mathematical problems, this study used Interactive Multi-Media Exercises (IMMEX;
www.immex.ucla.edu), which consists of a library of online multimedia simulations for problem
solving. IMMEX has a refined set of modeling tools for monitoring student’s performance and
progress (Stevens et al., 2004, Stevens & Soller, 2005; Soller & Stevens, 2007). I presented
students with 25 mathematics items on the IMMEX platform. For each problem, IMMEX
provided students with various menus to choose from, in a simulated problem space that is
composed of a finite but representative set of concepts, numbers, and equations that students can
combine to create a solution path. With these menus, students had to identify, define, and
represent their steps to solve the problem. They developed reasons for choosing information that
might or might not be productive in helping them find an answer (Baxter & Glaser, 1997;
Stevens, et al., 2003; 2004; 2005; 2006). This case-based paradigm simulates situations with
sufficient information that students with diverse experiences can successfully synthesize and thus
solve the problem. IMMEX allows for detailed investigation of the steps and procedures that
students use to complete a task, because all steps are documented by sequence and time spent per
step.

I analyzed students’ strategies to see which students planned and self-regulated their
learning. I found that the more successful problem solvers tended to take time initially to identify
the constraints of the problem and the steps necessary to solve it before embarking on those
steps. As a result, they took fewer steps, used fewer nonproductive steps, and looked ahead by
assembling their procedures before acting on them. By breaking a problem into manageable and
familiar steps, successful problem solvers can chunk information into concepts and hierarchies
that facilitate good problem solving (Baxter & Glaser, 1997; Paek, 2002; Siegler, 1988). This
finding suggests that teachers should help students become more purposeful in (or regulate) their
problem solving strategies.

Procedure

Participants first completed a retired SAT-I test from the 10 Real SATs (College Board,
1997) under standard SAT time constraints using paper and pencil. Next, students used IMMEX
to solve 31 SAT-I mathematics problems that came from the same SAT-I test the students had
completed for the first part of the study. Eight students also participated in a focus group to
reflect on their problem solving strategies and to explain their thinking behind their steps.

Developing the IMMEX Problem Space

I created a specific problem space in IMMEX for solving each of the 31 mathematics
items along with a method for using and coding the search path maps. The problem space
included formulas, definitions of concepts, and a breakdown of the process for arriving at a
solution. The problem space for this study was developed based on a formal task analysis the
researcher had already conducted, for which students listed the steps they used to solve certain
math items. Additionally, the researcher incorporated into the problem space common errors that
students make in arithmetic, basic algebra, and geometry (Tatsuoka, 1990, 1995). These two

19
elements (the formal task analysis and the incorporation of common errors) helped to determine
the menus and submenus needed for the IMMEX platform in this study so that the majority of
students could solve the problems using the information given (Mislevy, Yamamoto, & Anacker,
1991).

The problem space was the same for each of the 31 items in IMMEX, except for some of
the submenus, which were changed to correspond with the proper substeps and the numbers and
equations related to each problem. The problem space included the math concepts necessary for
a correct solution as well as bugs and distractors, which were included to track where students
made arithmetic errors or had misconceptions about the problem. Students could easily navigate
through all the menus and submenus and still not be able to correctly solve a problem—to reach
an accurate solution, they needed to know what kinds of information were pertinent and be able
to order that information correctly. Within the IMMEX problem space, each problem was
presented with five main menus. The problem and the five main menus were always at the top of
the screen, even when students navigated through the submenus. Each main menu represented
one of five math concepts: arithmetic, angles, area, perimeter, and solving equations. Clicking on
one of these menus revealed a host of submenus also representing math concepts. Clicking on a
submenu led to a series of equations and/or numbers. These equations/numbers were represented
in expanded form, so that the student had to decide where to combine or collapse terms. The
menu structure included shortcuts so that students could collapse several steps into one. For
example, consider the equation 2x + 10 = 5 – 3x. The traditional way of solving it would be first
to move the numbers to one side by subtracting 10: 2x + 10 – 10 = 5 – 10 – 3x. To get the
variables on one side of the equal sign, the next move would be to add 3x to both sides: 2x + 3x =
–5 – 3x + 3x. Then both sides of the equation would be divided by 5: 5x/(–5) = –5/5 to arrive at
the final response of x = –1. In IMMEX, the menu shortcuts enabled students to collapse these
three steps, computing the information in their heads so they could move to the answer in one
step.

The IMMEX problem space also included common arithmetic errors students could make
that were associated with the distractors offered on the paper-and-pencil test. In the problem
above, for example, students could incorrectly subtract by 3x so that the response would be 2x –
3x = –5, resulting in a final answer of x = 1. These types of mistakes were included as options in
the submenus and the equation structures. The purpose of these incorrect paths was to document
where students might go wrong in coming up with their final answer choice.

After completing all the work and arriving at an answer, students entered their responses
after clicking the “solve problem” button. The study was structured to give students two chances
to solve the problem so that they could reconsider each step they had taken and so that the
researcher would have an opportunity see how students revised their steps to answer the problem
correctly on the second try. The second chance also allowed students who might have made a
simple arithmetic error to backtrack and correct it.

20
 Research results

Number of Steps Taken

Using the data IMMEX collected, the number and types of steps the students took to
solve each problem were analyzed. In general, the fewer steps a student had taken in attempting
to solve a problem, the more likely it was that the student had solved the problem correctly—as
students who unsuccessfully completed a problem tended to take more irrelevant steps that were
not helpful to solving the problem. In addition, the number of steps taken differed between males
and females. On average, males took two fewer steps than females did to solve a problem: males
averaged four or fewer steps (M = 3.91, SD = 2.06), whereas females averaged six or more steps
(M = 6.12, SD = 1.77). Even with these differences, females attempted more IMMEX items,
took longer to solve each problem, and answered more problems correctly than males did; the
reason for the higher success of females on these items appears to be that they verified their
steps, not that they were inefficiently taking extra steps. On the paper-and-pencil SAT-I
mathematics sections, however, females scored lower than did the males. The IMMEX test had
no time constraints, so it may be that females performed better in the untimed situation than
males did.

Informal interviews with the participants suggested that the females liked to be sure of
their answers and would use any available information to verify them. They wanted their
answers to be correct on the first attempt, and they took more steps and more time to ensure
correctness. The interviews suggested that males, on the other hand, tended to be inclined toward
an answer and would select it, knowing they had a second chance if they got it wrong. This
method resulted in fewer steps and less time taken per problem. Males indicated they were also
more likely to guess once they had eliminated some choices.

Amount of Time Spent per Step and per Problem

I also analyzed time spent per step and per problem. Females tended to take 2 s more per
step (M = 19, SD = 21) than did males (M = 17, SD = 18), which resulted in females taking about
55 s more per problem (M = 2:06, SD = 2:51) than males (M = 1:11, SD = 1:45). These
differences are statistically significant (p < 0.01). This difference in time spent per step, coupled
with the fact that females took more steps than males did, may well help to explain why females
tend to score lower on standardized mathematics tests: They are not able to complete as many
problems.

Number of Attempts Made

Students were given two opportunities to solve each IMMEX problem, so I could
document the changes they made in their steps from the first to the second attempt. The majority
(61%) of students solved the problems correctly on the first try, and an additional 23% answered
the problems correctly on their second try. The steps students took on the second tries for both
correct and incorrect answers were analyzed. Students who correctly solved a problem on the
second try demonstrated an orderly process in which they deliberately retraced their steps to
verify their answers and more systematically regulated their steps, indicating that these students

21
knew what they were doing but had made a small error in computation at some point. Students
who did not solve the problem correctly on the second try, on the other hand, showed less
organization and planning in their process.

Observing the processes used by the participants gives an inside view of how they
regulated their learning as they solved problems, and suggests reasons for the performance
differences documented between females and males. The amount of time and number of steps to
solve each problem varied between males and females, with females taking extra steps to verify
their answers and therefore taking more overall steps per problem than males. This verification
process resulted in females spending more time on each problem than did males. The extra steps
and time, however, paid off in the females performing slightly higher than the males on the
IMMEX problems.

Discussion

The results confirmed the outcomes from previous research (e.g., Gallagher, 1990; 1992;
Gallagher & De Lisi, 1994) about differences in gender in problem solving: females tended more
than males to follow algorithms and verify their steps and the answer before moving on to the
next problem. In the interviews, females articulated a greater need for verification in their work.

This research shows the importance of understanding in detail the steps that students take
when solving mathematics problems. Tracing students’ steps allows researchers to better
understand how students organize information when coming up with an answer. In the present
study, tracing students’ steps illuminated some of the reasons that females’ math test scores are
typically lower than males’ scores. Finding out what knowledge students possess no longer needs
to be surmised only from final answers and scores, as the use of IMMEX in this study
demonstrates. Researchers can and must continue to probe the processes that students employ
and the knowledge they bring to bear when confronted with a mathematics problem. The more
deeply that educators and researchers can analyze student thinking, the better we can measure
students’ competence, knowledge, and abilities—and thus the better we can design tools and
practices for teachers to teach them effectively.

This study contributes to the literature on the mechanics of students’ problem solving
(e.g., Baxter & Glaser, 1997; Lajoie, 2003; Marshall, 2005; Stevens, et al., 2003; 2004; 2005;
2006). The findings from this study also show how technology such as IMMEX can provide two
key strategies for improving mathematics teaching and learning: it can provide teachers with
access to student thinking that is usually not obvious and thus enable them to modify instruction
appropriately (Pellegrino, et al. 2001), and it can provide opportunities for students to reflect and
fine-tune their problem solving strategies, giving them a strong context for thinking about and
being successful in mathematics. A main implication of this study is for educators to find ways
to increase students’ metacognitive skills in mathematics so that when they participate in
assessments, their performance reflects their actual understanding rather than their habitual
approaches to problem solving.

The differences that the present study found in self-regulation of problem solving
demonstrate the importance of teaching students how to plan their problem-solving. An

22
underutilized strategy in instruction is having students practice using an overarching schema for
new and novel problems. Other strategies include having students plan their steps before they
actually begin to solve a problem, and then having them reflect on these plans and actions. In
fact, a five-stage problem-solving process that is recommended in most textbooks and resources
emphasizes that students should begin by reading and understanding the statement of the
problem, then analyze the given facts, and then propose steps to solve the problem. These three
stages are the planning phases, which should take place before students actually work the
problem. The final two stages are carrying out the planned steps and then verifying the solution.
On the basis of this study, a sixth stage is recommended: Teachers should provide opportunities
for students to reflect on the steps they took to solve the problem. This allows students time to
reflect on their previous work and, if necessary, plan a better set of steps for upcoming problems.

References

Augustine, N. R., National Academies Committee on Prospering in the Global Economy of the
21st Century. (2007). Rising above the gathering storm: Energizing and employing
America for a brighter economic future. Washington, D.C.: National Academies Press.

Baxter, G. P. & Glaser, R. (1997). An approach to analyzing the cognitive complexity of science
performance assessments. National Center for Research on Evaluation, Standards, and
Student Testing, Center for the Study of Evaluation, Graduate School of Education and
Information Studies, University of California, Los Angeles.

Chi, M. T. H., & Glaser, R. (1985). Problem-solving ability. Report no. LRDC-1985/6.
Pittsburgh, PA: University of Pittsburgh Learning Research and Development Center.

College Board (1997). 10 Real SATs. Forrester Center, WV: Author.

Ericsson, K. A., & Simon, H. A. (1993). Protocol analysis: Verbal reports as data. Revised
edition. Cambridge, MA: MIT Press.

Ericsson, K.A. (2004). Deliberate practice and the acquisition and maintenance of expert
performance in medicine and related domains. Academic Medicine 79(10), 70–81.

Gallagher, A. M. (1990). Sex differences in the performance of high-scoring examinees on the

SAT-M (RR 90-27). Princeton, NJ: Educational Testing Service.

Gallagher, A. M. (1992). Strategy use on multiple-choice and free-response items: An analysis of

sex differences among high-scoring examinees on the SAT-M (RR 92-33). Princeton, NJ:
Educational Testing Service.

Gallagher, A. M., & De Lisi, R. (1994). Gender differences in the Scholastic Aptitude Test-
Mathematics problem solving among high-ability students. Journal of Educational
Psychology, 86(2). 204–211.

23
Hartley, K., & Bendixen, L.D. (2001). Educational research in the Internet age: Examining the
role of individual characteristics. Educational Researcher, 30(9), 22–26.

Inquiry Synthesis Project, Center for Science Education, Education Development Center, Inc.
(EDC) (2006, April). Technical report 2: Conceptualizing inquiry science instruction.
Retrieved December 13, 2007, from. http://cse.edc.org/products/inquirysynth/pdfs/
technicalReport2.pdf.

Lajoie, S. P. (2003). Transitions and trajectories for studies of expertise. Educational Researcher,
32 (8), 21–25.

Marshall, S. P. (1995). Schemas in problem solving. New York, NY: Cambridge University
Press.

Mislevy, R. J., Yamamoto, K., & Anacker, S. (1991). Toward test theory for assessing student
understanding (RR 91-32-ONR). Princeton, NJ: Educational Testing Service.

Newell, A., & Simon, H. A. (1972). Human problem solving. Englewood Cliffs, NJ: Prentice
Hall.

National Research Council. (2006). America’s lab report: Investigations in high school science.
Washington, DC: National Academy Press.

Organization for Economic Co-operation and Development (OECD). (2004). Problem solving
for tomorrow’s world: First measures of cross-curricular competencies from PISA 2003.
Claire Shewbridge and Andreas Schleicher (Eds), Programme for International Student
Assessment, Organization for Economic Co-operation and Development (OECD). Paris,
France.

Paek, P.L. (2002). Problem solving strategies and metacognitive skills on SAT mathematics
items. Doctoral dissertation, University of California, Berkeley). Dissertation Abstracts
International 63(09), 3139.

Pellegrino, J. W., Chudowsky, N., & Glaser, R. (Eds.). (2001). Knowing what students know:
The science and design of educational assessment. Washington, DC: National Academy
Press.

Schank, R. C. (1990). Case-based teaching: Four experiences in educational software design.

Interactive Learning Environments, 1(4), 31–53.

Schoenfeld, A. H. (1988). Problem solving in context(s). In R. I. Charles & E. A. Silver (Eds.),

The teaching and assessing of mathematical problem solving (pp. 82–92). Reston, VA:
National Council of Teachers of Mathematics.

Siegler, R. S. (1998). Individual differences in strategy choices: Good students, not-so good
students, and perfectionists. Child Development, 59, 833–851.

24
Soller, A., & Stevens, R. (2007). Applications of stochastic analyses for collaborative learning
and cognitive assessment. In G. R. Hancock and K. M. Samuelsen (Eds.), Advances in
Latent Variable Mixture Models. Charlotte, NC: Information Age Publishing.

Song, L. & Hill, J. R. (2007). A conceptual model for understanding self-directed learning in
online environments. Journal of Interactive Online Learning, 6(1). Retrieved March 12,
2008, from
http://ncolr.org/jiol/issues/viewarticle.cfm?volID=6&IssueID=19&ArticleID=98.

Stevens, R. H. (1991). Search path mapping: a versatile approach for visualizing problem-
solving behavior. Academic Medicine, 66 (9), S72–S75.

Stevens, R. H., & Casillas, A. (2006). Artificial neural networks. In D. M. Williamson, I. I.

Bejar, & R. J. Mislevy (Eds.), Automated Scoring of Complex Tasks in Computer Based
Testing. Mahwah, NJ: Lawrence Erlbaum Associates, pp. 259–312.

Stevens, R. H, & Palacio-Cayetano, J. (2003). Design and performance frameworks for

constructing problem-solving simulations. Cell Biology Education, 2(3),162–179.

Stevens, R. H., & Soller, A. (2005). Machine learning models of problem space navigation: The
influence of gender. ComSIS, 2(2), 83–98.

Stevens, R., Soller, A., Cooper, M., & Sprang, M. (2004). Modeling the development of problem
solving skills in chemistry with a web-based tutor. In J. C. Lester, R. M. Vicari, & F.
Paraguaca (Eds.), Intelligent Tutoring Systems. Springer-Verlag Berlin Heidelberg,
Germany. 7th International Conference Proceedings, pp. 580–591.

Tatsuoka, K. K. (1990). Toward an integration of item-response theory and cognitive error

diagnosis. In N. Frederiksen, R. Glaser, A. Lesgold, & M. G. Shafto (Eds.), Diagnostic
monitoring of skill and knowledge acquisition. Hillsdale, NJ: Lawrence Erlbaum
Associates.

Tatsuoka, K. K. (1993). Item constructive and psychometric models appropriate for constructed
responses. In R .E. Bennett & W. C. Ward (Eds.), Construction versus choice in cognitive
measurement: Issues in constructed response, performance testing and portfolio
assessment. Hillsdale, NJ: Lawrence Erlbaum Associates.

Tatsuoka, K. K. (1995). Architecture of knowledge structures and cognitive diagnosis: A

statistical pattern recognition and classification approach. In P. D. Nichols, S. F.
Chipman, & R. L. Brennan (Eds.), Cognitively diagnostic assignment. Hillsdale, NJ :
Lawrence Erlbaum Associates.

25
 Formulating mathematical conjectures in learning activities, assisted with technology
Fernando Barrera Mora Aarón Reyes Rodríguez
barrera@uaeh.edu.mx aaron.reyes.rdz@gmail.com

Universidad Autónoma del Estado de Hidalgo

Summary
What types of activities should professional development programs include to revise and extend
high school teachers’ mathematical and pedagogical knowledge? We propose a route to engage
high school teachers in an inquiry approach to reflect on their current practice and to construct
hypothetical learning trajectories that can eventually guide or orient the development of their
lessons. In this report, we focus on the activities that were worked within a professional
community that include the participation of mathematicians, mathematics educators and
doctoral students.

Introduction
What mathematical and pedagogical knowledge should the education of high school
mathematics teachers include? Who should participate in the educational programs to prepare
mathematics teachers? What should be the role of mathematics departments or the faculty of
education in preparing prospective and practicing teachers? What types of educational programs
should practicing teachers participate in order to revise and extend their mathematical knowledge
and to incorporate research results from mathematics education into their practices? Traditional
ways to prepare high school teachers normally involve the participation of both mathematics
departments and the faculty of education. Mathematics departments offer courses in mathematics
while the faculty of education provides the didactical or pedagogical courses. This model of
preparing teachers has not rendered solid basis to help teachers provide an instructional
environment in which they exhibit mathematical sophistication to interpret and prompt students’
responses and to organize and implement meaningful learning activities for their students.
Indeed, it is common to read that university instructors complain that their first year university
students lack not only fundamental mathematical knowledge; but also strategies or resources to
solve problems that require more than the use of rules or formulae.
Many practicing teachers, for different reasons, have not learned some of the content they are now required to
teach, or they have not learned it in ways that enable them to teach what is now required. …Teachers need
support if the goal of mathematical proficiency for all is to be reached. The demands this makes on teacher
educators and the enterprise of teacher education are substantial, and often under-appreciated (Adler, et al., 2005,
p. 361).
Davis and Simmt (2006) suggest that teachers’ preparation programs should focus more on
teachers’ construction of mathematical ideas or relations to appreciate their connections,
interpretations, and the use of various types of arguments to validate and support those relations,
rather than the study of formal mathematics courses. Thus, the context to build up their
mathematical knowledge should be related to the needs associated with their instructional
practices. “… [mathematical knowledge] needed for teaching is not a watered version of formal
mathematics, but a serious and demanding area of mathematical work” (Davis and Simmt, 2006,
p. 295). In this work, we report that teachers’ mathematical knowledge can be revised and
enhanced within an interacting intellectual community that fosters an inquisitive approach to

26
develop mathematical ideas and to promote problem-solving activities. The core of this
community should include mathematicians, mathematics educators, and practicing teachers. This
community promotes collaborative work to construct potential learning trajectories to guide or
orient the teachers’ instructional practices. Teachers need to be interacting within a community
that supports and provides them with collegial input and the opportunity to share and discuss
their ideas in order to enrich their mathematical knowledge and problem solving strategies. In
this context, we illustrate the importance of using computational tools to represent and explore
various ways of approaching mathematical tasks.

Research questions
Several research works (Santos-Trigo, 2004; Schoenfeld, 1994, 2000; NCTM, 2000) emphasize
the importance of formulating and validating conjectures when learning and developing
mathematics. Conjecturing processes involve several dimensions when technology is used
systematically. For instance, the idea of generalizing is widely amplified and at the same time
one has the opportunity to ask questions about the way in which a particular computational
system works. The questions that guided this research are: What type of mathematical reasoning
might be developed by high school teachers in order to reconstruct or enhance their mathematical
knowledge when using technology to explore hypothetical learning trajectories? What type of
mathematical arguments might high school teachers use to explain unexpected computer
mathematical results?

Conceptual Framework
The conceptual framework is structured around two main theoretical issues: (i) problem solving
and technology and (ii) hypothetical learning trajectories. We have chosen these constructs since
learning mathematics is achieved through problem solving, which is enhanced by using
technological tools. In this regard, we argue that promoting an inquiring approach when learning
mathematics can be attained effectively by formulating questions and elaborating conjectures
systematically. This path is strongly related with the finding and exploration of different
hypothetical learning trajectories.

Problem solving and the use of technology

In problem solving activities it has been recognized the relevance of an inquiring or inquisitive
approach for teachers to work on mathematical tasks or to think of and reflect on their
instructional activities. In this context, learning mathematics or developing mathematical
knowledge through problem solving is conceptualized as working with tasks where:

A task, or goal-directed activity, becomes a problem (or problematic) when the “problem solver” (which may be
a collaborating group of specialists) needs to develop a more productive way of thinking about the given
situation (Lesh and Zawojewski , 2007, p. 782).

It is important to clarify what is understood by a productive way of thinking. According with the
same authors “…Developing a ‘productive way of thinking’ means that the problem solver needs
to engage in a process of interpreting the situations, which in mathematics means modeling” (p.
782).

In this context, an important component is to develop an inquiring way of thinking to formulate

questions, to identify and investigate dilemmas, to search for evidences or information, to discuss

27
solutions, and to present or communicate results. This means willingness to wonder, to pose and
examine questions, and to develop mathematical understanding within a community that values
both collaboration and constant reflection. At this point Schoenfeld (1994) argues:
“Mathematicians develop much of that deep mathematical understanding by virtue of
apprenticeship in to that community [mathematical community]–typically in graduate school and
as young professionals” (p. 68). A mode of inquiry involves necessarily the challenges of the
status quo and a continuous re-conceptualization of what is learned and how knowledge is
constructed.

[In a community of inquiry] participants grow into and contribute to continual reconstitution of the community
through critical reflection; inquiry is developed as one of the forms of practice within the community and
individual identity develops through reflective inquiry (Jaworski, 2006, p. 202).
Taking this view into account, and considering that the use of technology has been playing an
important role in the process of mathematical learning by enhancing different elements of
mathematical thinking, particularly formulating and validating conjectures, it is relevant to ask:
what is the role of a computer system in the process of posing and justifying conjectures? How
trustable are the results obtained with the aid of a computer system?

Concerning the first question Santos (2007) argues: “A relevant aspect when representing a task
with the aid of a dynamical software is that students have the opportunity to pose questions about
the structure of some elements of the configuration” (p. 124).

Regarding the second question, Dick (2007) has introduced the term Mathematical Fidelity “to
emphasize that the mathematics of the tool does not always represent the mathematics as it is
understood by the mathematics community” (p.1174). In the example that we will discuss, it will
be pointed out the strong necessity of providing mathematical arguments to deal with
discrepancies between the computers results and the expected ones.

Hypothetical Learning Trajectories

To promote the teachers’ inquiring approach to their practice we rely on the construction of
Hypothetical Learning Trajectories (HLT). These trajectories emerge from examining potential
routes of solution of mathematical tasks. Simon and Tzur (2004) state that the construction of
hypothetical learning trajectories is based on the following assumptions:
1. Generation of an HLT is based on understanding of the current knowledge of the students involved.
2. An HLT is a vehicle for planning learning of particular mathematical concepts.
3. Mathematical tasks provide tools for promoting learning of particular mathematical concepts and are,
therefore, a key part of the instructional process.
4. Because of the hypothetical and inherently uncertain nature of this process, the teacher is regularly involved in
modifying every aspect of the HLT (p. 93).
In this perspective, we suggest that teachers together with other community members (including
mathematicians), work on various ways to approach the tasks and to identify relevant concepts
and problem solving strategies needed to solve them. We argue that this type of teachers’
interaction becomes relevant to foster and to develop a problem solving approach that involves:
Seeing the mathematical content in mathematically unsophisticated questions, seeing underlying similarity of
structure in apparently different problems, facility in drawing on different mathematical representations of a

28
 problem, communicating mathematics meaningfully to diverse audiences, facility in selecting and using
appropriate modes of analysis (“mental”, paper and pencil, or technological), and willingness to keep learning
new material and techniques (Cohen, 2001, p. 896).
In addition, we also recognize that the use of computational tools offers to teachers the
opportunity to enhance relevant aspects of mathematical thinking as well as to represent and
examine mathematical tasks in terms of questions that can lead them to develop or reconstruct
some mathematical results. For instance, the use of a dynamic software allows teachers to
represent problems dynamically in order to recognize and explore mathematical relations within
a geometrical configuration, and to identify loci described by members of the configuration when
others are moved. In this context, the use of computational tools becomes important for teachers
to discuss pedagogical paths associated with the hypothetical learning trajectories that can be
useful to guide or orient their instructional practices.

We claim that the inquiring process is strongly intertwined with the appearance of hypothetical
learning trajectories derived from a problem solving activity. By this we mean that in the process
of formulating questions, there arises the opportunity to learn or reconstruct new mathematical
concepts that emerge while pursuing those questions.

Research Design, Methods and General Procedures

The Center for Research and Applied Mathematics, that is part of a public university, is in charge
of developing and implementing a professional program to revise and improve high school
teachers’ mathematical and didactical knowledge. As a part of the program, we coordinated a 40
hrs instructional module, out of four, whose main aim was to illustrate and discuss the strengths
and limitations of using computational tools in problem solving activities. To this end, a group
that includes two mathematicians, one mathematics educator, and two doctoral students met
together during two months, in sessions of three hours a week, to select and discuss the tasks that
later would be used during the development or implementation of the activities with the teachers.
During each session, one member of the group presented one problem and provided information
related to its relevance or rationale and ideas about possible solutions. Then, all participants
became engaged in the solution process and at the end they discussed and summarized the main
ideas and ways used to approach the task. In this report, we focus on presenting general features
of hypothetical learning trajectories that emerged as a result of working on those tasks that later
we used to structure the development of the sessions with the actual group of teachers who
participated in this module. Our unit of analysis is the work done and reported by the participants
(the mathematicians, the math educator and the doctoral students) as a group. Thus, in presenting
the results we use the word group to identify and characterize that work. The selected problems
came from textbooks and research articles. There were also those that include the construction of
an initial dynamic configuration (using dynamic software) in which teachers could formulate
their own questions or problems.

The task: This problem involves an extension of a task discussed in Santos, et al. (2006, p. 125).
In particular, the working group constructed a hypothetical route for teachers to develop an
inquiring approach to the tasks in which the use of technology is encouraged. The problem arises
from analyzing invariance and structure of simple components of a geometric configuration in
order to identify an instructional path to foster the teachers’ construction of mathematical
relations.

29
 R
Given a straight line L, a point P in L and a
point Q not in L, draw the segment PQ, a
line L1 perpendicular to PQ through Q and a
line L2 perpendicular to L through P. Call R Q
L2 L1
the intersection of L1 and L2 . What is the
L
locus of R when P runs on L?
P

Figure 1: What is the locus of point R when

point P is moved along line L?
Research Results and Discussion
Figure 2 shows the different stages that appeared during the solution process of the task as well
as main features that guide the construction of the different hypothetical learning trajectories. It
should be mention that the diagram shows only one instructional route that the group proposed
during the discussion of the task, however there were two more possible routes (finding the
triangle with minimum area and the case where the locus is a hyperbola), whose discussion is not
presented.

In this perspective, the meaning associated with the main stages that characterize the potential
instructional trajectory involves: (i) the recognition of the high school teachers’ knowledge base
to represent and explore the initial task, (ii) the recognition that the aim of the developed task is
to provide conditions in order that high school teachers reinforce and reconstruct their
mathematical concepts in such way that this would help them to design and guide learning
activities in the classroom, (iii) the discussed problem arises from analyzing minimal elements in
a geometric configuration with the objective of designing learning tasks and (iv) the possibility
that the teachers will bring into the discussion additional elements to modify every aspect of the
hypothetical learning trajectory after they have solved the task.

One of the members of the discussion group suggested to approach the problem using Cabri-
Geometry to construct the geometric configuration, after this, using the tool Locus, it was asked
the software to describe the locus drawn by point R when P moves on L. Cabri-Geometry shows
a graph that looks like a parabola, Figure 1. With this information, some of the members of the
group went further in conjecturing, using Cabri-Geometry's tool Equation or coordinate: the
equation of the locus described by R corresponds to a parabola. At this point there was
consensus that formal arguments were needed in order to continuous with the analysis to find
connections and generalizations.
Using a coordinate system. An algebraic approach becomes important to construct an argument
to show that the locus is a parabola. Here, the group used a Cartesian System in a proper position
to facilitate algebraic operations.

Without loss of generality, one can assume that L coincides with the x axes, P = (t ,0) and
Q = (a, b) . In order to determine the coordinates of R, one finds the equations of L1 , which turns

30
out to be y " b = (( t " a) b)( x " a) , the equation of L2 is x = t . Solving the system determined by
2
these two equations yields y " b = ( x " a) b … (*), which is in fact the equation of a parabola
since a and b are fixed.
!
!

Figure 2: Hypothetical learning process in the context of a particular task.

At this stage, the dynamic representation of the task becomes a departure point to identify and
explore diverse mathematical relations. Here, we document ways in which the working group
explored the following general cases:

(a) Same assumptions on L, P and Q but now, it was taken an additional point Q' on the segment

31
 PQ and the line L1 that passes through the point Q'. What is the locus described by R when P
moves along L? How does the locus change when Q' moves along the segment PQ?

An interesting part of the use of Cabri Geometry to formulate conjectures is that after proving the
result the discussion group obtain more accurate information about the parabola. For example,
knowing the focus and the directrix, the group could formulate the result in terms of synthetic
geometry.

Let L be a line, Q a point not in L, P ! L, L1 the line that passes through Q and P. Take a point
Q’ ! L1 and draw the perpendicular line to L1 that passes through Q’, calling it L2. Through P,
Q’ and Q draw perpendicular lines to L, calling these lines L3, L4 and L5, respectively. Let T, S
and R be the points of intersection of the lines L and L4; L and L5; L2 y L3, respectively. Through
Q’ draw a perpendicular line to L4 that intersects L3 and L5 at E and V respectively. Let F and W
be points on L5 such that WV = VF = QS 2 4Q' T . Let L6 be the perpendicular to L5 that passes
through W and intersects L3 at U. Then L6 and F are the directrix and focus of a parabola with
vertex at V.

Proof. The claim!is equivalent to show that UR=FR. We have:

L
FR 2 = VE 2 + (UR " 2VF) 2 …(1)
From the similar triangles PQ'T and PQS one
has:
! QS VE
= ,
Q'T Q' E

and from this, one obtains:

SQ
! VE = Q' E .
TQ'
Figure 3: We have to prove that UR = FR.

Substituting the value of VF and VE in equation (1) and developing the binomial one arrives to:
!
SQ2Q' E 2 QS 2 QS 4
FR 2 = + UR 2
" UR +
Q'T 2 Q'T 4Q'T 2
SQ2 # Q' E 2 &
= UR 2 + % " UR + FV (.
Q'T $ Q'T '
From the triangle RQ'P we have Q' E 2 = ( PE )( ER) ; on the other hand PE=Q'T, hence from the
previous equation one concludes that:
!
SQ2
FR 2 = UR 2 + ( ER " UR " FV ) .
Q'T
We also have ER-UR=-EU=-VW=-VF; from which the conclusion follows.

In the last result, the !

group assumed that the second coordinate of the point Q’ does not change;
with this in mind some of the members of the group ask a very natural question. What would

32
 happen if this condition is replaced by: the distance from Q to Q’ remains constant?

(b) Assuming that L, P and Q are as above, but now the point Q' is the intersection of the line L',
passing through P and Q, and the circle C of radius r with center at Q. The lines L1 and L2 are
constructed as before, and so is R. What is the locus described by R when P moves along L?
How does the locus behave when r approaches zero?
y y

R
R
L'
Q
Q
Q' Q'
L2 L2
1
L1 1
L x L1 x
1
1
P L P

Figure 4: What is the locus of point R when Figure 5: What is the locus of point R when
point P moves along line L? point P moves along line L?

In discussing part (b), with the use of Cabri Geometry the group has the chance to experiment
and observe the behavior of the locus generated by R. One first approach shows results as shown
in Figure 5, and it seems that the locus is a parabola, the Equation tool from Cabri Geometry
even suggests that we are dealing with a parabola.

Nevertheless, taking a closer look at the y

geometrical behavior of the locus generated

by R, there appears a graph as the one shown
in Figure 6, which cannot be identified with
the graph of a parabola. With this evidence, it 1
Q
L1
L2
is natural to ask for formal arguments to find L
1
x

P
out which kind of geometric object is
described by point R. After performing Q'

calculations using a Coordinate System the

group found that: R L'

# 1# 2 2 &&
R = % x, %( x " a) + b 2 ± r ( x " a) + b 2 (( , Figure 6: What locus is described by point R?
$ b$ ''

where the center of the circle is (a,b) . It should be noticed that the second coordinate of R

! [ 2
]
approaches ( x " a) b + b when r approaches zero, which is the same result as (*), page 6. This
result is consistent with the process of generalizing, an important aspect of the mathematical
thinking.
Also
! the participants asked questions related with the way that Cabri performs geometric
transformations. This led to think about the reliability of mathematical results obtained with the
aid of a computer system. Here the group had the opportunity to point out the necessity of

33
analyzing the process and results obtain from the technological tool and ask questions related to
the axiomatic system of it.

Closing remarks
Mathematical tasks are key elements of any professional development program that aims to
revise and enhance teachers’ mathematical and didactical knowledge. How should those tasks be
discussed with teachers in order to identify explicitly ways of reasoning that are consistent with
mathematical practice? We argue that tasks or problems should be addressed openly within an
inquisitive community that promotes collaboration and mathematical reflection. In this process,
the use of computational tools becomes relevant to represent some tasks dynamically and
visualize diverse mathematical relations embedded in those tasks. It is evident that the
conceptualization of the task as dilemmas, provide the opportunity to identify and explore
relations, to open diverse lines of thinking or reflection that can lead the community or the
problem solver to approach the task from diverse angles or perspectives. For example, the visual
and empirical approach becomes important to identify relevant information, possible relations,
and plausibility of solutions. The use of dynamic software offers the opportunity of utilizing
particular heuristic strategies (searching for partial solutions) to solve the problem. Thinking of
various approaches to the problem, another relevant problem solving activity, allows the problem
solver to identify fundamental properties of the solution and possible relations or connections.
Thus, problem solving is a continuous activity in which contents (from various domains),
resources and strategies are used to initially construct a hypothetical learning trajectory that can
be useful to orient and structure the practice of mathematical teachers. Finally, the group that
worked on the task recognizes the relevance of approaching them within an inquisitive or
inquiring community. The participants have developed a guide to implement the tasks. Of
course, the plan and activities to implement the tasks in the professional development program
were based on considering the trajectories that emerged during the group sessions.
An aspect, which is of crucial importance when using technological tools for solving
mathematical problems, is related to providing support or formal arguments to results produced
through the use of the tools. It is well accepted that technology is a powerful tool, however the
results obtained should be examined rigorously in order to be accepted or rejected. In this respect
Dick (2007, p. 1175) has introduced the term mathematical fidelity and has identify three areas
in which a lack of mathematical fidelity can emerge: (i) mathematical syntax, (ii)
underspecifications in mathematical structures and (iii) limitations in representing continuous
phenomena with discrete structures and finite precision numerical computation. However these
areas might not consider aspects related with reliability such as the results in the discussed
example. We think that results that disagree with the expected ones has to do with the internal
processing of the tool; related with this we suggest that a closer examination of the mathematical
structure of the tool has to be done. At this respect our opinion agrees with Zbiek et al. (2007)
whose statement is:
As technology becomes an increasingly integrated part of school mathematics, careful analysis of issues of
mathematical fidelity [and reliability] will be needed. This type of research will necessitate intense collaboration
involving mathematicians, computer scientist, and mathematics education researchers (p. 1176).
We also consider that this analysis should include categorizing levels of reliability of the tool.
For instance we claim that the basic arithmetic operations (addition and multiplication within the
precision range of calculators and computers) are 100% reliable. This is not the case for more
sophisticated mathematical operations.

34
Acknowledgment
The first author acknowledges the support received from Conacyt through research project with
reference #61996.
References

1. Adler, J., Ball, D., Krainer, K, Fou-Lai, L., & Novotna, J. (2005). Reflections on an emerging
field: Researching mathematics teacher education. Educational Studies in Mathematics, 60,
359-381.
2. Cohen, A. (2001). Two reactions to the mathematical education of teachers. Notices of the
AMS, 48(9), 985-991.
3. Davis, B. & Simmt, E. (2006). Mathematics-for-teaching: An ongoing investigation of the
mathematics that teachers (need) to know. Educational Studies in Mathematics, 61, 293-319.
4. Dick, T. (2007). Keeping the faith: Fidelity in technological tools for mathematics education.
In G. W. Blume & M. K. Heid (Eds.), Research on Technology and the Teaching and
Learning of Mathematics: Syntheses, cases, and perspectives. Vol. 2: Cases and Perspectives
(pp. 333-339). Greenwich, CT: Information Age.
5. Jaworski, B. (2006). Theory and practice in mathematics teaching development: Critical
inquiry as a mode of learning in teaching. Journal of Mathematics Teacher Education, 9(2),
187-211.
6. Lesh, R. & Zawojewski, J. (2007). Problem solving and modeling. In F. K. Lester Jr. (Ed.),
Second Handbook of Research on Mathematics Teaching and Learning (pp. 763-804).
Charlotte, NC: Information Age.
7. NCTM (2000). Principles and Standards for School Mathematics. Reston, Virginia: National
Council of Teachers of Mathematics.
8. Santos-Trigo, M. (2004). The role of dynamic software in the identification and construction
of mathematical relationships. Journal of Computers in Mathematics and Science Teaching,
23 (4), 399-413.
9. Santos, M. (2007). La resolución de problemas matemáticos. Fundamentos cognitivos.
México: Trillas.
10. Santos, M. et al. (2006). Constructing a Parabolas’ World Using Dynamic Software to
Explore Properties and Meanings. International Journal for Technology in Mathematics
Education, 2 (3), 125-134.
11. Schoenfeld, A. (1994). Reflections on doing and teaching mathematics. In A. Schoenfeld
(Ed.), Mathematical thinking and problem solving (pp. 53-70). Hillsdale, NJ: Lawrence
Erlbaum.
12. Schoenfeld, A. (2000). Purposes and Methods of Research in Mathematics Education,
Notices of the AMS, 47 (6), 641-649.
13. Simon, M. A. & Tzur, R. (2004). Explicating the role of mathematical tasks in conceptual
learning: An elaboration of the hypothetical learning trajectory. Mathematical Thinking and
Learning, 6(2), 91-104.
14. Zbiek, R. M. et al. (2007). Research on Technology in Mathematics Education. In F. K.
Lester Jr. (Ed.), Second Handbook of Research on Mathematics Teaching and Learning
(pp.1169-1207). Charlotte, NC: Information Age.

35
 PROBLEM POSING PERFORMANCE OF GRADE 9 STUDENTS IN SINGAPORE
ON AN OPEN-ENDED STIMULUS

CHUA, Puay Huat <puayhuat.chua@nie.edu.sg>

National Institute of Education, Nanyang Technological University, Singapore

YEAP, Ban Har <banhar.yeap@nie.edu.sg>

National Institute of Education, Nanyang Technological University, Singapore

Abstract
This is an exploratory study into the individual problem posing characteristics of 152 Grade 9
students (aged 15) from four secondary schools in Singapore. The subjects were novice problem
posers in that they were not given any training in problem posing skills. Each student was asked
to write down a problem for their friends with the final answer as 600. Students also solved their
own problems. The relationship between the structures of the posed problems, the topics
involved in the problems and the solutions were discussed. Students’ self-reported metacognitive
regulatory strategies, the effects of achievement levels and of gender were also discussed. It was
found that direct proposition type of problems occurred in about half of the posed problems. The
presence of problem over-conditioning was not significant across achievement levels and gender.
Students’ confidence in their posed problems were found to be related to some of the
metacognitive strategies at the property noticing phase, problem construction stage and during
solution checking.
Keywords: Mathematical Problem Posing, Problem Solving, Angle Measure, Metacognition

The importance of problem posing in relation to mathematical explorations has been highlighted
in various literatures. There are studies linking mathematical problem posing to creativity (Silver,
1994, Haylock, 1987) and to mathematical competence (Ellerton, 1986). Specifically the
relationship between mathematical problem posing performance and problem solving abilities has
also been studied in recent years. In fact, English (1997) noted that both problem posing and
problem solving are closely related and that the process of problem posing in fact draws heavily
on the processes of problem solving.
Silver and Cai (1996) made a distinction between two notions of problem posing. Firstly,
problem posing can be construed to be a case of a generation of new problems from a
mathematical situation. Secondly, it can also be interpreted as the reformulation of a given
problem in which there is an intention to uncover the deeper underlying structures of a given
question or problem. In this case, it is one strategy in problem solving where the solver tries to
answer related questions which will give insights to the original problem. For the present work,
the focus will be on the nature of problem posing itself and not as part of a problem solving
heuristic. Problem posing is used here as the formulation of new problems from a mathematical
stimulus.
The inclusion of activities in which students generate their own problems had also been strongly
endorsed by the National Council of Teachers of Mathematics (1991). It is believed that such
activities can provide a glimpse of students’ understanding of mathematical concepts and
processes and their attitudes towards problem solving. As part of the metacognitive aspect of the
national mathematics curriculum framework, problem posing is also strongly encouraged in the

36
classroom. In the Ministry of Education document, Mathematics Syllabus (Lower Secondary)
(CPDD, 2001), students are encouraged to “create, formulate or extend problems.” (p.16)
One strand of the studies in problem posing involves developing problem posing as an
instructional intervention to improve problem solving skills and to improve disposition towards
solving. Some of these include work done by Gonzales (1994, 1998) on using problem posing to
improve on preservice teacher training and Manouchehri (2001), who worked on an instructional
model for promoting problem posing in a sixth grade classroom. Another strand of work goes
into analyzing the problems posed in terms of their surface structures. For example, Marshall’s
schema theory (Marshall, 1995) was used in the study by Charalambous, Kyriakides and
Philippou (2003) on the problem posing skills of primary school students. For both strands, the
contexts of these studies mainly involved mathematical word problems and largely on arithmetic.
Subjects of these studies varied between students of various grade levels to undergraduates in pre-
service teacher preparation courses.
Lesser work is being done in the area on the cognitive processes of mathematical problem posing
itself and the regulation of these processes. Christou, Mousoulides, Pittalis, Pantazi and Sriraman
(2005) had proposed in their study on 143 Grade 6 students in Cyprus, a few processes that can
be used to describe problem posing. Selecting quantitative information is one of the processes
involved in posing problems. It is mostly linked to tasks that require students to pose problems
that are appropriate to specific given answers. Such a process involves the ability to focus on the
context of the problem structure and the relationship between the given initial information and the
subsequent information that the students created to make the posed problems coherent. This is an
important skill in building connections across domains of knowledge and sense making in
mathematical exploration. One purpose of this study is to look at the selecting process involved in
problem posing in the area of school geometry in Singapore. Students’ posed problems in this
area will shed light into how they perceive linkages between the different topics within school
geometry and into how they construct their problems. The control of this cognitive process is
also an important aspect in the study of problem posing.
Livingston (2003) referred to metacognitive regulatory processes as those that one uses to control
cognitive activities and more importantly to see to the meeting of the cognitive goal. These
involve planning, monitoring of cognitive activities and checking of outcomes of those activities.
The other purpose of this study is to illuminate some of these regulatory processes that are
involved in problem posing.
Method
a) Subjects
This is an exploratory study about the individual problem posing characteristics of 152 Grade 9
(aged 15) students from four secondary schools in Singapore. The subjects were novice problem
posers. Besides their classroom experience in asking questions, they were not given any specific
training in problem posing prior to this study. The decision to locate the study with these students
was that few such studies had been made in this area in Singapore.
b) Task
Each student was asked to freely write down a problem for his or her friends to solve with the
final answer as 600. Students also solved the problems they had posed. By solving their own
problems, the students can make explicit the selecting process as they construct the problem
structures. Silver (1990) in her study on problem posing involving number sense, also noted that

37
such open ended stimulus tends to provide good opportunities for students to be engaged in
generative aspects of mathematical thinking.
Much of what constitutes a problem is dependent on the context in which the problem is posed.
When students posed problems to friends, they do so with some perceived knowledge of their
friends’ familiarity of topics and methods, cognizance of common errors made by friends and the
time taken for their friends to complete the tasks. Such perceived knowledge is captured in
students’ posed problems.
c) Questionnaire
Immediately after completing the task, students were asked to complete an 18-item questionnaire
as shown in Table 1. The purpose of the questionnaire is to get a snapshot of their metacognitive
regulatory strategies during their posing and solving. Each item has a 4 point Likert scale with 1
being strongly disagree and 4 being strongly agree. This instrument is an adaptation of Goos,
Galbraith and Renshaw (2000) metacognitive survey for secondary students in the Australian
state of Queensland in their study on the metacognitive aspects of students solving combinatorics
problems. In order to make the questionnaire more appropriate for the students in this study,
some of the questions were changed. Further modifications had also been made to take into
account the different phases of metacognitive regulatory behaviour in problem posing. These
phases were the results of an earlier work made by the authors as they worked on the think-aloud
Table 1
Metacognitive Regulatory Strategies

Code Statements Phase

4A I read carefully when there is an important information PN
6A I am good at recalling what my teacher had taught PN
8A I ask myself questions about the information before I begin PN
9 I use my own examples to make what is given more meaningful PN
13A I try to use my own words when I read the new information PN
1A I check periodically if I am getting the problem that I want PC
2AA I consider other possibilities to a problem before I ask it PC
10AA I find myself checking on my understanding as I posed the problem PC
11A I draw diagrams to help me understand while posing the problem PC
17A I think about the method of solution first before I pose the problem PC
5A I ask if I have considered all possibilities to my problem while solving it CS
14A I go through over new information that is not clear CS
15AA I constantly look back at the problem as I start the solution CS
16A I check my solution as I worked on it CS
3A I know how well I have done once I finish the problem LB
7A I ask if there was an easier way to pose after I finish posing the problem LB
12A I ask if I could have posed a different problem after I finished LB
18A I like the problem that I posed LB
PN: Property Noticing, PC: Problem Construction, CS: Checking Solution, LB: Looking Back

38
protocols of 10 students prior to this study. Students were then engaged in the same task of
creating problems with the final answer as 600. Property noticing describes the initial phase
before students start the active construction of their problems. In this phase, students make
associations with the topics that first come to their mind when confronted with this stimulus.
Within the problem construction phase, students draw upon their earlier experiences about topics
to come up with problems. Simultaneously they are also checking the solutions to their posed
problems and retrospectively going back to their earlier posed problems (and modifying when
necessary) and see if their solutions make sense. In the last phase, students reflect back and
evaluate their work.
One limitation to this study is that the sample of 152 students can not meant to be representative
of the students in all the secondary schools in Singapore. The sample size also does not allow for
factor analysis of the metacognitive regulatory strategies.

Results
The results of analysis of students’ posed problems, their solutions and the questionnaire
responses are presented in two parts. In the first part, problems posed are described in terms of
the types of problem structures, the domains of knowledge used and the solutions to the problems.
Secondly, discussions are made on students’ self-reported metacognitive regulatory strategies.
Specifically the relationship between the different strategies used and the types of posed problems,
students’ gender and students’ achievement levels are also discussed. All results are discussed at
5% level of significance.
a) Characteristics of Posed Problems
i) Problem structure
There are direct proposition problems where the solutions require single-step solutions as
shown in Figure 1. Each of the direct proposition problems involves a single topic. These
problems account for 50.7% of the total posed problems. Their solutions involved some
forms imitative reasoning. For example, the solution may involve the recalling of a
Figure 1
Example of a direct proposition problem

simple algorithm like finding the angle sum in a triangle as in Figure 1 or a possible
memorized answer like cos-1(0.5) = 600 to solve “what is x if cos x = 0.5?” These may
suggest that students just created them from what first came into their mind without
trying to create linkages with other topics. The other possibility is that these are problems
which students perceived their friends are able to solve. They also reflect what can be

39
commonly found as exercises in school textbooks or perhaps problems which they
commonly encountered in their classroom learning experiences.
Figure 2
Example of a multiple topic problem

The rest of the problems involve multiple steps and are situated in a combination of
topics. One example is shown in Figure 2. These questions are good examples of how
students were able to link topics together in their problem construction. In the problem in
Figure 2, the student involved the uses of the geometric properties of a parallelogram, a
trapezium and a triangle. Justifications of the steps were also made by using of the
properties of alternate angles and the angle sum of a triangle.
Within these non-direct propositions, there are problems which are over-conditioned. An
over-conditioned problem contains extraneous information which does not contribute to
the solution. In Figure 3, angle 400 is not needed as part of the solution. This occurs in
Figure 3
Example of an over-conditioned problem

20.4% of the total posed problems. Yet there are other problems that contain
inconsistencies in their structures. The angle sum of a triangle is violated in Figure 4.

40
 The problem was constructed without considering the linkage to the other parts of the
diagram. The student perhaps was trying to impress his or her friends the sophistication
the problem by including more information. Of the posed problems, 17.1% of them are
inconsistent problems.

Figure 4
Example of an inconsistent problem

ii) Domains of Knowledge

The problem in Figure 1 involves a single topic about the sum of angles in a right angle
triangle. For single topic problems, students also made use of the topics on alternate
angles, complementary/supplementary angles, corresponding angles and trigonometric
ratios. Multiple topics, namely, the use of the properties of angles in a triangle and
alternate angles are found in the problem in Figure 2. Across the four schools, students
used a variety of topics in school geometry in their posed problems. Of these topics, the
use of angle sum in a triangle is most prevalent. It is observed in 43.4% of the work with
others using the topics about complementary/supplementary angles (24.3%), circle
properties (21.1%) and alternate angles (15.1%).
As shown in Table 2, the use of the topic on triangle is also prevalent among problems
that involve multiple topics (64.2%) rather than in single topic problems (27.1%).
Perhaps students are more familiar with this topic. Over-conditioning is also not
significantly associated to problems with multiple topics, χ 2 (1) = 0.588, p = .366.

Table 2
Analysis of posed problems involving multiple topics
Multiple Topics (P10) Asymp Sig.
χ 2 (1)
Absent Present Total (2 sided)
Use of Angle Sum Absent 62 (72.9%) 24 (35.8%) 86
21.01 .000
in Triangle (K3) Present 23 (27.1%) 43 (64.2%) 66
Use of Circle (K5) Absent 78 (91.8%) 42 (62.7%) 120
19.06 .000
Present 7 (8.2%) 25 (37.3%) 32

iii) Solution
There are varying numbers of steps in students’ solutions to their posed problems. They
range from the single step direct proposition type to more involved types like the problem
in Figure 2. In the course of working through their solutions, students demonstrated they

41
 were able to justify their steps appropriately. Most of the solutions involve the recall of
solution algorithms in school geometry and trigonometry. Except for four solutions that
have computational errors, the rest of the students’ solutions are found to be correct.
b) Achievement and Gender
Across the four schools, students’ scores in the standardized national Primary School Leaving
Examination at Grade 6 are used to classify the achievement levels. Students are classified either
as High Achievers (HA), Average Achievers (AV) or Low Achievers (LA). The distribution of
gender and achievement levels is shown in Table 3.
Table 3
Distribution of Gender and Achievement Levels
Male Female Subtotal
LA 11 29 40
AV 31 41 72
HA 17 23 40
Subtotal 59 93 152

The students’ achievement levels are found to be not significantly associated to the types of posed
problems, domains of knowledge and their solutions. For example, there is no strong evidence to
suggest that HA students produce more multiple topics type of questions compared to students in
the other levels. In Table 4, achievement levels and the presence of over-conditioned problems
are also found not to be significantly related in this study. This suggests that the problem posing
performance of students to the given open-ended stimulus is not strongly influenced by how well
they had performed in their standardized tests.
Across the achievement levels, there are also no significant associations with the students’
metacognitive regulatory strategies except for the presence of checking during the solution phase.
There is some association between HA students and their self-declared use of checking in their
solutions. Lower number of HA students checked their solutions compared to the other two
groups of students. This perhaps reflects the HA students’ confidence in their solutions to their
posed problems and hence the lesser need for checking.
Table 4
Achievement Profiles, Over-Conditioning and Use of Checking

Achievement Profiles
χ 2 (2) Asymp Sig.
LA AV HA Total (2 sided)

Absent 31 (77.5%) 54 (75.0%) 36 (90.0%) 121

Over 3.71 .156
Conditioning (P8) Present 9 (22.5%) 18 (25.0%) 4 (10.0%) 31

Use of Checking No 17 (42.5%) 43 (59.7%) 29 (72.5%) 89

in Solution (CS) 7.49 .024
Yes 23 (57.5%) 29 (40.3%) 11 (27.5%) 63
(Q16A)

Problem structures are also found not to be significant in discussing gender. Like in achievement
levels, the over-conditioning feature in problems is also found not to be significant across gender.
Table 5 shows some significant results from the questionnaire survey and gender. At property
noticing phase, more females than males agree that they were not good at recalling what the

42
teachers had taught. For both gender, close to half of the students (46.7%) felt that they were
good at recalling what had been taught.
More males reported that they had considered other possibilities to the posed problems as they
were solving the problems. But most of them (69.5%) did not check their solutions. More males
than females are also found to like the problems they had posed compared to females. For all the
students, 88 (57.9%) reported they like their posed problems.
Table 5
Questionnaire Responses and Gender

Gender Asymp Sig.

χ 2 (1) (2 sided)
Female Male Total
Good at recall (PN) No 56 (60.2%) 26 (44.1%) 82
(Q6A) 3.788 .038
Yes 37 (39.8%) 33 (55.9%) 70
Think of method of No 46 (49.5%) 18 (30.5%) 64
solution first (PC) 5.320 .016
(Q17A)
Yes 47 (50.5%) 41 (69.5%) 88

Consider all possibilities

No 45 (48.4%) 18 (30.5%) 63
4.755 .022
to problem (CS) (Q5A) Yes 48 (51.6%) 41 (69.5%) 89

Use of Checking in
No 48 (51.6%) 41 (69.5%) 89
4.76 .029
Solution (CS) (Q16A) Yes 45 (48.4%) 18 (30.5%) 63

Like posed questions

No 46 (49.5%) 18 (30.5%) 64
5.32 .021
(LB) (Q18A) Yes 47 (50.5%) 41 (69.5%) 88

From Table 6, among students who reflected that they knew how well they had done once they
had finished posing their problems, 87.8% of them checked their solutions and 86.7% drew
diagrams to help them understand as they constructed their problems. Similarly, of those who felt
that they had done well, a high number also reported that they asked questions about the
information during the property noticing phase of their problem posing. The strategy of asking
questions at the property noticing phase, drawing diagrams during their problem construction
phase and checking of their solutions appears to account for the higher confidence in their
knowing of how well they have done in their problem posing.
Table 6
Questionnaire Responses with Knowing How Well When Done

Know how well when done (Q3A) Asymp Sig.

χ 2 (1) (2 sided)
No Yes Total
Ask questions about No 30 (55.6%) 34 (34.7%) 64
information (PN) (Q8A) 6.216 .010
Yes 24 (44.4%) 64 (65.3%) 88
Draw diagrams to help No 20 (37.0%) 13 (13.3%) 33
understand (PC) (Q11A) 11.575 .001
Yes 34 (63.0%) 85 (86.7%) 119
Use of Checking in No 31 (57.4%) 12 (12.2%) 43
Solution (CS) (Q16A) 35.005 .000
Yes 23 (42.6%) 86 (87.8%) 109

43
 Conclusion and Implications
To encourage a variety of problem structures, the classroom teacher needs to broaden the types of
problem experiences being presented to students. The teacher can capitalize on the informal
activities situated in students’ daily activities and get students to the habit of recognizing
mathematical situations wherever they might be and making connections to various aspects of
school geometry. Otherwise, students would only be comfortable with constructing direct
proposition type of problems which does not allow them to explore the inter-connectedness of
topics. This ability to make connections is an important skill in mathematical exploration.
Perhaps, to get students to have more confidence in their problem posing, the teacher can
encourage students to ask more questions about the given stimulus during the property noticing
phase, to draw diagrams during the problem construction phase and to check their solutions.
These metacognitive strategies appear to help novice problem posers in this study to have more
confidence in their work.
Since achievement levels and gender are not significant across problem posing performance,
classroom problem posing activities should be encouraged for all students. The teacher can also
make use of students’ problem posing work as teaching points. For example, the teacher in
discussing students’ posed problems, can sensitize the class to issues about the inconsistencies in
problem structures or to the notion of over-conditioning in problem construction. Perhaps such
discussions may help to produce better problem posers and may contribute to students’
engagement in more quality mathematical inquiry in the classroom.
The teacher in teaching is also involved in posing problems. The school curriculum planner can
look into ways of promoting the teacher’s competency in problem posing just like the way
problem solving heuristics are made known to teachers. The very way in which the teacher asks
questions can affect that shared spirit of investigation between the teacher and the students.
Appropriate use of varied problem types which may depart from the textbook exercises may bring
about a better quality of classroom interaction. But such teacher’s behaviour is also dependent on
the teacher’s beliefs and perceptions about problem posing itself and about the teachers’ views on
the nature of mathematics. This is an issue that warrants further investigation.

References
Christou, C., Mousoulides, N., Pittalis, M., Pantazi, D.P., & Sriraman, B. (2005). An empirical
taxonomy of problem posing processes. Zentralblatt f! r Didaktik der Mathematik, 37(3),
Retrieved September 21, 2007, from
http://www.umt.edu/math/reports/sriraman/Int_Reviews_Preprint_Cyprus_Sriraman.pdf

Charalambos, C., Kyriakides, L., & Philippou, G. (2003). Testing a comprehensive

model for measuring problem solving and problem posing skills of primary
pupils. In N.A. Pateman, B.J. Dougherty & J. Zilliox (Eds.), Proceedings of the
27th PME, vol 2 (pp. 205-212). Honolulu : PME.
Ellerton, N. F. (1986). Children’s made-up problems: A new perspective on talented
mathematicians. Educational Studies in Mathematics, 17, 261-271.
English, L. D. (1997). Promoting a problem-posing classroom. Teaching Children
Mathematics, 4, 172-181.

44
Gonzales, N.A. (1994). Problem Posing: A neglected component in mathematics courses
for prospective elementary and middle school teachers. School Science and
Mathematics, 94(2), 78-84.
Gonzales, N. A. (1998). A blueprint for problem posing. School Science and
Mathematics, 98(8), 448-456.
Goos, M., Galbraith, P., & Renshaw, P. (2000). A money problem: a source of insight
into problem solving action. Retrieved September 5, 2007, from
www.cimt.plymouth.ac.uk/journal/pgmoney.pdf
Haylock, D.W. (1987). A framework for assessing mathematical creativity in school
children. Educational Studies in Mathematics, 18, 59-74.
Livingston, J.A. (2003). Metacognition: an overview. Maryland: NY: ERIC Clearinghouse No.
TM034808. (ERIC Document Reproduction Service ED 474 273)
Manouchehri, A. (2001) A four-point instructional model. Teaching Children
Mathematics, 8, 180-186.
Marshall, S.P. (1995). Schemas in problem solving, (pp.62-112). New York: Cambridge
University Press.
Ministry of Education of Singapore. (2001). Mathematics syllabus (Lower
Secondary), (pp.16-92). Singapore: Curriculum Planning and Development
Division.
National Council of Teachers of Mathematics (1991). Professional standards for teaching
mathematics. Reston, VA : NCTM.
Silver, E. A. (1990). Contributions of research to practice: Applying findings, methods
and perspectives. In Cooney, T.J., Hirsch, C.R. (Eds.), Teaching and learning
mathematics in the 1990s, (pp. 1-11). Reston, VA: NCTM.
Silver, E. A. (1994). On mathematical problem posing. For the Learning of
Mathematics, 14(1), 19-28.
Silver, E. A., & Cai, J. (1996). An analysis of arithmetic problem posing by middle
school students. Journal of Research for Mathematics Education, 27(5), 521-
539.

45
 FUTURE DIRECTIONS AND PERSPECTIVES FOR PROBLEM
SOLVING RESEARCH AND CURRICULUM DEVELOPMENT

Lyn English Richard Lesh Thomas Fennewald

Queensland University of Technology Indiana University Indiana University
l.english@qut.edu.au ralesh@indiana.edu tfennewa@indiana.edu

ABSTRACT
Since the 1960s, numerous studies on problem solving have revealed the complexity of the
domain and the difficulty in translating research findings into practice. The literature
suggests that the impact of problem solving research on the mathematics curriculum has
been limited. Furthermore, our accumulation of knowledge on the teaching of problem
solving is lagging. In this first discussion paper we initially present a sketch of 50 years of
research on mathematical problem solving. We then consider some factors that have held
back problem solving research over the past decades and offer some directions for how we
might advance the field. We stress the urgent need to take into account the nature of problem
solving in various arenas of today’s world and to accordingly modernize our perspectives on
the teaching and learning of problem solving and of mathematical content through problem
solving. Substantive theory development is also long overdue—we show how new
perspectives on the development of problem solving expertise can contribute to theory
development in guiding the design of worthwhile learning activities. In particular, we explore
a models and modeling perspective as an alternative to existing views on problem solving.

INTRODUCTION
Research on mathematical problem solving has received a good deal of attention in past
decades. Among the notable developments have been Polya’s (1945) seminal work on how to
solve problems, studies on expert problem solvers (e.g., Anderson, Boyle, & Reiser, 1985),
research on teaching problem solving strategies, and heuristics and fostering metacognitive
processes (e.g., Charles & Silver, 1988; Lester, Garofalo, & Kroll, 1989), and, more recently,
studies on mathematical modeling (e.g., Lesh, in press; English, 2007). Existing, long-
standing perspectives on problem solving have treated it as an isolated topic, where problem
solving abilities are assumed to develop through initial learning of concepts and procedures
followed by practice on “story problems,” then through exposure to a range of strategies (e.g.,
“draw a diagram,” “guess and check”), and finally, through experiences in applying these
competencies to solving “novel” or “non-routine problems.” As we discuss later, when taught
in this way, problem solving is seen as independent of, and isolated from, the development of
core mathematical ideas, understandings, and processes. Despite these decades of research
and associated curriculum development, it seems that students’ problem solving abilities still
require substantial improvement especially given the rapidly changing nature of today’s
world (Kuehner & Mauch, 2006; Lesh & Zawojewski, 2007; Lester & Kehle, 2003).
This current state of affairs has not been helped by the noticeable decline in the amount of
problem solving research that has been conducted in the past decade. A number of factors
have been identified as contributing to this decline. These include the discouraging cyclic
trends in educational policy and practices, limited research on concept development and

46
problem solving, insufficient knowledge of students’ problem solving beyond the classroom,
the changing nature of the types of problem solving and mathematical thinking needed
beyond school, and the lack of accumulation of problem solving research (Lesh &
Zawojewski, 2007). Before considering each of these contributing factors, we offer an
overview of research on mathematical problem solving over the past 50 years.

A BRIEF SKETCH OF FIFTY YEARS OF RESEARCH ON MATHEMATICAL

PROBLEM SOLVING
In mathematics education, research on problem solving has focused primarily on word
problems of the type emphasized in school textbooks or tests – where “problems” are
characterized as activities that involve getting from givens to goals when the path is not
obvious. With such situations in mind, Polya’s book How to Solve It (1945) introduced the
notion of heuristics – such as draw a picture, work backwards, look for a similar problem, or
identify the givens and goals (later referred to as strategies by mathematics educators) –
which mathematics education researchers immediately recognized to be useful for generating
after-the-fact descriptions of past behaviors for many expert problem solvers. But, even for
less experienced problem solvers, these same heuristics also were expected to provide useful
answers to the question: “What should I do when I’m stuck?”
Unfortunately, for reasons we describe briefly in this, our first of two papers for ICME 11,
the past 50 years of research have not provided validation for these latter expectations.
Nonetheless, some hope remains! Most past research has leaped ahead to investigate the
questions: (a) Can Polya-style heuristics be taught? (b) Do learned heuristics/strategies have
positive impacts on students’ competencies? There exists almost no research that has
provided useful operational definitions to answer more fundamental questions such as: (a)
What does it mean to “understand” Polya-style heuristics? (b) How (and in what ways) do
these understandings develop? (c) What is the nature of primitive levels of development? (d)
How can development be reliably observed, documented, and measured (or assessed)? Until
researchers develop useful responses to these latter two questions, it is not reasonable to
expect significant progress to be made on the former two questions.
In spite of the apparent face validity of Polya’s heuristics, Begel’s (1979) comprehensive
review of the research literature in mathematics education concluded that there was little
evidence to support the claim that general processes that experts use to describe their past
problem solving behaviors also should provide prescriptions to guide novices’ next-steps.
Similarly, Silver’s (1985) assessment of the literature on problem solving concluded that,
even in studies where some successful learning has been reported, transfer of learning has
been unimpressive. Furthermore, successes generally occurred only when world-class
teachers taught long and complex courses in which the size and complexity of the
“treatments” made it unclear why performance improved. Perhaps, suggested Silver,
improvements in problem solving performance simply resulted from students learning
relevant mathematics concepts - rather than from learning problem solving strategies,
heuristics, or problem solving processes!
Similar conclusions again were stated in the NCTM’s 1992 Handbook for Research on
Mathematics Teaching and Learning (Grouws, 1992), where Schoenfeld’s (1992) chapter on
problem solving concluded that attempts to teach students to use Polya-style heuristics and
processes generally had not proven to be successful. However, Schoenfeld went on to suggest
that one reason for this lack of success might be because many of Polya’s heuristics appear to
be descriptive but not prescriptive. That is, most are really just names for large categories of

47
processes rather than being well defined processes in themselves. Therefore, in an attempt to
go beyond “descriptive power” to achieve “prescriptive power,” Schoenfeld suggested that
problem solving research and teaching should: (a) Help students develop larger numbers of
more specific problem solving strategies that link more clearly to specific classes of
problems, (b) Teach metacognitive strategies1 so that students learn when to use their
problem solving strategies and content knowledge, and (c) Develop ways to improve
students’ beliefs about the nature of mathematics, problem solving, and their own personal
competencies.
Unfortunately, ten years after Schoenfeld’s proposals were made, Lester and Koehle (2003)
again reviewed the literature and again concluded that research on problem solving still had
little to offer to school practice. One explanation for this lack of success appeared to be
because Schoenfeld’s proposal simply moved the basic shortcoming of Polya’s heuristics to a
higher level. That is, regardless of whether attention focuses on Polya-style heuristics or on
Shoenfield-style metacognitive processes or beliefs, short lists of descriptive processes or
rules tend to be too general to have prescriptive power. Yet, longer lists of prescriptive
processes or rules tend to become so numerous that knowing when to use them becomes the
heart of what it means to understand them. This shortcoming tends to be exacerbated by the
fact that regardless of whether attention focuses on Polya’s heuristics or on Schoenfeld’s
metacognitive processes or beliefs, virtually all such processes and rules are
counterproductive in some situations. For example, even the seemingly-sensible admonition
for students to carefully plan-monitor-assess their work tends to be explicitly set aside during
periods of productive “brainstorming” during initial stages of solving complex problems. In
fact, the defining characteristic of brainstorming is that problem solvers are supposed to
rapidly generate a diverse collection of ideas – by temporarily avoiding criticism, assessment,
and concerns about long-range implications. So, knowing when and why to use such
techniques emerges as one of the most important parts of what it means to understand them.
In response to such conclusions about the state of problem solving research, Schoenfeld’s
plenary address for the 2007 NCTM Research Pre-session proposed another embellishment
of his same basic theory. The heart of his recommendation was that researchers should focus
on something that we might call meta-meta-cognitive processes – or rules which are expected
to operate on lower-level metacognitive processes, heuristics, strategies, knowledge, or skills.
But again, just as in the case of the beliefs and metacognitive processes that Schoenfeld
proposed fifteen years earlier, meta-meta-cognitive processes were described as being
explicitly executable rules (e.g., cost-benefit rules which operate on lower-level rules).
Consequently, it is unclear why meta-metacognitive rules should be expected to avoid the
same shortcomings that were associated with past notions of heuristics, strategies, or meta-
cognitive processes. That is, short lists of descriptive rules lack prescriptive power, and
longer lists of prescriptive rules involve knowing when and why to use them.
Lesh and Zawojewski (2007) concluded that, when a field of research has experienced more
than 50 years of failure using continuous embellishments of rule-governed conceptions of
problem solving competence, perhaps the time has come to consider other options – and to
re-examine foundation-level assumptions about what it means to understand mathematics
concepts and problem solving processes. In particular, it is time to re-examine foundation-
level assumptions about what it means to understand a small number of big ideas in
elementary mathematics. One alternative is to use theoretical perspectives and accompanying
research methodologies that we call models & modeling perspectives (MMP) on mathematics
problem solving, learning, and teaching (Lesh & Doerr, 2003). But, before describing

1
Metacognitive processes are processes that operate of lower-order knowledge or abilities.

48
relevant aspects of MMP, we briefly identify some of the major reasons why past problem
solving research has produced so little success.

LIMITING FACTORS IN ROBLEM-SOLVING RESEARCH

Pendulum Swings Fuelled by High-Stakes Testing
Over the past several decades, we have seen numerous cycles of pendulum swings between a
focus on problem solving and a focus on “basic skills” in school curricula. These
approximately 10-year cycles, especially prevalent in the USA but also evident in other
nations, appear to have brought few knowledge gains with respect to problem solving
development from one cycle to the next (Lesh & Zawojewski, 2007). Over the past decade or
so, many nations have experienced strong moves back towards curricula materials that have
emphasized basic skills. These moves have been fuelled by high-stakes national and
international mathematics testing, such as PISA (Programme for International Student
Assessment: http://www.pisa.oecd.org/) and TIMMS (Third International Mathematics and
Science Study: http://timss.bc.edu/timss2003i/intl_reports.html).
These test results have led many nations to question the substance of their school
mathematics curricula. Indeed, the strong desire to lead the world in student achievement has
led several nations to mimic curricula programs from those nations that score highly on the
tests, without well-formulated plans for meeting the specific needs of their student and
teacher populations (Sriraman & Adrian, 2008). This teaching-for-the test has led to a “New
Push for the Basics” as reported in the New York Times, November 14, 2006. Unfortunately,
these new basics are not the basics needed for future success in the world beyond school, as
we indicate later. With this emphasis on basic skills, at the expense of genuine real-world
problem solving, the number of articles on research in problem solving has declined. What is
needed is research that explores students’ concept and skill development as it occurs through
problem solving.
Limited Research on Concept Development and Problem Solving
As we discuss in our second paper, relationships are unclear between concept development
and the development of problem solving competencies (Lester & Charles, 2003; Schoen &
Charles, 2003). One shortcoming of past problem solving research is that it has not been clear
how concept development is expected to interact with the development of relevant problem
solving heuristics, beliefs, dispositions, or processes. In fact, in many curriculum standards
documents (e.g., NCTM, 2000, 2008 http://standards.nctm.org/document/chapter3/prob.htm),
problem solving tends to be listed as the name of a chapter-like topic similar to algebra,
geometry, or calculus. In other words, the implicit assumption is conveyed that problem
solving ability is expected to increase by: (a) first, mastering relevant concepts, (b) second,
mastering relevant problem solving heuristics, strategies, beliefs, dispositions, or processes,
and (c) third, learning to put these concepts and processes together to solve problems.
Consequently, when such assumptions are coupled with the flawed belief that students must
first learn concepts and processes as abstractions before they can put them together and use
them in “real-life” problem solving situations, problem solving tends to end up never getting
taught at all in many classrooms. So, one of the most critical challenges for future problem
solving research is to clarify the nature of relationships that should exist between concept
development and the development of problem solving competencies.
Limited Knowledge of Students’ Problem Solving Beyond the Classroom

49
As we have highlighted, problem solving is a complex endeavor involving, among others,
mathematical content, strategies, thinking and reasoning processes, dispositions, beliefs,
emotions, and contextual factors. Future studies of problem solving need to embrace the
complexity of problem solving as it occurs in school and beyond, as we discuss later.
However, to date, most research on problem solving has not really addressed students'
problem solving capabilities beyond the classroom—we need to know why students have
difficulties in applying the mathematical concepts and abilities (that they presumably have
learned in school) outside of school—or in other classes such as those in the sciences. To
assist us here we need more interdisciplinary problem solving experiences that mirror
problem solving beyond the classroom (English, in press). For example, experiences that
draw upon the broad field of engineering provide powerful links between the classroom and
the real world, enabling students to apply their mathematics and science learning to the
solution of authentic problems (Kuehner & Mauch, 2006).
Changing Nature of the Types of Problem Solving and Mathematical Thinking needed
beyond School
Today, experts outside of schools consistently emphasize that new technologies for
communication, collaboration, and conceptualization have led to significant changes in the
kinds of mathematical thinking that are needed beyond school—and to significant changes in
the kinds of problem solving situations in which some form of mathematical thinking is
needed. For example, in just a few decades, the application of mathematical modeling to real-
world problems has escalated. Traffic jams are modeled and used in traffic reports; the
placement of cell-phone towers is based on mathematical models involving 3-D topography
of the earth; and the development of internet search engines is based on different
mathematical models designed to find new and more efficient ways to conduct searches.
Unfortunately, the types of problems students meet in the classroom are often far removed
from reality—we need to redress this state of affairs as we consider fresh perspectives on
problem solving in the curriculum.
Research on problem solving beyond school also suggests that, although professionals in
mathematics-related fields draw upon their school learning, they do so in a flexible and
creative manner, unlike the way in which they experienced mathematics in their school days
(Gainsburg, 2006; Hall, 1999; Hamilton, 2007; Noss, Hoyles, & Pozzi, 2002; Zawojewski &
McCarthy, 2007). Furthermore, problem solvers beyond the classroom often are not isolated
individuals but instead are teams of diverse specialists (Hutchins, 1995a, 1995b; Sawyer,
2007). These specialists often offload important aspects of their thinking using powerful
technology-based tools which make some functions easier (such as information storage,
retrieval, representation, or transformation) but which make others far more complex and
difficult (such as interpretation and communication). So, relevant knowledge and abilities
tend to be distributed across a variety of tools, and across individuals within groups. Critical
abilities often are those associated with the mathematics of description, explanation, and
communication at least as much as the mathematics of computation and deduction (Lesh,
Middleton, Caylor & Gupta, 2008), and progress tends to resemble the evolution of a
community or interacting organisms – rather than movement along a path (Lesh & Yoon,
2004). Unfortunately, research on mathematical problem solving has not kept pace with the
rapid changes in the mathematics and problem solving needed beyond school.
Lack of Accumulation of Problem solving Research
As we also discuss in our second paper, there has been a lack of accumulation of problem
solving research. Failed or flawed concepts or conjectures have continued to be recycled or
embellished – with no significant changes being made in the underlying theoretical

50
perspectives. Mathematics education researchers have generally avoided tasks that involve
developing critical tools for their own use. Unlike their counterparts in more mature sciences
(physics, chemistry, biology), where some of the most significant kinds of research often
involve the development of tools to reliably observe, document, or measure the most
important constructs, mathematics educators have developed very few tools for observing,
documenting, or measuring most of the understandings and abilities that are believed to
contribute to problem solving expertise. We return to this concern later in this paper.
Furthermore, partly because operational definitions and tools have not been developed for
most constructs that have been considered important in problem solving development, there
is a tendency to repeatedly elaborate on or recycle apparently failed or flawed concepts. For
example, the use of Polya-style heuristics, problem solving strategies, and various
metacognitive and meta-metacognitive processes (Schoenfeld, 2007) is an example of
continuous embellishment of a theory that focuses on explicitly learned rules. In our second
paper, we extend our discussion on theory development and explore alternative research
methodologies for advancing the field.

ADVANCING THE FIELD OF PROBLEM SOLVING RESEARCH AND

CURRICULUM DEVELOPMENT
Although we have highlighted some of the issues that have plagued problem solving research,
there are emerging signs that the situation is starting to improve. We believe the pendulum of
change is beginning to swing back towards problem solving on an international level,
providing impetus for new perspectives on the nature of problem solving and its role in
school mathematics (Lester & Kehle, 2003). For example, a number of Asian countries have
recognized the importance of a prosperous knowledge economy and have been moving their
curricular focus toward mathematical problem solving, critical thinking, creativity and
innovation, and technological advances (e.g., Maclean, 2001; Tan, 2002). In refocusing our
attention on problem solving and how it might become an integral component of the
curriculum rather than a separate, often neglected, topic we explore the following issues:
• What is the nature of problem solving in various arenas of today’s world?
• What future-oriented perspectives are needed on the teaching and learning of problem
solving including a focus on mathematical content development through problem
solving?
• How can studies of problem solving expertise contribute to theory development that
might guide the design of worthwhile learning experiences?
• Why is a models and modeling perspective a powerful alternative to existing views on
problem solving?
The Nature of Problem Solving in Today’s World
Concerns have been expressed by numerous researchers and employer groups that schools are
not giving adequate attention to the understandings and abilities that are needed for success
beyond school. For example, potential employees most in demand in mathematics/science
related fields are those that can (a) interpret and work effectively with complex systems, (b)
function efficiently and communicate meaningfully within diverse teams of specialists, (c)
plan, monitor, and assess progress within complex, multi-stage projects, and (d) adapt quickly
to continually developing technologies (Lesh, in press). Research indicates that such
employees draw effectively on interdisciplinary knowledge in solving problems and
communicating their findings. Furthermore, although they draw upon their school learning,
these employees do so in a flexible and creative manner, often creating or reconstituting

51
mathematical knowledge to suit the problem situation, unlike the way in which they
experienced mathematics in their school days (Gainsburg, 2006; Hamilton, 2007; Lesh, in
press; Zawojewski & McCarthy, 2007). In fact, these employees might not even recognize
the relationship between the mathematics they learned in school and the mathematics they
apply in solving the problems of their daily work activities.
Identifying and understanding the differences between school mathematics and the work-
place is critical in formulating a new perspective on problem solving. As we address later,
one of the notable findings of studies of problem solving beyond the classroom is the need to
master mathematical modeling. Many new fields, such as nanotechnology, need employees
who can construct basic yet powerful constructs and conceptual systems to solve the
increasingly complex problems that confront them. Being able to adapt previously
constructed mathematical models to solve emerging problems is a critical component here.
Future-Oriented Perspectives on the Teaching and Learning of Problem Solving
We have argued that future-oriented perspectives on problem solving should transcend
current school curricula and national standards and should draw upon a wider range of
research across disciplines (English, 2008; Lesh, in press). Most research on problem solving
has commenced with the assumption that the researchers already possess clear and accurate
understandings about what it means to "understand" problem solving. This is not necessarily
the case, as we have indicated (e.g., retrospective descriptions of observed problem solving
do not necessarily provide useful forward-looking prescriptions for what problem solvers
should do as "next steps" during problem solving sessions).
A critical component of any agenda to advance the teaching and learning of problem solving
is the clarification of the relationships and connections between the development of
mathematical content understanding and the development of problem solving abilities, as we
have emphasized earlier in this paper. If we can clarify these relationships we can inform
curriculum development and instruction on ways in which we can use problem solving as a
powerful means to develop substantive mathematical concepts. In so doing, we can provide
some alternatives to the existing approaches to teaching problem solving. These existing
approaches include instruction that assumes the required concepts and procedures must be
taught first and then practiced through solving routine “story” problems that normally do not
engage students in genuine problem solving (primarily a content-driven perspective). Another
existing approach, which we have highlighted earlier, is to present students with a repertoire
of problem solving heuristics/strategies such as “draw a diagram,” “guess and check,” “make
a table” etc. and provide a range of non-routine problems to which these strategies can be
applied (primarily a problem solving focus). Unfortunately, both these approaches treat
problem solving as independent of, or at least of secondary importance to, the concepts and
contexts in question.
A powerful alternative to these approaches is one that treats problem solving as integral to the
development of an understanding of any given mathematical concept or process. This
perspective (Lesh & Zawojewski, 2007) also reflects the recognition that the problem solving
of novices and experts differs in ways that go beyond their observed behaviors, that is, what
they actually do in solving a problem. Novices and experts see (interpret and re-interpret)
problem situations differently—experts focus on the underlying structural features of a
problem situation so for them, problem solving involves an interplay between problem
structure (content) and problem solving processes. We continue this discussion in the next
section.

52
 Studies of Problem solving Expertise and their Contributions to Theory Development
The seminal work of Krutetskii (1976) has shown how gifted mathematics students have a
repertoire of ideas, strategies, and representations that seem to be organized into a highly
sophisticated network of knowledge, equipping them with powerful ways to approach
problem solving situations. As noted above, experts readily perceive the underlying structures
of problem situations, project ahead to remove unnecessary steps in the solution process, and
are able to generalize broadly. When we explore expert problem solving beyond the
classroom, we see other factors that play a key role. For example, the knowledge of experts in
workplace environments that require heavy use of mathematics tends to be more organized
around the mathematics of the situation than around general problem solving strategies or
traditional mathematical topics (e.g., Gainsburg, 2006; Hall, 1999).
Although such studies have provided rich insights into how experts perform in given problem
solving situations, they do “not guarantee that one is studying the experts at what actually
makes them experts” (Lester & Kehle, 2003, p. 504). In other words, how do experts become
experts? We need new studies on the nature and development of expertise—how expertise
evolves within episodes of problem solving and over many experiences. Presumably,
students’ understandings of problem solving are not so different from their understandings of
other aspects they are to learn in mathematics. For example, students’ understandings of
problem solving heuristics probably develop. And, development should be able to be traced.
So, we need more studies about problem solving that are similar to the studies that
mathematics educators have conducted about the development of concepts and abilities in
topic areas such as early number concepts, rational number concepts, early algebra concepts,
and so on.
Rather than just describe the behavior we observe as experts solve problems, we need to
know how they interpret the problem situations, how they mathematize them, how they
quantify them, how they operate on quantities, and so on (Lesh & Zawojewski, 2007).
Furthermore, we need to look beyond the assumption that experts initially learn content, then
acquire problem solving strategies, and then learn ways to apply the mathematics and
strategies they have developed. As Zawojewski and Lesh (2003) and others have argued, the
development of problem solving expertise appears as a synergistic, holistic development of
varying degrees of mathematical content, problem solving heuristics/strategies, higher-order
thinking, and affect—all of which are situated in particular contexts.
Theory Development: A Models and Modeling Perspective (MMP) on the Development
of Problem Solving in and beyond the Classroom
Before we explore theory development, we need to offer a more appropriate definition of
problem solving, one that does not separate problem solving from concept development as it
occurs in real-world situations beyond the classroom. We adopt here the definition of Lesh
and Zawojewski (2007):
A task, or goal-directed activity, becomes a problem (or problematic) when the “problem
solver” (which may be a collaborating group of specialists) needs to develop a more
productive way of thinking about the given situation (p. 782).
Thinking in a productive way requires the problem solver to interpret a situation
mathematically, which usually involves progression through iterative cycles of describing,
testing, and revising mathematical interpretations as well as identifying, integrating,
modifying, or refining sets of mathematical concepts drawn from various sources (Lesh &
English, 2005; Lesh & Zawojewski, 2007). These processes are the rudiments of
mathematical modeling. Seeing problem solving from a models and modeling perspective

53
(MMP) contrasts with the traditional definition of problem solving as searching for a way to
progress from the “givens” to the “goals.” Rather, from a models and modeling perspective,
problem solving involves iterative cycles of understanding the givens and the goals of a
problem. In the remainder of this section we highlight some of the key features of problem
solving from a models and modeling perspective.
1. When the solution to a problem involves the development of a model (or artifact or
conceptual tool), and when the underlying conceptual systems are expressed in forms that can
be examined and assessed by students themselves, solutions tend to involve sequences of
iterative express-test-revise cycles similar to the kind that are involved in the first-, second-,
and nth-drafts in the development of other kinds of symbolic or graphic descriptions of
situations. Furthermore, if the underlying conceptual system is one that Piaget-inspired
researchers have investigated, then the modeling cycles that problem solvers go through
during a single 60-90 minute problem solving session are often strikingly similar to the stages
of development that Piagetians have documented over time periods of several years.
Consequently, we have sometimes referred to such sessions as local conceptual development
sessions (Lesh & Harel, 2003) – because students’ thinking often evolves through several
stages similar to those recognized by the Piagetians during a single 60-90 minute episode.
2. When significant conceptual adaptations occur within a single problem solving session,
researchers are able to go beyond observing sequential states of knowledge to also directly
observing processes that lead from one state to another. And, such observations have made
it clear that it is seldom appropriate to think of solution processes as activities in which
students connect previously-mastered-but-disconnected concepts and processes. Nor do
solutions involve movement along a path which is formed by linking together concepts,
processes, facts, and skills. Instead, problem solvers’ early interpretations tend to involve a
collection of partly-overlapping-yet-undifferentiated partial interpretations of different
aspects of the situations of conceptual systems. So, regardless of whether the problem solver
is an individual or a group, model development tends to involve gradually sorting out,
clarifying, revising, refining, and integrating conceptual systems that are at intermediate
stages of development.
3. When solutions to problems involve the development of mathematically significant
artifacts or tools, when the underlying design is an important part of the product that is
designed to help solve the problem, and when the product needs to be powerful (for the
specific situation in which it was first created), sharable (with other people), and reusable (in
situations different to the one in which it was first created), the knowledge and abilities that
are embodied2 in these products tend to be generalizable and transferable.
4. Heuristics that are intended to help problem solvers make productive adaptations to
existing ways of thinking tend to be significantly different from heuristics that are intended to
help problem solvers figure out what to do when they are stuck (with no apparent concepts
available). Their functions tend to have far less to do with helping students know what to do
next, and have far more to do with helping them interpret the situation (including alternative
ways of thinking about givens, goals, personal competencies, and “where they are” in
solution processes). Furthermore, such heuristics often function tacitly rather than as
explicitly executed rules. So, learning them is similar to situations in which athletes or
performing artists analyze videotapes of their own performances (or those of others). It is
useful to develop languages (interpretation systems and conventions for making
interpretations) for describing these past performances. But, such languages usually are not
2
Here, we use the term “embodiment” in the way used by Zoltan Dienes who introduced the notion of
concrete embodiments of mathematical concepts (Dienes, 1960).

54
intended to give rise to prescriptive rules about what to do at specific points in future
performances. Instead, the language and imagery tends to be aimed mainly at helping
students make sense of things during future performances. In other words, they are aimed
mainly at the development of more powerful models.
5. When problem solving involves model development, heuristics and metacognitive
processes tend to evolve in ways that are quite similar to the dimensions of development that
apply to other types of concepts or abilities that mathematics educators have studied. For
example, Vygotsky’s (1978) concept of internalizing external functions often results in early
understandings of heuristics that are distinctly social in character. So, instead of “looking at a
similar problem” students may find it more useful to think of themselves as “looking at the
same problem from another point of view” (and to be aware of the fact that one’s current
point of view is not the only possible point of view).
6. In fields like engineering it is considered to be “common knowledge” that realistic
solutions to realistically complex problem solving situations nearly always need to integrate
concepts and procedures drawn from a variety of textbook topic areas or theories. Likewise,
when students develop realistically useful models (or other conceptual tools) for making
sense of realistically complex “real-life” situations, they often need to integrate ideas and
procedures drawn from more than a single textbook topic area (measurement, geometry,
probability, statistics, and algebra). One reason for this is because useful solutions often
involve trade-offs involving conflicting goals associated with multiple agents. These goals
may involve low costs but high quality, or low risk but high gain, or rapid but thorough
development. The models that are produced are “chunks of knowledge” that represent
inherently connected ideas that need to be unpacked in follow-up teaching and learning
activities. Even after connected ideas are unpacked, students’ knowledge often continues to
be organized around experience as much as it is organized around abstractions.
7. When problem solvers describe or design things mathematically, they tend to do more than
simply engage logical-mathematical systems; they also engage feelings, values, beliefs, and a
variety of problem solving processes, facts, and skills. So, the development of processes,
skills, attitudes, beliefs, is part of the development of specific models. Skills, attitudes, and
beliefs are not developed separately in the abstract before they are connected to concepts or
conceptual systems; skills, attitudes, and beliefs are engaged and developed when the relevant
models are engaged. Thus, skills, attitudes, and beliefs are integral parts of relevant models.

CONCLUDING POINTS
We have argued in this paper that research on mathematical problem solving has stagnated
for much of the 1990s and early part of this century. Furthermore, the research that has been
conducted does not seem to have accumulated into a substantive, future-oriented body of
knowledge on how we can effectively promote problem solving within and beyond the
classroom. This lack of progress is mainly due to the many years of repeated elaborations of
rule-governed conceptions of problem solving competence.
The time has come to consider other options for advancing problem solving research and
curriculum development—we have highlighted the need to re-examine foundation-level
assumptions about what it means to understand mathematics concepts and problem solving
processes. One powerful alternative we have advanced is to utilize the theoretical
perspectives and accompanying research methodologies of a models & modeling perspective
(MMP) on mathematics problem solving, learning, and teaching. Our second paper elaborates
further on this perspective and on the associated research methodologies.

55
Adopting an MMP means researchers who study students’ models and modeling
developments naturally utilize integrated approaches to exploring the co-development of
mathematical concepts, problem solving processes, metacognitive functions, dispositions,
beliefs, and emotions. These researchers also view problem solving processes
developmentally, in a similar way they would in studying the development of mathematical
concepts in topic areas such as early number, geometry, and algebra. In addition, the
problems used are simulations of appealing, authentic problem solving situations (e.g.,
selecting sporting teams for the Olympic Games) and engage students in mathematical
thinking that involves creating and interpreting situations (describing, explaining,
communication) at least as much as it involves computing, executing procedures, and
reasoning deductively.

REFERENCES
Anderson, J. R., Boyle, C. B., & Reiser, B. J. (1985). Intelligent tutoring systems. Science,
228, 456-462.
Begle, E. G. (1979). Critical Variables in Mathematics Education. Washington D. C.: the
Mathematics Association of America and the National Council of Teachers of
Mathematics.
Charles, R. & Silver, E., (1988). The teaching and assessing of mathematical problem
solving. Reston, VA: National Council of Teachers of Mathematics.
Dienes, Z. (1960). Building up Mathematics. London: Hutchins Education Ltd.
English, L. D. (2007). Complex systems in the elementary and middle school mathematics
curriculum: A focus on modeling. In B. Sriraman (Ed.), Festschrift in Honor of Gunter
Torner. The Montana Mathematics Enthusiast, (pp. 139-156). Information Age
Publishing.
English, L. D. (2008). Mathematical modeling: Linking Mathematics, Science, and the Arts
in the Elementary Curriculum. In B Sriraman, C. Michelsen, & A. Beckmann, & V.
Freiman (Eds.), Proceedings of The Second International Symposium on Mathematics
and its Connections to the Arts and Sciences (MACAS2, pp. 5-36). University of
Southern Denmark Press.
English, L. D. (In press, 2009). Promoting interdisciplinarity through mathematical
modelling. ZDM: The International Journal on Mathematics Education, 41(1).
Gainsburg, J. (2006). The mathematical modeling of structural engineers. Mathematical
Thinking and Learning, 8(1), 3-36.
Grouws, D., (1992). (Ed.), Handbook of research on mathematics teaching and learning: A
project of the National Council of Teachers of Mathematics (pp. 334-370). New York,
NY: Macmillan Publishing Co.
Hall, R. (1999). Case studies of math at work: Exploring design-oriented mathematical
practices in school and work settings (NSF Rep. No. RED-9553648), Arlington, VA:
National Science Foundation.
Hamilton, E. (2007). What changes are needed in the kind of problem solving situations
where mathematical thinking is needed beyond school? In R. Lesh, E. Hamilton, & J.
Kaput (Eds.), Foundations for the future in mathematics education (pp. 1-6). Mahwah,
NJ: Lawrence Erlbaum.

56
Hutchins, E. (1995a) Cognition in the Wild. Cambridge, MA: MIT Press.
Hutchins, E. (1995b) How a cockpit remembers its speeds. Cognitive Science. 19, 265-288.
Krutetskii, V. (1976). The psychology of mathematical abilities in school children. Chicago,
IL: University of Chicago Press.
Kuehner, J. P., & Mauch, E. K. (2006). Engineering applications for demonstrating
mathematical problem solving methods at the secondary education level. Teaching
Mathematics and its Applications, 25(4), 189-195.
Lesh, R. & Doerr, H. (2003). Foundation of a Models and Modeling Perspective on
Mathematics teaching and Learning. In R.A. Lesh & H. Doerr (Eds.), Beyond
Constructivism: A models and modeling perspective on mathematics teaching, learning,
and problem solving. 9-34. Mahwah, NJ: Erlbaum.
Lesh, R. & English, L. D. (2005). Trends in the evolution of models and modeling
perspectives on mathematical learning and problem solving. In H. Chick & J. Vincent
(Eds.), Proceedings of the 29th Annual Conference of the International Group for the
Psychology of Mathematics Education. (pp. 192-196). University of Melbourne.
Lesh, R. & Harel, G. (2003). Problem solving, modeling, and local conceptual development.
International Journal of Mathematics Thinking and Learning, 5, 157-189.
Lesh, R. & Yoon, C., (2004) Evolving communities of mind-In which development involves
several interacting and simultaneously developing strands. Mathematical Thinking and
Learning, 6(2), 205-226.
Lesh, R. & Zawojewski, J. S. (2007). Problem solving and modeling. In F. Lester (Ed.). The
Second Handbook of Research on Mathematics Teaching and Learning. (pp. 763-804).
Charlotte, NC: Information Age Publishing.
Lesh, R. (In press). Directions for future research and development in engineering education.
In J. Zawojewski, H. Diefes-Dux, & K. Bowman (Eds.), Models and modeling in
Engineering Education: Designing experiences for all students. Rotterdam: Sense
Publications.
Lesh, R., Middleton, J., Caylor, E., & Gupta, S., (2008) A science need: Designing tasks to
engage students in modeling complex data. Educational studies in Mathematics, 68(2),
113-130.
Lester, F. K. & Charles, R. I. (Eds.) (2003). Teaching mathematics through problem solving:
PreK - 6. Reston, VA: National Council of Teachers of Mathematics.
Lester, F. K. & Kehle, P. E, (2003). From problem solving to modeling: The evolution of
thinking about research on complex mathematical activity. In R. A. Lesh & H. M.
Doerr (Eds.), Beyond constructivism: Models and modeling perspectives on
mathematics problem solving, learning, and teaching (pp. 501-518). Mahwah, NJ:
Lawrence Erlbaum Associates.
Lester, F. K., Garofalo, J., & Kroll, D. L. (1989). Self-confidence, interest, beliefs, and
metacognition. Key influences on problem solving behavior. In D. B. McLeod & V. M.
Adams (Eds.), Affect and mathematical problem solving: A new perspective (pp. 75-
88). New York: Springer-Verlag.
Maclean, R. (2001). Educational change in Asia: An overview. Journal of Educational
Change 2, 189-192.edies. London: World Scientific, 2003.

57
National Council of Teachers of Mathematics. (2000). Principles and standards for school
mathematics. Reston, VA: Author.
NCTM Math Standards (2008) http://standards.nctm.org/document/chapter3/index.htm
Accessed: 22 May 2008
Noss, R., Hoyles, C., & Pozzi, S. (2002). Abstraction in expertise: A study of nurses’
conceptions of concentration. Journal for Research in Mathematics Education, 33(3),
204-229.
Petroski, H. (2003). Early education. American Scientist, 91, 206-209.
Polya, G. (1945). How to solve it. Princeton, NJ: Princeton University Press.
Sawyer, R. K. (2007). Group Genius: The Creative Power of Collaboration. New York:
Basic Books.
Schoen & Charles, (2003). (Eds.), Teaching mathematics through problem solving: Grades
6-12. Reston, VA. National Council of Teachers of Mathematics.
Schoenfeld, A. (1992). Learning to think mathematically: Problem solving, metacognition,
and sense making in mathematics. In D. A. Grouws (Ed.), Handbook of research on
mathematics teaching and learning: A project of the National Council of Teachers of
Mathematics (pp. 334-370). New York, NY: Macmillan Publishing Co.
Schoenfeld, A. (2007). Problem Solving Reconsidered: Toward a Theory of Goal-Directed
Behavior. Presentation given at the 2007 NCTM national conference pre-session.
Address delivered in Atlanta, GA, USA. 19 March 2007.
Silver, E.A. (1985). Research on teaching mathematical problem solving: Some under
represented themes and needed directions. In E. A. Silver (ed.), Teaching and learning
mathematical problem solving. Multiple research perspectives (pp. 247-66). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Sriraman, B., & Adrian, H. (2008). A critique and response to multicultural visions of
globalization. Interchange 39(1), 119-130.
Tan, J. (2002). Education in the twenty-first century: Challenges and dilemmas. In D. da
Cunha (Ed.) Singapore in the new millennium: Challenges facing the citystate. (p. 154-
186). Singapore: The Institute of Southeast Asian Studies.
Vygotsky, L.S. (1978). Mind and society: The development of higher psychological
processes. Cambridge, MA: Harvard University Press.
Zawojewski, J. & Lesh, R. (2003). A Models and Modeling Perspective on Problem Solving
Strategies In R.A. Lesh & H. Doerr (Eds.), Beyond Constructivism: A models and
modeling perspective on mathematics teaching, learning, and problem solving. 9-34.
Mahwah, NJ: Erlbaum.
Zawojewski, J. & McCarthy, L. (2007). Numeracy in practice. Principal Leadership, 7(5),
32-38
Zawojewski, J., Hjalmarson, J. S., Bowman, K., & Lesh, R. (In press). A modeling
perspective on learning and teaching in engineering education. In J. Zawojewski, H.
Diefes-Dux, & K. Bowman (Eds.), Models and modeling in engineering education:
Designing experiences for all students. Rotterdam: Sense Publications.

58
 TEACHING MATHEMATICS IN THE CLASSROOM THROUGH PROBLEM SOLVING

Norma S. G. Allevato
Universidade Cruzeiro do Sul - UNICSUL/ São Paulo/Brazil

Lourdes R. Onuchic
Universidade Estadual Paulista Julio de Mesquita Filho – UNESP/Rio Claro/Brazil

ABSTRACT
We present an approach to teaching-learning-evaluating mathematics through problem solving.
The historical context is briefly described leading up to the current guidelines of the National
Council of Teachers of Mathematics (USA), on which we base a characterization of teaching
through problem solving. Considered a teaching method, its foundations and general guidelines
for implementation in the classroom are presented. Mathematics teaching-learning-evaluation
through problem solving has been used and studied systematically, at all educational levels and
in teacher education activities, by the Problem Solving Work and Study Group (Grupo de
Trabalho e Estudos em Resolução de Problemas – GTERP), based at UNESP, Rio Claro, São
Paulo, Brazil. The research developed by the group follows essentially qualitative approaches,
with the main objective of reflecting on and analyzing the possibilities this method offers for
increasing learning and improving teaching processes, as well as promoting improvement of the
practices of mathematics teachers. Our studies show that students’ construction of knowledge
related to mathematical concepts and contents is more meaningful and effective, and when
applied in teacher education activities, it favors significant improvements in their teaching
practice.

INTRODUCTION

Mathematical problems have occupied a central position in school mathematics curricula

since Antiquity. Records of mathematical problems can be found in ancient Egyptian, Chinese,
and Greek history, and problems are still found in mathematics textbooks from the 19th and 20th
Centuries, and up to the current day. According to Stanic and Kilpatrick (1989), the main point
to consider in the examples presented in these books is they reflect a very limited view of
learning and problem solving. The role of problem solving in the school mathematics
curriculum is the result of conflicting forces linked to old and enduring ideas regarding the
benefits of studying mathematics and a variety of events that occurred at the beginning of the
20th Century.

59
 Problem solving as a classroom teaching method, here denoted Mathematics Teaching-
Learning-Evaluation through Problem Solving, is a fairly new concept in mathematics education,
despite the long history of problem solving in school mathematics. Consequently, the method
has not been the object of very much research.

PROBLEM SOLVING AND RESEARCH IN THE 20TH CENTURY

The definition presented by Leder (1998) for “educational research” outlines the various
aspects that it can address, including: the purposes of education, teaching and learning
processes, professional development, organizational resources, policies, and strategies. The
volume of studies on library shelves indicates that it has become an immense enterprise, and that
the search for new knowledge not only continues, but has been widely documented. In recent
years, the volume, scope, and diversity of educational research, in general, and research in
mathematics education, in particular, have grown substantially.
Having begun as a field of systematic studies with the work of Felix Klein at the
beginning of the 20th Century, mathematics education had grown into a vast and intricate
endeavor by the end of the century. Felix Klein was one of the most important mathematicians of
the late 19th Century, and one of the last, together with Gauss, Riemann, and Poincaré, to break
the barrier of specialization and provide the fundamental elements that gave impetus to the
mathematics of the 19th and early 20th Century. It was then that he wrote his book Elementary
Mathematics from an Advanced Standpoint.
Klein stated, in his autobiography written in 1923, that the totality of all knowledge and
the ideal of a complete education should not be neglected because of specialized studies, and that
universities should be concerned with preparatory teaching in the schools. He emphasized
teacher education, in particular. He was a brilliant mathematician who was sincerely and
seriously concerned about issues related to teaching.
At that time, mathematics teaching was characterized by work based on repetition, where
the memorization of facts was considered important. Years later, a different orientation began to
emphasize that students should learn with comprehension, and should understand what they were
doing.

60
 It was then that talk began of solving problems. George Polya (1944) emerged as a
reference, emphasizing the importance of discovery and of encouraging students to think by
means of problem solving. In his book How to Solve It, he states “A great discovery solves a
great problem, but there is always a bit of discovery in the solution of any problem”. In 1949, he
wrote that solving problems is the specific realization of intelligence, and that if education does
not contribute to the development of intelligence, it is obviously incomplete. In 1948, the work
developed by Herbert F. Spitzer in basic arithmetic, in the U.S.A., was based on learning with
comprehension, always using problems-situations; and in 1964, in Brazil, the teacher Luis
Alberto S. Brasil defended teaching mathematics using problems that generated new concepts
and contents.
In the 1960s and 1970s, mathematics teaching in Brazil and in other countries of the
world was influenced by a reform movement known as Modern Mathematics. This reform
dominated the scene, and like the others before it, failed to include the participation of classroom
teachers. It presented a mathematics based on structures of logic, algebra, topology, and order,
and emphasized set theory. It highlighted many properties, reflected excessive concern with
mathematical abstractions, and used a universal, precise, and concise language. However, it
emphasized the teaching of symbols and complex terminology, which compromised learning. In
this reform, teaching was approached with excessive formalization, distancing itself from
practical issues.
According to Onuchic & Allevato (2005), these reforms were not as successful as
expected. The questions continued: Are these reforms aimed at preparing a citizen who is useful
to the society in which he/she lives? Do they seek to teach mathematics in a way that prepares
students for a world of work that demands mathematical knowledge? In addition to this,
particularly in the 1970s, there was a growing concern with a mathematics curriculum that was
initially aimed at increasing test scores, also known as computational ability tests.
Concomitantly, at the beginning of the 1970s, systematic investigation of problem
solving and its implications for curricula was initiated. Thus, the importance attributed to
problem solving is relatively recent, and only in this decade did mathematics educators come to
accept the idea that the development of problem-solving abilities deserved more attention. At the
end of the 1970s, problem solving emerged, gaining greater acceptance around the world. In

61
1976, at the 3rd International Congress on Mathematical Education, in Karlsruhe, Germany,
problem solving was one of the themes addressed.
Discussions in the field of mathematics education in Brazil and around the world
demonstrated the need to adapt school work to new trends that could lead to improved ways of
teaching and learning mathematics.
In the U.S.A., the National Council of Teachers of Mathematics (NCTM) responded to
this concern with a series of recommendations for the progress of school mathematics, published
in 1980 in a document entitled An Agenda for Action (NCTM, 1980). All interested groups and
individuals were called to collaborate in the work to seek, through a massive cooperative effort, a
better mathematical education for all. The first recommendation in the document was that
“problem solving be the focus of school mathematics in the 1980s.”
During the 1980s, many resources were developed to facilitate work with problem
solving in the classroom, such as collections of problems, lists of strategies, suggestions for
activities, and guidelines to evaluate student work involving problem solving. Much of this
material contributed to helping teachers make problem solving the central point of their work.
Nevertheless, possibly due to differences in conceptions regarding the significance of
problem solving becoming “the focus of school mathematics”, the work during that decade failed
to achieve a good level of progress (ONUCHIC, 1999). Schroeder & Lester (1989) presented
three different approaches to problem solving that help us reflect on these differences: theorizing
about problem solving; teaching how to solve problems; and teaching mathematics through
problem solving. Teachers who teach about problem solving seek to emphasize Polya’s model,
or a derivation of it. When teaching how to solve problems, the teacher concentrates on the
manner in which mathematics is taught, and what of this can be applied in the resolution of
routine and non-routine problems. In this view, the essential purpose for learning mathematics
was to be able to use it. As the 1980s ended with all these recommendations for action,
researchers began to question the teaching and the effect of strategies and models, and to discuss
the didactic-pedagogical perspectives of problem solving.
Problem solving, as a teaching method, became a focus of research and studies in the
1990s. This new view of mathematics teaching-learning was based especially on studies
developed by the NCTM that culminated in the publication of the Standards 2000, officially
Principles and Standards for School Mathematics.(NCTM, 2000). Problem solving was

62
emphasized as one of the standards for the process of teaching mathematics, and teaching
through problem solving was strongly recommended. (ONUCHIC; ALLEVATO, 2005).
Drawing on the ideas of the NCTM Standards, the PCN (National Curriculum
Parameters) were created in Brazil (BRASIL; 1997, 1998, 1999), which pointed to the
development of the capacity to solve problems, explore them, generalize from them, and even
propose new problems based on them, as one of the purposes of mathematics teaching. They
indicated problem solving as the point of departure for mathematics activities, and discussed
ways to do mathematics in the classroom.
Today, at the beginning of the 21st Century, some of the greatest challenges faced by
mathematics educators in past decades have persisted, changed, or proliferated, as teaching and
society have grown more complex. In “Unfinished Business: Challenges for Mathematics
Educators in the Next Decades”, Kilpatrick & Silver (2000) outline what they believe to be the
main challenges: guarantee mathematics for all; promote students’ understanding; maintain
balance in the curriculum; use evaluation as an opportunity for learning; and develop
professional practice.
Cai (2003) emphasizes, however, that although little is known regarding how students
attribute meaning and learn mathematics through problem solving, many ideas associated with
this approach – the change in the teacher’s role, the selection and elaboration of problems,
collaborative learning, among others – have been researched intensively, offering answers to
various frequently-asked questions regarding this way of teaching.

MATHEMATICS TEACHING-LEARNING-EVALUATION THROUGH PROBLEM SOLVING

The compound word “teaching-learning-evaluation” was chosen intentionally to express

the idea that teaching and learning should take place simultaneously during the construction of
knowledge, with the teacher as guide and the students as co-builders of this knowledge. In
addition, this methodology integrates a more updated conception regarding evaluation. It is
constructed during the problem solving, integrated with the teaching to follow students’ growth,
increasing learning and reorienting practice in the classroom when necessary.
According to Van de Walle (2001), truly effective mathematics teachers should involve
four basic components in their work: (1) an appreciation of the discipline of mathematics itself,

63
meaning “doing mathematics”; (2) an understanding of how students learn and construct ideas;
(3) the ability to plan and select tasks so that students learn mathematics in a problem solving
environment; (4) the ability to integrate evaluation with the process to increase learning and
improve teaching from day to day.
Teaching-Learning-Evaluation of Mathematics through Problem Solving differs from
approaches that privilege rules regarding “how to”. It “reflects a tendency to react to past
characteristics, like a set of facts, mastering algorithmic procedures, or the acquisition of
knowledge through routine or mental exercise”. (ONUCHIC, 1999, p.203).
It corresponds to work in which a problem is the point of departure for learning, and the
construction of knowledge occurs in the process of solving it. Teacher and students develop the
work together, and learning takes place collaboratively in the classroom (ALLEVATO,
ONUCHIC, 2007; ONUCHIC; ALLEVATO, 2005). The methodology is similar to the Japanese
approach to teaching mathematics through problem solving. In Problem Solving as a Vehicle for
Teaching Mathematics: a Japanese Perspective, Yoshinori Shimizu (2003) writes that “Japanese
teachers in elementary schools often organize an entire mathematics lesson around multiple
solutions to a single problem in a whole-class instructional mode. This organization is
particularly useful when introducing a new concept or a new procedure during the initial phase
of a teaching unit.”(p.206).
For Van de Walle (2001), a problem is any task or activity for which students have no
prescribed or memorized methods or rules, and no perception that a specific method for arriving
at the correct solution exists. Adding a subjective character to this question, in the context of the
methodology presented here, we consider that problem refers to “everything that we do not know
how to do, but are interested in doing”.
There are no rigidly defined ways to put this methodology into practice (SHIMIZU, 2003;
KRULIK; RUDNICK, 2005; ONUCHIC; ALLEVATO, 2005; VAN DE WALLE; LOVIN,
2006). One proposal is to organize activities according to the following stages:
1) Form groups and hand out the activity. The teacher presents the problem to the
students, who, divided into small groups, read and try to interpret and understand the problem. It
should be emphasized that the mathematical content necessary, or most appropriate, to solve the
problem has not yet been presented in class. The problem proposed to the students, which we call

64
the generative problem, is what will lead to the content that the teacher plans to construct in that
lesson.
2) Observe and encourage. The teacher no longer has the role of transmitter of
knowledge. While students attempt to solve the problem, the teacher observes, analyzes
students’ behavior, and stimulates collaborative work. The teacher mediates in the sense of
guiding students to think, giving them time to think, and encouraging the exchange of ideas
among students.
3) Help with secondary problems. The teacher encourages students to use their previous
knowledge, or techniques that they already know, to solve the problem, and stimulates them to
choose different methods based on the resources they have available. Nevertheless, it is
necessary to assist students with their difficulties, intervening, questioning, and following their
explorations, and helping them to solve secondary problems when necessary. These refer to
doubts presented by the students in the context of the vocabulary present in the statement of
problem; in the context of reading and interpretation; as well as those that might arise during the
problem solving, e.g. notation, the passage from vernacular to mathematical language, related
concepts, and operational techniques, to enable the continuation of the work.
4) Record solutions on the blackboard. Representatives of the groups are invited to
record solutions on the blackboard. Correct as well as incorrect solutions, as well as those done
for different processes, should be presented for all the students to analyze and discuss.
5) Plenary session. The teacher invites all students to discuss solutions with their
colleagues, to defend their points of view and clarify doubts. The teacher acts as a guide and
mediator in the discussions, encouraging the active and effective participation of all students, as
this is the richest moment for learning.
6) Seek consensus. After addressing doubts and analyzing resolutions and solutions
obtained for the problem, the teacher attempts to arrive at a consensus with the whole class
regarding the correct result.
7) Formalize the content. At this moment, called “formalization”, the teacher makes a
formal presentation of the new concepts and contents constructed, highlighting the different
operative techniques and properties appropriate for the subject.
It should be reiterated that, in this methodology, the problem is proposed to the students
before the mathematical contents necessary or most appropriate for solving it (planned by the

65
teacher according to the program for that grade level) have been formally presented. Thus, the
teaching-learning of a mathematical topic begins with a problem that expresses key aspects of
this topic, and mathematical techniques should be developed in the search for reasonable answers
to the problem given.
For Van de Walle (2001), problem solving should be seen as a main teaching strategy,
and he points to the importance of beginning the work from the point where students are,
contrary to other ways of teaching that begin from the point where the teachers are, ignoring
what the students bring with them to the classroom. He goes on to state that teaching with
problems has great value, and that despite the difficulties, there are good reasons to engage in the
effort.
Without a doubt, teaching mathematics through problem solving is an approach
consistent with the recommendations of the NCTM (NCTM, 2000) and the Brazilian National
Curriculum Parameters (BRASIL, 1997, 1998, 1999), as the mathematical concepts and abilities
are learned in the context of problem solving. bThe development of thinking processes at the
university level will be promoted through these experiences, and the work of teaching
mathematics will take place in an environment of investigation guided by problem solving.
In agreement with Krulik & Rudnick (2005), and always with the objective of carrying
out ongoing evaluation, new problems related to the generative problem are proposed to the
students following the stage of formalization, with the aim of analyzing whether or not the
essential elements of the mathematical contents introduced in that lesson were understood. In our
view, the understanding of mathematics by students involves the idea that understanding is
essential relating. It should be emphasized that indications that a student understands,
misunderstands, or fails to understand specific mathematical ideas emerge often as he/she solves
a problem.
Rather than being the focus of mathematics teaching, when considered as a teaching
method, problem solving makes understanding its central focus and objective. In this way, the
emphasis attributed to problem solving is not removed, but its role in the curriculum is
broadened. It moves beyond a limited activity to engage students in the application of
knowledge, following the acquisition of certain concepts and techniques, to be a means of
acquiring new knowledge as well as a process in which the students can apply previously
constructed knowledge (ONUCHIC, 1999).

66
 Mathematics has always played an important role in society. This role is more significant
today, and may become even more important in the future. People do not always think
mathematically, nor are they aware that, if they were to do so, they might make better decisions.
This lack of awareness may be a failure of the mathematics taught as well as the way it is taught.
Often the teaching of mathematics produces students with over-simplified conceptions and
strategies that are excessively mechanical to resolve problems. For Hiebert & Behr(1989), rather
than considering knowledge as a package that is ready and finished, teaching should encourage
students to construct their own knowledge.

RESEARCH AND SCIENTIFIC PRODUCTION RELATED TO PROBLEM SOLVING AT UNESP

Romberg (1998) considers the objective of mathematics education to be the production of

new knowledge about the teaching and learning of mathematics, and that since students learn
most of “their mathematics” in school, research should identify the main classroom components
that promote mathematical understanding, and point out the organizational characteristics that
impede or contribute to the good functioning of those classrooms.
In this paper, we intend to highlight Mathematics Teaching-Learning-Evaluation through
Problem Solving, a teaching method which has been used and studied systematically, at all
educational levels and in teacher education activities, by the Problem Solving Work and Study
Group (Grupo de Trabalho e Estudos em Resolução de Problemas – GTERP), based at UNESP,
Rio Claro, São Paulo, Brazil. The group has been the generative nucleus of investigations,
scientific production, and continuing education in problem solving.
The research conducted composes part of a larger, broader project whose main objective
is to reflect on and analyze the possibilities offered by Mathematics Teaching-Learning-
Evaluation through Problem Solving for increasing learning and improving teaching processes,
as well as promoting improved practices among teachers in the mathematics classroom.
These studies, using a qualitative methodology, consist essentially of interventions in the
realm of participant research and action research. Problem solving activities are elaborated by the
researchers and applied in the classroom. Thus, a large part of the theses, dissertations, and other
work produced by the group narrate and analyze situations involving pedagogical interventions
realized by its members in the classroom or in the sphere of teacher education. The scientific

67
production encompasses a wide variety of mathematical contents at all educational levels –
elementary, high school, and higher learning.
This set of research constitutes a wide spectrum of research possibilities in mathematics
education. A description of the theses and dissertations, as well as the work carried out by the
group up until 2005, can be found in Onuchic (1999) and Onuchic & Allevato (2005).

CONCLUSIONS

Considering mathematics educators are people who are professionally concerned about
mathematics teaching and learning at any educational level, we can testify to their dedication and
relevant production by the volume and quantity of studies in mathematics education carried out
in the 20th Century. Students are currently the beneficiaries of the large variety of instructional
materials created. Certainly if we compare teachers at the beginning of the century with the
teachers of today, we can say that they are much better prepared pedagogically and
mathematically. The majority of school mathematics curricula are richer than those at the
beginning of the century. In spite of all this, the same complaints are heard today: that students
do not like and do not learn mathematics well enough; that teachers do not know mathematics
and do not know how to teach it; that school curricula are superficial, repetitive, and
fragmented… All these complaints, and data obtained from other sources (research, evaluations,
etc), suggest that students leave school poorly prepared, not knowing how to make use of the
mathematics they worked with throughout their many years of schooling. As we already said,
people are often unable to make decisions in life. These people do not always think
mathematically, nor are they aware that, if they did, they would make better decisions.
The teaching method presented here constitutes a way of working in the classroom using
generative problems as a point of departure. Using Mathematics Teaching-Learning-Evaluation
through Problem Solving, students’ construction of knowledge related to mathematical concepts
and contents occurs more meaningfully and effectively. Experiences in research with students
and teacher education activities, in which this approach was used, have favored significant
advances in the understanding of mathematical concepts and contents and improvements in
teaching practice.

68
REFERENCES

ALLEVATO, N. S. G.; ONUCHIC, L. R. O Ensino de Números Racionais e Proporcionalidade

através da Resolução de Problemas. In: Conferência Interamericana de Educação Matemática,
12., 2007. Anais ... Santiago de Queretaro: Benemérita Escuela Normal de Querétaro, 2007. 1
Cd-rom.
BRASIL. MEC. Parâmetros Curriculares Nacionais: Matemática – 1o, 2o, 3o e 4o ciclos,
Ensino Médio: Matemática. Brasília, 1997, 1998, 1999.
CAI, J. What Research Tells Us about Teaching Mathematics through Problem Solving. In:
LESTER JR, F. K. (Ed.). Teaching Mathematics through Problem Solving. Prekindergarten-
Grade 6. Reston/VA: NCTM, 2003. p.141-253.
HIEBERT, J.; BEHR, M. Introduction: capturing the major themes. In: ______. (Eds) Number
concepts and operations in the middle grades. 3 ed. Reston/VA: NCTM, 1989. p.1-18.
KILPATRICK, J.; SILVER, E. A. Unfinished Business: Challenges for Mathematics Educators
in the Next Decades. In: BURKE, M. J.; CURCIO, F. R. (Eds.). NCTM 2000 Yearbook-
Learning Mathematics for a New Century. Reston/VA: NCTM, 2000. p.223-235.
KRULIK, S.; RUDNICK, J. A. Problem-Driven Math: Applying the Mathematics Beyond
Solutions. Chicago, IL: Wright Group/McGrawHill, 2005.
LEDER, G. C. The Aims of Research. In: SIERPINSKA, A.; KILPATRICK, J. (Eds.).
Mathematics Education as a Research Domain – a Search for Identify. Book 1. Kluver
Academic Publisher. 1998. p.131-140.
NCTM. Principles and Standards for School Mathematics. Reston, VA: National Council of
Teachers of Mathematics, 2000.
ONUCHIC, L R.; ALLEVATO, N. S. G. Novas reflexões sobre o ensino-aprendizagem de
matemática através da resolução de problemas. In: BICUDO, M. A. V.; BORBA, M. C. (Orgs).
Educação Matemática - pesquisa em movimento. 2.ed. São Paulo: Cortez, 2005. p.213-231.
ONUCHIC, L. R. Ensino-Aprendizagem de Matemática através da Resolução de Problemas. In:
BICUDO, M. A. V.(Org.). Pesquisa em Educação Matemática. São Paulo: Editora UNESP,
1999. cap.12, p.199-220.
POLYA, G. How to Solve It. Princeton: Princeton University Press, 1944.

69
ROMBERG, T. A.. The Social Organization of Research Programs in Mathematical Sciences
Education. In: SIERPINSKA, A.; KILPATRICK, J. (Eds.). Mathematics Education as a
Research Domain – a Search for Identify. Book 2. Kluver Academic Publisher. 1998. p.379-
389.
SHIMIZU, Y. Problem Solving as a Vehicle for Teaching Mathematics: A Japanese Perspective.
In: LESTER JR, F. K. (Ed.) Teaching Mathematics through Problem Solving.
Prekindergarten-Grade 6. Reston/VA: NCTM, 2003. p.205-214.
STANIC, G. M. A.; KILPATRICK, J. Historical Perspectives on Problem Solving in the
Mathematics Curriculum. In: CHARLES, R. I.; SILVER, E. A. (Ed.). The Teaching and
Assessing of Mathematical Problem Solving. Virginia: Laurence Erlbaum Associates, 1989.
p.1-22.
VAN DE WALLE, J. A; LOVIN, H. L. Teaching Student-Centered Mathematics. New York:
Pearson, 2006.

70
 






TSG-19






               

           
                 

               
            
              



           


               
           
            
              


           

            
                



          

  







                



71


           




            


            
              

             
             



 

     

           




  



            

       
 

             



            
            
             


             
               
             



72
              
             
          

     
         
            



          
            
             
              
         


           
             
               
             




             

           
           

            
             


               
             
             
           
             


               


           
          

            



73
             
            
            
             


  
        
 
 
     
 

    
 
    

 
  

    







               


                
             

            



             
            





 








74



           
             
              




           





   
  
      
 
     
  

        
  
         
   
        
    
 
     
    


        
   
 
 
         
 
  
    
    


         

 
 






            
          


         

75
            

            




             
             
            
          
             

             


               


            
           



            



            
            


  
               


            


                 
            
             


 
                
            


76



            
               
             
               





            
              
             


       
     
      
         
 
         
    
     
     
     
  
 
         

      
     

       
        
     

 

 

             
         
 
            



77
           
           


 








    
        


       
        
       
        
 

      
 
       
 
      



        

       


       

 

       


       

        
 




           
              
           

          
    

            
          
 

78
             
             
           

            
           
          



              


            

               

            
              










            

           















          



79







          














        









          
          






        





       

          


80
 An ICT environment to assess and support students’ mathematical problem-solving
performance in non-routine puzzle-like word problems

Paper to present in TSG19 at ICME11

Theme (d): Research and development in problem solving with ICT technology

Angeliki Kolovou*
Marja van den Heuvel-Panhuizen*#
Arthur Bakker*
Iliada Eliax

*FIsme, Utrecht University, the Netherlands

#IQB, Humboldt University Berlin, Germany
x
Department of Education, University of Cyprus

Summary
This paper reports on a small-scale study on primary school students’ problem-solving
performance. In the study, problem solving is understood as solving non-routine puzzle-like word
problems. The problems require dealing simultaneously with multiple, interrelated variables. The
study employed an ICT environment both as a tool to support students’ learning by offering them
opportunities to produce solutions, experiment and reflect on solutions, and as a tool to monitor
and assess the students’ problem solving processes. In the study, 24 fourth-graders were involved
from two schools in the Netherlands. Half of the students who belonged to the experimental group
worked in pairs in the ICT environment. The analysis of the students’ dialogues and actions
provided us with a detailed picture of students’ problem solving and revealed some interesting
processes, for example, the bouncing effect that means that the students first come with a correct
solution and later give again an incorrect solution. The test data collected before and after this
“treatment” did not offer us a sufficient basis to draw conclusions about the power of ICT
environment to improve the students’ problem-solving performance.

1. INTRODUCTION

Problem solving is a major goal of mathematics education and an activity that can be seen as the
essence of mathematical thinking (Halmos, 1980; NCTM, 2000). With problems tackled in
problem solving typically defined as non-routine (Kantowski, 1977), it is not surprising that
students tend to find mathematical problem solving challenging and that teachers have difficulties
preparing students for it. Despite the growing body of research literature in the area (Lesh &
Zawojewski, 2007, Lester & Kehle, 2003, Schoenfeld, 1985), there is still much that we do not
know about how students attempt to tackle mathematical problems and how to support students in
solving non-routine problems.
In order to get a better understanding of Dutch primary school students’ competences in
mathematical problem solving, the POPO study started in 2004. In this study, 152 fourth-grade
students who are high achievers in mathematics were administered a paper-and-pencil test on non-
routine problem solving. In a few items, students were asked to show their solutions strategies.
The results were disappointing. Students did not show a high performance in problem solving,
despite their high mathematics ability (Van den Heuvel-Panhuizen, Bakker, Kolovou, & Elia, in
preparation). Although the students’ scribbling on the scrap paper gave us important information
about their solution strategies, we were left with questions about their solution processes.
Moreover, after recognizing that even very able students have difficulties with solving the
problems, we wondered what kind of learning environment could help students to improve their
problem solving performance. The POPO study thus yielded a series of questions. To answer these
questions we started the iPOPO study which – in accordance with the two main questions that
emerged from the POPO study – implied a dual research goal.
First, the iPOPO study aimed at gaining a deeper understanding of the primary school
students’ problem solving processes, and, second, it explored how their problem-solving skills can

81
be improved. For this dual goal of assessing and teaching, the study employed ICT both as a tool
to support students’ learning by offering them opportunities to produce solutions, experiment and
reflect on solutions, and as a tool to monitor and assess the students’ problem solving processes. In
particular, we designed a dynamic applet called Hit the target, which is based on one of the paper-
and-pencil items used in the POPO study. Like several of these items, it requires students to deal
with multiple, interrelated variables simultaneously and thus prepares for algebraic thinking.
This paper focuses on the following two research questions: Which problem-solving
strategies do fourth-grade students deploy in this Hit the target environment? Does this ICT
environment support the students’ problem-solving performance?

2. THEORETICAL BACKGROUND

2.1. Mathematical problem solving

The term “problem solving” is used for solving a variety of mathematical problems, ranging from
real-life problems to puzzle-like problems. Our focus is on the latter. We consider problem solving
as a cognitive activity that entails strategic thinking, and that includes more than just carrying out
calculations. An episode of problem solving may be considered as a small model of a learning
process (D’Amore, & Zan, 1996). In problem solving, the solution process often requires several
steps. First the students have to unravel the problem situation. Subsequently, they have to find a way
to solve the problem by seeking patterns, trying out possibilities systematically, trying special cases,
and so on. While doing this they have to coordinate relevant mathematical knowledge, organize the
different steps to arrive at a solution and record their thinking. In sum, in our view problem solving
is a complex activity that requires higher order thinking and goes beyond standard procedural skills
(cf., Kantowski, 1977).
An example of a mathematical problem used in the POPO study is shown in Figure 1. Someone
who knows elementary algebra might use this knowledge to find the answer to this problem by, for
example, solving the equation 2x – 1(10 – x) = 8. Fourth-grade students, however, have not yet learned
such techniques, but can still use other strategies such as systematic listing of possible solutions or trial
and error. Grappling with such problems might be a worthwhile experiential base for learning algebra
in secondary school (cf., Van Amerom, 2002).

Figure 1: Problem used in the POPO Study

Within the complexity that characterizes problem-solving activity, D’Amore and Zan (1996) identify
the involvement of three interrelated discrete variables, as follows: the subject who solves the task;
the task; and the environment in which the subject solves the task. This study primarily focuses on
the third variable, referring to the conditions, which may help a subject to improve his problem
solving abilities.

82
 The research questions stated in Section 1 address two different aspects that are closely related:
monitoring learning and supporting that learning. We have chosen to use ICT for both of these aspects,
because – as Clements (1998) recognized – ICT (1) can provide students with an environment for
doing mathematics and (2) can give the possibility of tracing the students’ work.

2.2. ICT as a tool for supporting mathematical problem solving

A considerable body of research literature has shown that computers can support children in
developing higher-order mathematical thinking (Suppes, 1966; Papert, 1980; Clements & Meredith,
1993; Sfard & Leron, 1996; Clements, 2000; Clements, 2002). Logo programming, for example, is a
rich environment that elicits reflection on mathematics and one’s own problem-solving (Clements,
2000). Suitable computer software can offer unique opportunities for learning through exploration and
creative problem solving. It can also help students make the transition from arithmetic to algebraic
reasoning, and emphasize conceptual thinking and problem solving. According to the Principles and
Standards of the National Council of Teachers of Mathematics (NCTM, 2000) technology supports
decision-making, reflection, reasoning and problem solving.
Among the unique contributions of computers is that they also provide students with an
environment for testing their ideas and giving them feedback (Clements, 2000). In fact, feedback
is crucial for learning and technology can supply this feedback (NCTM, 2000). Computer-assisted
feedback is one of the most effective forms of feedback because “it helps students in building cues
and information regarding erroneous hypotheses”; thus it can “lead to the development of more
effective and efficient strategies for processing and understanding” (Hattie & Timperley, 2007,
p.102). More generally, computer-based applications can have significant effects on what children
learn because of “the computer’s capacity for simulation, dynamically linked notations, and
interactivity” (Rochelle, Pea, Hoadley, Gordin, & Means, 2000, p. 86).
This learning effect can be enhanced by peer interaction. Pair and group work with
computer software can make students more skilful at solving problems, because they are
stimulated to articulate and explain their strategies and solutions (Wegerif & Dawes, 2004).
Provided there is a classroom culture in which students are willing to provide explanations,
justifications, and arguments to each other, we can expect better learning.

2.3. ICT as a window onto students’ problem solving

Several researchers have emphasized that technology-rich environments allow us to track the
processes students use in problem-solving (Bennet & Persky, 2002). ICT can provide mirrors to
mathematical thinking (Clements, 2000) and can offer a window onto mathematical meaning
under construction (Hoyles & Noss, 2003, p. 325). The potential of computer environments to
provide insight into students’ cognitive processes makes them a fruitful setting for research on
how this learning takes place.
Because software enables us to record every command students make within an ICT
environment, such registration software allows us to assess their problem solving strategies in
more precise ways than can paper-and-pencil tasks. Therefore, computer-based tasks as opposed
to conventional paper-and-pencil means have received growing interest in the research literature
for the purposes of better assessment (Clements 1998; Pellegrino, Chudowsky, & Glaser, 2001;
Bennet & Persky, 2002; Burkhardt & Pead, 2003; Threlfall, Pool, Homer, & Swinnerton, 2007;
Van den Heuvel-Panhuizen, 2007).
Where early-generation software just mimicked the paper-and-pencil tasks, recent research
shows that suitable tasks in rich ICT environments can also bring about higher-order problem
solving. For example, Bennet and Persky (2002) claimed that technology-rich environments tap
important emerging skills. They offer us the opportunity to describe performance with something
more than a single summary score. Furthermore, a series of studies indicated that the use of ICT
facilitates the assessment of creative and critical thinking by providing rich environments for
problem solving (Harlen & Deakin Crick, 2003).
By stimulating peer interaction we also expect that students will articulate more clearly their
thinking than when working individually. Thus, student collaboration has a twofold role: it helps them
shape and broaden their mathematical understandings and it offers researchers and teachers a nicely
bounded setting in order to observe collaboration and peer interaction (Mercer & Littleton, 2007).

83
 3. METHOD

3.1. Research design and subjects

The part of the iPOPO study described in this paper is a small-scale quasi-experiment following a
pre-test-post-test control group design. In total, 24 fourth-graders from two schools in Utrecht
participated in the study. In each school, 12 students who belonged to the A level according the
Mid Grade 4 CITO test – in other words to the 25% best students according to a national norm –
were involved. Actually, the range of the scores that correspond to level A of the Mid Grade 4
CITO test is between 102 and 151 points. In both schools, the average mathematics CITO score of
the class was A and the average “formation weight” of the class and the school was 1. This means
that the students were of Dutch parentage and came from families in which the parents had at least
secondary education. First, of each school six students were selected for the experimental group.
Later on, the group of students was extended with six students to be in the control group. These
students also belonged to the A level, but, unfortunately, their average score was lower than that
of the experimental group. The teacher obviously selected the more able students first.
An ICT environment was especially developed for this study to function as a treatment for
the experimental group. Before and after the treatment, a test was administered as pre-test and
post-test. The control group did the test also two times, but did not get the treatment in between.
The quasi-experiment was carried out in March-April 2008. The complete experiment took about
four weeks: in the first week the students did the test, in the second week the experimental group
worked in the ICT environment and in the fourth week the students did again the test.

3.2. Pre-test and post-test

The test that was used as pre-test and post-test was a paper-and-pencil test consisting of three non-
routine puzzle-like word problems, titled Quiz (see Figure 1), Ages, and Coins. The problems are of the
same type and require that the students deal with interrelated variables. The test sheets contain a work
area on which the students had to show how they found the answers. The students’ responses were
coded according to a framework that was developed in our earlier POPO study. The framework covers
different response characteristics including whether the students gave specific strategy information,
how they represented that strategy and what kind of problem-solving strategies they applied.

3.3. Applet used as treatment

The treatment consisted of a Java applet called Hit the target. 1 It is a simulation of an arrow
shooting game. The screen shows a target board, a score board featuring the number of gained points,
and the number of hit and missed arrows, a frame that contains the rules for gaining or loosing points,
and an area in which the number of arrows to be shot can be filled in. A hit means that the arrow hits
the yellow circle in the middle of the target board; then the arrow becomes green. A miss means that
the arrow hits the gray area of the board; in that case, the arrow becomes red.

Figure 2: Screen view of applet in the computer-shooting mode

1
The applet has been programmed by Huub Nilwik.

84
The applet has two modes of shooting: a player shoots arrows by him or herself or lets the
computer do the shooting (see Figure 2). In case the player shoots, he or she has to drag the arrows
to the bow and then draw and unbend the bow. The computer can do the shooting if the player
selects the computer-shooting mode and fills in the number of arrows to be shot. Regarding the
rules for gaining points there are also two modes: the player determines the rules or the computer
does this. The maximum number of arrows is 150 and the maximum number of points the player
can get by one shot is 1000.
As the player shoots arrows or lets the computer do so, the total score on the scoreboard
changes according to the number of arrows shot and the rules of the game. The player can actually
see on the scoreboard how the score and the number of hits and misses change during the shooting.
The player can also remove arrows from the target board, which is again followed by a change in the
total score. When the player wants to start a new shooting round, he or she must click on the reset
button. The player can change the shooting mode or the rules of the game at any time during the
game.
The aim of the applet is that the students obtain experience in working with variables and
realize that the variables are interrelated (see Figure 3); a change in one variable affects the other
variables. For example, if the rules of the game are changed, then the number of arrows should be
also changed to keep the total points constant.

Total points

Number of arrows
Rules of the game - hits and misses
- total arrows
Figure 3: Variables involved

The 12 students of the experimental group worked for about 30 minutes in pairs with the applet.
The pairs were chosen by the researcher in such a way that all of them would have about the same
average CITO score and consisted of a boy and a girl. The dialogue between the students and their
actions on the applet were recorded by Camtasia software, which captures the screen views and
the sound in a video file. Scrap paper was also available to the students. Before the students
started working, it was explained to them that they should work together, use the mouse in turns,
explain their thinking to each other, and justify their ideas.
The work with the applet started with five minutes of free playing in which the students
could explore the applet. Then, they had to follow a pre-defined scenario containing a number of
directed activities and three questions (see Table 1). The first two questions (A and B) are about
the arrows while the rules of the game and the gained points are known. In the third question (C),
which consists of two parts, the rules of the game are unknown.

Table 1: Questions in the pre-defined scenario

Arrows Rules Gained
points
A. How many hits and misses? Hit +3 points, miss –1 point 15 points
B. How many hits and misses? Hit +3 points, miss +1 point 15 points
15 hits and 15 misses C1. What are the rules? 15 points
C2. Are other rules possible to get
the result 15 hits-15 misses-15 points?

The directed activities were meant to assure that all the students had all the necessary experiences
with the applet. During these activities, the students carried out a number of assignments in order
to become familiar with the various features of the applet: the player-shooting mode, the
computer-shooting mode, the rules of the game, and the total score. First, the students had to shoot

85
one arrow, followed by shooting two arrows and then a few more, in order to get five arrows on
the target board. Their attention was then drawn to the scoreboard; they had five hits and zero
misses and their total score was zero since the rules of the game had been initially set to zero.
After that, the rules were changed so that a hit meant that three points were added. Then, the
students had to shoot again five arrows in both shooting modes, each resulting in a total score of
15 points. Afterward, the rule was changed again. A miss then meant that one point had to be
subtracted. At this point, Question A was asked, followed by Questions B and C.

4. RESULTS

4.1. The students’ problem-solving strategies in the ICT environment

All pairs were successful in answering the Questions A, B, and C. The solutions were found based
on discussions and sharing ideas for solutions. In all cases, explanations were provided and the
talk between the students stimulated the generation of hypotheses and solutions. However, some
students provided more elaborate explanations and suggested more successful problem-solving
strategies than others.
In order to identify the problem-solving strategies the students applied, we analyzed all
dialogues between the students. In this paper, however, we will only discuss our findings with
respect to Questions C1 and C2, which triggered the richest dialogues.
Characteristic for Question C is that the number of hits and misses, and the number of points
were given, but that the students had to find the rules. All pairs were able to answer Questions C1
and C2, and most of them could generalize to all possible solutions (“It is always possible if you do
one less”), albeit on different levels of generalization. The Tables 2 and 3 show which strategies the
pairs used when solving Questions C1 and C2. Each pair of students is denoted with a Roman
numeral. Pairs I, II, and III belong to school A, while Pairs IV, V, and VI belong to school B.

Table 2: Problem-solving strategies when solving C1

Strategy Pairs
I II III IV V VI
Average CITO score per pair
111 111 114 110 111 107
1a Directly testing a correct solution (+2 –1 or +1 +0) 1* 1 1 1
2a Testing incorrect canceling-out solution (+1 –1) 1
2b Testing other incorrect solution(s) 1
3 Adapting the rules of the game until a correct 2 2
solution is reached
Number of trials 1 1 1 2 1 3
* The numbers in the cell indicate the order in which the strategies were applied

When answering Question C1 (see Table 2), four out of the six pairs directly came up with a correct
solution. Pair VI found the correct solution in the third trial. The most interesting strategy came from
Pair IV. This pair found the correct solution in the second trial. The pair started with a canceling-out
solution (+1 –1) resulting in a total score of zero and then changed the solution to get 15 points in
total.
Table 3 shows that having found a correct solution in C1 did not mean that the students had
discovered the general principle (or the correct solution rule) of getting “15 hits-15 misses-15
points”. Even after finding the correct solution rule and generating a series of correct solutions,
some students tested wrong solutions again (we could call this the “bouncing effect”). Perhaps
they were not aware that there is only one correct solution rule; the difference between the number
of points added for every hit and the number of points subtracted for every miss (or vice versa)
should be 1, or the difference between the number of hit-points and miss-points should be 15
points. The highest level of solution was demonstrated by Pair VI, who recognized that the
difference between the points added and the points subtracted should be 15 (and that explains why
the difference between the number of points added for every hit and the number of points
subtracted for every miss – or vice versa – should be 1). A clever mathematical solution came
from the Pairs I and II. These students just used the correct solution to C1 in the reverse way to get
the required result of 15 points in total.

86
Table 3: Problem-solving strategies when solving C2
Strategy Pairs
I II III IV V VI
Average CITO score per pair
111 111 114 110 111 107
4a Repeating the correct solution to C1 2
4b Reversing the correct solution to C1 to find 1* 1/3
another correct solution (–1 +2 or –0 +1/+0 +1)
5a Generating a correct solution rule based on 2 4 6 1 4
testing of (a) correct solution(s) for which the
difference between the number of points added
for every hit and the number of points
subtracted for every miss (or vice versa) is 1
5b Generating a correct solution rule based on 1
understanding that the difference between hit-
points and miss-points is 15
5c Generating a general correct solution rule (“the 8
difference of 1 also applies to 16-16-16”)
6 Testing more correct solutions from a correct 3 7 2 2
solution rule
2b Testing other incorrect solution(s) 4 2 1/3/5 1/3/5
7 Generating an incorrect solution rule (keeping 2/4
ratio 2:1 or using rule +even number –odd
number) based on correct solution(s)
* The numbers in the cell indicate the order in which the strategies were applied

Besides strategies that directly or indirectly lead to a correct solution or rule, some other
characteristics were found in the solution processes (see Table 4). Four pairs altered or ignored
information given in the problem description. It is noteworthy that during subsequent attempts to
answer Question C2, some students insisted on keeping the rules constant and changing the
number of hits and misses in order to get a total of 15 points. Pair V, for example, changed the
problem information (15 hits and 15 misses) and started C2 with trying out the solution 1 hit is 15
point added and 1 miss is 15 points subtracted. The total score then became zero; subsequently,
they set the number of hits to 30 and the number of misses to 15, which resulted into a high score.
Even though at that point the researcher repeated the correct problem information, the students
ignored it persistently. In their third attempt, they changed the number of hits and misses to 1 and
0 respectively and the total score became 15 instead of the reverse (15 hits and 15 misses resulting
in 15 points). Only when the researcher repeated the question they considered the correct
information and tried out the solution +4 –2 with 15 hits and 15 misses. However, the total score
was 30 points and they suggested doubling the number of misses to 30 so that the number of total
points would be halved. This is clearly an example of a wrong adaptation. Another example is
from Pair VI. After having +3 and –1 as the rule of the game, resulting in a total of 30 points, the
students change the number of hits into 10 in order to get 15 points as the result but forgetting that
the number of hits should be 15.

Table 4: Other characteristics of the solution processes

Characteristics C1 C2
Pairs Pairs
I II III IV V VI I II III IV V VI
Altering or ignoring X X X X
information
Exploring large X X X X X
numbers (!1000)

Another characteristic of the solution processes was testing rules including large numbers. Four of
the six pairs tried out numbers bigger than 1000. These explorations all took place when answering

87
the second part of Question C. The students found this working with large numbers quite amusing,
since they then could get a large amount of total points. That the students worked with numbers
larger than 1000 was quite remarkable, because it was not possible to fill in numbers of this size in
the applet. Consequently, the students had to work out the results mentally. It is also worth noting
that some students understood that one could go on until one million or one trillion (Pair IV). This
means that several students knew that there are infinite solutions, as it was made explicit by one pair
(see Pair II). Furthermore, most of the students used whole numbers and no one used negative
numbers. In one occasion, a student (from Pair II) suggested adding 1½ points for a hit, but the
applet does not have the possibility to test solutions with fractions or decimals.
Observing the students while working on the applet revealed that the students demonstrated
different levels of problem-solving activity. For example, there were students that checked the
correctness of their hypotheses by mental calculation, while others just tried out rules with the
help of the applet. None of them questioned the infallibility of the applet; when they used the
applet after they had found out that a rule was wrong, they did this to make sure that they were
really wrong. Furthermore, the students also showed differences in the more or less general way in
which they expressed their findings. One of the students articulated that the general rule “a hit is
one point more (added) than the number of points (subtracted) by a miss” also applies to other
triads such as 16 hits-16 misses-16 points and in general to all triads of equal numbers.
To conclude this section about the ICT environment, we must say that observing the
students while working with the applet gave us quite a good opportunity to get closer to the
students’ problem-solving processes.

4.2. Does the ICT environment support the students’ problem-solving performance?
In this section, we discuss the results from the pre-test and the post-test in the experimental and
control group. Figure 4 shows the average number of correct answers per student in both groups in
school A and school B.

3.0
2.5
of correct answers
Average number

2.0
1.5 Pre-test
1.0 Post-test
0.5
0.0
Exp Contr Exp Contr Exp Contr
School A School B Total
Figure 4: Average number of correct answers per student in the pre and the post-test in both groups

114
112
CITO-score

110
Average

108
106
104
102
Exp Contr Exp Contr Exp Contr
School A School B Total
Figure 5: Average CITO score of the experimental and control group

As can been seen in Figure 4, if the group of students is taken as a whole, the experimental group
gained slightly from the treatment. However, we have too few data to give a reliable answer to the
research question. Only 12 students from school A and 12 students from school B were involved
in this study and among these schools, the results were quite different. Only in school A, there is a

88
considerable improvement in the scores of the post-test. Another issue is the mismatch between
experimental and control group (see also Section 3.1). In both schools, the control group scored
lower than the experimental group. This mismatch was more evident in school A. A plausible
explanation for these differences could be that although all students had an A score in
mathematics, the average CITO scores of the experimental group and the control group were
different in school A and school B (see Figure 5).
In fact, the differences between the average CITO score of the experimental and control
group in each school, presented in Figure 5, are similar to the differences between the average
scores of these groups in the paper-and-pencil test. In school A, the control group has a lower
CITO score than the experimental group. The same holds for school B, but there the difference is
smaller than in school A.

5. DISCUSSION

We started this study with two questions that emerged from the earlier POPO study. To investigate
these questions, we set up, as a start, a small-scale study in which an ICT environment played a
crucial role. The dialogues between the students and their actions when working in the ICT
environment gave us a first answer to the first research question. The collected data provided us
with a detailed picture of students’ problem solving and revealed some interesting processes, for
example, the bouncing effect and the making of wrong adaptations.
Our second question is difficult to answer. The sample size, and the number of the test
items were not sufficient to get reliable results and the time we had at our disposal was not enough
to gather and analyze more data. Moreover, the time that the experimental group worked in the
ICT environment was rather limited to expect an effect. Despite these shortcomings, we decided to
carry out a small-scale study in order to try out the test items and the ICT environment with a
small group of students first.
Clearly, more data (more students, more schools and more problems) are needed to confirm
or reject our conjecture that having experience with interrelated variables in a dynamic, interactive
ICT environment leads to an improvement in problem solving performance. For this reason, we
will extend our study to larger groups of students, involving students of higher grades and
different mathematical ability levels as well. Moreover, to see more of an effect we will enlarge
the working in the ICT environment substantially. In addition, we will extend the study by
analyzing the students’ problem-solving strategies when solving paper-and-pencil problems. Our
experiences from the present study will serve as a basis for doing this future research.

REFERENCES

Bennett, R. E., & Persky, H. (2002). Problem solving in technology-rich environments. In C.

Richardson (Ed.), Assessing gifted and talented children (pp. 19-33). London:
Qualifications and Curriculum Authority.
Burkhardt, H., & Pead, D. (2003). Computer-based assessment: a platform for better tests? In C.
Richardson (Ed.), Whither Assessment? (pp. 133-148). London: Qualifications and
Curriculum Authority, UK. Retrieved from
http://www.nottingham.ac.uk/education/MARS/papers/whither/index.htm.
Clements, D. H., & Meredith, J. S. (1993). Research on Logo: Effects and Efficacy. Journal of
Computing in Childhood Education, 4(3-4), 263-290.
Clements, D. H. (1998). Computers in Mathematics Education Assessment. In G. W. Bright & J. N.
Joyner, Classroom Assessment in Mathematics: Views from a National Science Foundation
Working Conference (pp. 153-159). Lanham, MD: University Press of America.
Clements, D. H. (2000). From exercises and tasks to problems and projects. Journal for Research
in Mathematics Education, 19(1), 9-47.
Clements, D. H. (2002). Computers in Early Childhood Mathematics. Contemporary Issues in
Early Childhood, 3(2), 160-181.

89
D’Amore, B, & Zan, R. (1996). Italian research on problem solving 1988-1995. In A. Gagatsis &
L. Rogers (Eds.), Didactics and History of Mathematics (pp. 35-51). Thessaloniki: Erasmus
ICP-95-G-2011/11.
Halmos, P. (1980). The heart of mathematics. American Mathematical Monthly, 87(7), 519-524.
Harlen, W., & Deakin Crick, R. (2003). A systematic review of the impact on students and
teachers of the use of ICT for assessment of creative and critical thinking skills. London:
EPPI-Centre, Social Science Research Unit, Institute of Education, University of London.
Retrieved from http://eppi.ioe.ac.uk.
Hattie, J., & Timperley, H. (2007). The Power of Feedback. Review of Educational Research.
77(1), 81-112.
Hoyles, C., & Noss, R. (2003). What can digital technologies take from and bring to research in
mathematics education? In A. J. Bishop (Ed.), Second International Handbook of
Mathematics Education (pp. 323-349). Dordrecht; Boston: Kluwer Academic Publishers.
Kantowski, M. G. (1977). Processes Involved in Mathematical Problem Solving. Journal for
Research in Mathematics Education, 8(3), 163-180.
Lesh, R., & Zawojewski, J. S. (2007). Problem solving and modeling. In F. K. Lester (Ed.),
Second Handbook of Research on Mathematics Teaching and Learning (Vol. 2, pp. 763-
804). Charlotte, NC: Information Age Publishing.
Lester, F. K., & Kehle, P. E. (2003). From problem solving to modeling: The evolution of thinking
about research on complex mathematical activity. In R. Lesh & H. H. Doerr (Eds.), Beyond
Constructivism: Models and Modeling Perspectives on Mathematics Problem Solving,
Learning and Teaching (pp. 501-518). Mahwah, NJ: Erlbaum.
Mercer, N., & Littleton, K. (2007). Dialogue and the Development of Children’s Thinking: a
Sociocultural Approach. London and New York: Routledge.
National Council of Teachers of Mathematics (2000). Principles and Standards for School
Mathematics. Reston, VA: Author.
Papert, S. (1980). Mindstorms: Children, Computers, and Powerful Ideas. New York: Basic
Books.
Pellegrino, J. W., Chudowsky, N., & Glaser, R. (2001). Knowing What Students Know: The
Science and Design of Educational Assessment. Washington, DC: National Academy Press.
Roschelle, J. M., Pea, R. D., Hoadley, C. M., Gordin, D. N., & Means, B. M. (2000). Changing
how and what children learn in school with computer-based technologies. The Future of
children 10(2), 76-101.
Schoenfeld, A. H. (1985). Mathematical Problem Solving. New York: Academic Press.
Sfard, A., & Leron, U. (1996). Just Give me a Computer and I will Move the Earth: Programming
as a Catalyst of a Cultural Revolution in the Mathematics Classroom. International Journal
of Computers for Mathematical Learning, 1(2), 189-195.
Suppes, P., & Morningstar, M. (1969). Computer-assisted instruction. Science 166(903), 343-350.
Threlfall, J., Pool, P., Homer, M., & Swinnerton, B. (2007). Implicit aspects of paper and pencil
mathematics assessment that come to light through the use of the computer. Educational
Studies of Mathematics 66(3), 335-348
Van Amerom, B. A. (2002). Reinvention of early algebra: developmental research on the
transition from arithmetic to algebra. Utrecht: CD-ß Press, Center for Science and
Mathematics Education, Utrecht University.
Van den Heuvel-Panhuizen, M. (2007). Changes in purpose, content and format: Trends in
assessment in mathematics education. In B. Choksi & C. Natarajan (Eds.), The epiSTEME
Reviews, Vol. 2, Research Trends in Science, Technology and Mathematics Education (pp.
241-266). New Delhi, India: Macmillan.
Van den Heuvel-Panhuizen, M., Bakker, A., Kolovou, A., & Elia I. (in preparation) Mathematical
problem solving by high achieving primary school students.
Wegerif, R., & Dawes, L. (2004). Thinking and Learning with ICT. Raising Achievement in
Primary Classrooms. London and New York: Routledge Falmer.

90
THE COMPUTER TOOL FOR VERIFICATION HYPOTHESES IN PARAMETRICAL PROBLEMS
SOLVING
D. Mantserov, D. Petrichenko, S. Pozdnyakov
Summary
The article considers various aspects of using verification environment (so-called «Verifier») to sup-
port students’ activity in parametrical problems solving. The Verifier was created to support high
school students in their work with functions. It compares student’s answer with true one and shows
her counterexamples accompanied by their graphs and comments. So with Verifier’s support one may
improve her own solution applying various conjectures. A student should consider various conditions
and different cases; therefore the Verifier’s answers have to be given in a complex logical form, which
was a difficulty we got over in our work. We believe that our approach opens the way for expanding
types of problems, which will be useful for studying function properties.

The study and research questions. Relevant features of the theoretical

framework
Problem solution is important part of mathematics teaching methodology. In the process of finding a
solution to the problem some features of creative thinking dynamics are shown (Wertheimer, [1]):
- Collision with a problem;
- Ambiguity, incompleteness of a situation;
- Refinement of infringement areas;
- Using operations to change situation.
The solving process also includes conjectures making and partial solution construction (see, e.g. Polya
[3, 4]).
The heuristics play an important role in mathematical problem solving (see, e.g. Schoenfeld [2]). It is
interesting to consider possibilities for technological support of heuristic activities. For doing this we
should base on the
- essence of didactic task;
- psychology of intellectual activity;
- computer tools usage traits or patterns;

The main question for our research is:

How can we support productive activity of students in solving mathematical problems with logically
complex solutions via computer tools usage?
To find the answer we will explore an idea of providing a student with examples and counterexamples
during the solution process.
We will use M. Minsky conception [5] of two mind mechanisms for combining separate facts in new
entity:
1) creating new abstraction for these facts;
2) mechanical combining of facts in new collection.
In our research we will use both mechanisms: the first one for human-computer interaction, and the
second one for organizing computer support.

91
In problem solving process the students should construct their answers in the predicate form, so that
the Verifier could check them against the
set of existing examples. But one may ask
why should they be of such importance
for a student? First of all, it is so because
abstract conceptions in human mind are
closely connected with their examples
("by default") [5] and it is possible to
form the new concepts only through con-
sidering “pros” and “contras” [6].
The graphical interface of Verifier is
based on classical teaching routine - the
"IRE sequence", which consists of three
steps: a teacher initiates an interaction, a
student responds then, and the teacher
evaluates the response (see, e.g. Sinclair
& Coulthard, 1975 [7], Cazden, 1988 [8]).

Design and methodology

Object classes and their properties description
Verifier’s tasks are based on parametric classes for functions such as y=kx+b, y=a|x-b | + | x-c |,
y=cxm/n, y=logak (x-b), y = ax + kbx (Verifier’s environment allows to construct new classes). Every
class is connected with the set of examples associated with parameters' values - which are interpreted
by the Verifier to show function formulae and graphs.
Each task related to a function class may be interpreted as the class property.
Let's consider a class typical for studying in 10-11 grades of high school in Russia. It is the class of
logarithmic functions which is determined by the formula y=loga(bx+c) and a predicate which restricts
the domain of parameters, for example, a>0 & a≠1.
Various properties can be formulated as predicates of class parameters a, b, c.
Examples.
1) “y=loga(bx+c) is increasing function”: (a>1 & b>0) OR (a<1 & b<0).
2) “A domain of y=loga(bx+c) is all positive numbers”: b>0 & c=0.
3) “An intersection of function graph with y-axe is lying in upper semiplane”:
(a>1 & c>1) OR (a<1 & c<1 & c>0).
4) “Function graph neither has points inside the III quarter of coordinate plane nor on the quarter
boundaries”: (a<1 & b>0 & c<1 & c>0) OR (a>1 & b<0 & c>1).
One may note that the last condition may be rewrtten in other way:
((a<1 & b>0) OR (a>1 & b<0)) & ((a>1 & c>1) OR (a<1 & c<1 & c>0)),
It means that this property can be expressed through previous ones:
“Function graph neither has points inside the III quarter of coordinate plane nor on the quarter
boundaries” can be written as
(“y=loga(bx+c) is increasing function”) & (“Function graph neither has points inside the III quarter of
coordinate plane nor on the quarter boundaries”)

92
The last example shows that after describing some set of properties via predicates we can verify the
connections between these properties.

Examples of interactive answer detailing for Verifier problems

Let’s consider two basic examples and demonstrate various logical paths while solving problems with
Verifier.
Problem. Find conditions for fraction power function y=cxm/n with the range of all positive real num-
bers (m and n are coprime numbers, n>0).
First trial
Presumable student reasonings. «We know that fraction power function y=xm/n has only positive
values, therefore c must be positive».
Input: c>0
Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies your condition but doesn't satisfy the task condition.
Judgment: "We need to narrow fraction power function range. So we
must narrow power function’s domain. To narrow fraction power
function’s domain we need to make it integer power function".

Second trial

Input: c>0 & n>1

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies your condition but doesn't satisfy the task condition.
Judgment: «Yes, of course. The function must not be defined for
x=0 therefore m must be negative».

Third trial

Input: c>0 & n>1 & m<0

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies task condition but doesn't satisfy your condition.
Judgment: "Really. Narrowing fraction power function domain can
be done without narrowing the function domain. The function take
the same value twice. Let's return to our hypothesis of integer
power functions".

93
Fourth trial

Input: c>0 & m<0& n1

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies your condition but doesn't satisfy the task condition.

Judgment: "It is clear the function must be odd. Let's try to check
the idea by taking m=-2".

Fifth trial

Input: c>0 & m<0 & n1 & m=-2

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies task condition but doesn't satisfy your condition.

Judgment: "Yes the conjecture proves to be true. To make function

even we must take even m".

Sixth trial

Input: c>0 & m<0 & n1& m_is_even

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies task condition but doesn't satisfy your condition.

Judgment: "Ok. It appears that first idea works for functions with
domain of all positive numbers, and the second one works for
functions with domain of all numbers except zero. We must com-
bine both ideas in a more accurate way".
Seventh trial
Input: c>0 & m<0 & (n>1 | n=1 & m_is_even)

Output. Congratulations! It is not possible to find difference be-

tween your answer and true one using accessible examples.

Let's consider another way of deduction while solving the same

problem (it was proposed by a student).

94
Problem. Find conditions for fraction power function y=cxm/n to have range of all positive real num-
bers (m and n are coprime numbers, n>0).
First trial
Presumable student reasonings. «We know that fraction power
function y=xm/n has only positive values therefore c must be posi-
tive».
Input: c>0

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies your condition but doesn't satisfy the task condition
(here we see another counterexample).

Judgment: "Why this example doesn't satisfy the task condition? The range does contain zero! How to
eliminate it? We need to do power parameter negative!".
Second trial
Input: c>0 & m<0

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies your condition but doesn't satisfy the task condition
(here we see another counterexample).

Judgment: "It is not enough to eliminate zero value from the

range! We must eliminate all negative numbers too! Therefore the
power parameter must be even".

Third trial
Input: c>0 & m<0 & m_is_even

Output. Your answer has a mistake. Look at the picture. This func-
tion satisfies task condition but doesn't satisfy your condition.

Judgment: "I forgot that the power parameter can be a fraction!

We need to separate integer power parameters and fraction power
parameters".

Fourth trial
Input: c>0 & (m<0 & n>1 | n=1 & m<0 & m_is_even)

Output. Congratulations! It is not possible to find difference be-

tween your answer and true one using accessible examples.

95
Using Verifier to support problem reformulation

As M. Minsky noted [5], the reformulation of problem is one of the most important intellectual tools
for its solving. For our purposes we will consider the process of reformulation as a process of knowl-
edge representation forms alteration (Pozdnyakov, 1995, [9, 10]). For article's subject we will use four
basic forms of knowledge representation:
- algorithmic representation (formulae)
- predicative representation
- representation by objects properties
- representation by examples collection.
So there are many types of transitions from one form of representation to another.
The partial list of such transitions:
- from formula to graph
- from graph and formula to properties
- from the properties of a “class” object to a concrete example of this object with such properties
- from some set of properties to new one
- etc.
Beneath we will provide for analysis of some types of problems based on transitions from one form of
knowledge representation to another.
1. Type “from function graph and formula to func-
tion properties”

The type is characterized by giving the full

information to a student about the object under con-
sideration (function in our case). We give the set of
examples and counterexamples of functions to a
student with their formulae and graphs. The prob-
lem is to find an appropriate classification by for-
mulating the property which separates examples
and counterexamples.

2. Type “from function graph to function

properties”
In this type of tasks we narrow the informa-
tion about an object and provide the students with
function graphs without any formulae.
The problem for student is to find an appro-
priate classification by formulating the property to
separate examples and counterexamples based on
function graphs comparing.

96
Student must be aware of function graphs properties to succeed in solving this type of problems.
3. Type “from function graph and formula to
predicate”
By stating these problems we may form the
skills needed for solving inequalities. A student can
search for a predicate (it is the same as to describe a
solution by a set of inequalities) to specify a set of
objects. Formulae here play the role of additional
help in solution search.

4. Type “from function formula to predicate”

Such type of the problems is characterized by
function representation in a formula form. The
search of solution requires a comparison between
concrete formula and its class formula. This type of
problems helps to train the skills of generating con-
crete examples for a common class and formulate
characteristic properties of the object set.

5. Type “from function graph to predicate”

This is most important type of problems
which combine common skills for information
translation from one form to another with concrete
math skills of exploring functions properties, solv-
ing equations and inequalities.

6. Type “from function formula to function

properties”
The problems of such type are closely connected
with translation of information from formulae
language to verbal form. By such type problems
we form math skills of object's "recognizing".
For example: "what is an essential feature of
function formula (from class given) to be even"
or "what can be said about its domain.

97
 Further implications
There are other types of problems, so here we provide their non-exhaustive list. They imply the
examples construction in a symbolic or graph form.

7. Type “from predicate to func-

tion graph”.

8. Type “from predicate to func-

tion formula”.

9. Type “from predicate to func-

tion graph and formula”.

10. Type “from function proper-

ties to function graph and formula”.

Research results and discussion

Teachers’ answers on research questions

Beneath we cite some typical teachers’ answers which applied Verifier in their practice.

1. Is the material aimed to achieve new educational results?

"Yes. Students have learned to read attentively the problem description and to analyze it, noticing that
problems, which seem to be identical, are different in fact".
2. Is it possible that Verifier usage in educational process have raised the level of student’s educa-
tional independence and self-activity?
"Yes, it is possible. The Verifier's problems put the students in a situation of self-control and self-
estimations of actions".
3. Is it possible to state that using Verifier in educational process motivates students to learn math?
“Yes it is. The Verifier's problems are rather difficult and "time expensive", nevertheless students
solved plenty of them, while working at home”.
4. What type of changes in real educational process was brought with the use of Verifier and were
these changes necessary and effective?
“We didn’t feel any need in such changes”.
5. Does Verifier promote group and individual research activity in educational process?
“Yes. Verifier promotes different forms of group and individual research work”.
6. Does Verifier minimize labor expenses of the teachers when they prepare materials for lessons?
“Yes”.

98
7. Is Verifier suitable for schools, assuming it fits the hardware available at schools?
“Yes”.
8. What impact does Verifier have on quality of knowledge and skills?
“Students began to solve equations and inequalities with parameters more confidently. They also
started to show awareness of function graphs properties”.

Conclusion
The results of experiments with Verifier for supporting problems solving approve the authors’ posi-
tion that verification environment can be used by a student as a tool to support his thinking about the
complex calculus problems.
This conclusion is in good correspondence to works of L. Vygotsky in psychology about the role of
tools in intellectual skills forming [11].

References

1. M. Wertheimer. Productive Thinking.

2. Schoenfeld, A. (1985). Mathematical Problem Solving. New York: Academic Press.
3. G. Polya. How to Solve It.
4. G. Polya. Mathematical Discovery.
5. M. Minsky. Society of mind.
6. P. Winston. Artificial intelligence.
7. Sinclair, J., & Coulthard, M.: 1975, Towards an analysis of discourse, Oxford University Press,
London.
8. Cazden, C.: 1988, Classroom Discourse: The Language of Teaching and Learning, Heinemann,
Portsmouth, NH.
9. M. Bashmakov, S. Pozdnyakov, N.Reznik. Information environment of teaching process (rus.).
“Svet”, Saint-Petersburg, Russia.
10. Pozdnyakov S. Mathematical models of intellectual dialogue in Computer Aided Learning. World
Conference on Twenty-First Century Mathematics. Proseedings. School of Mathematical Sciences,
GC University, Lahore, Punjab (Pakistan), 2005.
11. Vygotsky, L.S. (1962). Thought and Language. Cambridge, MA: MIT Press.

99
 THE DECISION-MAKING AS A SCHOOL ACTIVITY
María Candelaria Espinel Febles
University of La Laguna (mespinel@ull.es)
Ana Teresa Antequera Guerra
IES Luis Cobiella Cuevas (aantegue@yahoo.es )

ABSTRACT
A problem is chosen from the PISA 2003 Report, and then extended to include questions
involving a choice of preferences and decisions. These questions are posed to secondary school
students aged 15 to 16, whose problem-solving and heuristic decision-making skills are then
analyzed. We discovered that students have difficulty using mathematics when assigning a
weight to achieve an objective, though they are able to recognize the functional use of the rules
of choice used in society.

INTRODUCTION
Some of the real world choices an individual may have to make require the use of mathematical
calculation tools in order to establish a preference or analyze a situation. Multi-criteria analysis
constitutes a way to model decision processes and involves the decision to be made, any
unknown events which may affect the results, the possible courses of action and the result(s) of
said actions. Using multi-criteria models, the decision maker assesses the possible implications
each course of action may entail so as to obtain a better understanding of the relationships
between the actions and the objectives. There are various mathematical procedures for
summarizing the values yielded by each alternative with respect to all the criteria considered in
the analysis. The best known mechanisms are those which use linear weighting (scoring), i.e. the
simple sum of each attribute’s contributions. This is a very fruitful field of Operational Research
in which improvements and variants are constantly being developed, along with growing
applications in various contexts (Berumen and Llamazares, 2007).
For the specific case of a discrete decision problem, the information available is realized through
an evaluation of n attributes for a set of m alternatives, and presented in a double entry matrix in
which the attributes are shown as columns, while the associated alternatives appear in the rows.
The decision is made by fixing the criterion to be used in deciding the best solution (Romero,
1993).

In “The Best car” activity in PISA 2003 (MEC, 2005), a decision problem is posed using a data
table with four attributes and five alternatives. It then asks for the value of a linear function to be
calculated which ranks the alternatives, so that a linear function can then be weighted to yield a
specific objective. We broadened this activity, asking that the best car be chosen by applying two
rules, the first taking into account the number of first places, and the second by deleting the high
and low values and adding the remaining scores.
In this paper we analyzed the students’ results for the activity proposed. The study was done with
secondary education students (abbreviated ESO in Spanish, Obligatory Secondary Education),
most of them around 15 years of age.
The objectives guiding our research were:
. To see whether the students understood the decision rules and applied them correctly. The
problem was easy to understand since the context, in which the problem was presented, to
choose the best car, is familiar to most youngsters.

100
. To see if the students could identify other situations or settings where the different rules of
choice presented in the activity, or other applicable rules, can be used.
. To observe whether the students used a mathematical process in other contexts requiring that a
choice or decision be made, since the goal was to promote the development of the student’s
ability to take a critical stance given the rules of choice that are applied in society.
. To verify the validity of the activity designed via a questionnaire on fair decisions and well-
founded judgments. This problem is part of wider research into the design and implementation of
activities involving game theory and negotiation models using mathematics.

Theoretical framework
The salient points and the framework guiding our research were oriented around the studies
conducted by the Organization for Economic Cooperation and Development (OECD) on the use
of mathematics by students, which in the case of Spanish students shows their unwillingness to
think beyond applicable procedures or routine problems in mathematics.
In Spain, one of the curriculum changes made to secondary education in 2007 involved the use
of competencies in the curriculum. Mathematical competencies are understood as an individual’s
ability to use mathematics to meet his needs as a constructive, committed and thinking citizen.
This is expedited through the promotion of activities that involve the problem solving and
discrete mathematics.
At the present time, problem solving is provided as part of the primary and secondary education
curricula. A problem is defined as a conflictive situation that requires a solution for which an
explicit procedure is not known beforehand. Many of the text books more commonly used by
school children include guidelines for the solving of problems, almost always following the
phases described by Polya (1945). These phases encourage the use of heuristics, considered as
exploratory methods or algorithms used in the problem-solving process in which the solution is
found by evaluating the progress made in the search for the final result. This is why heuristics are
“golden rules,” conjectures, intuitive solutions or simply common sense. A problem is no more
than a tool for thinking mathematically (Schoenfeld, 1992) and requires training individuals who
can think for themselves and who can critique and reflect on the solutions.
The curricular content dedicated to problem solving, mandatory until the age of 16, offers the
chance to work with discrete mathematics. This subject is, however, not well-known by
practicing teachers, and the lack of available material and activities further serves to prevent this
subject from being taught to the same level as in other countries (DeBellis and Rosenstein,
2004).
Discrete mathematics is one of the branches of mathematics which has seen the most progress in
the 20th century as a result of computers. Different learning centers and associations, such as the
Freudenthal Institute (Doorman et al., 2007), the National Council of Teachers of Mathematics
(NCTM), and the American Mathematical Society (AMS) have promoted its incorporation into
curricula to include experiments and research involving mainly the use of graphs, matrices and
combinatory, all of which are very useful in today’s mathematics (Kenney and Hirsch, 1991;
Rosenstein et al., 1997). Material and ideas for its development have also been provided by the
arrival of texts and projects such as COMAP (1988), based on a set of examples of the
applications of mathematics most relevant to everyday life. The past decade has seen a
proliferation of textbooks (Parks et al., 2000) and web pages along these same lines, such as
http://www.dimacs.rutgers.edu.

101
Some Mathematics Education researchers regard discrete mathematics as a chance to revitalize
mathematics in school (DeBellis and Rosenstein, 2004; Rosenstein et al., 1997). They see it as a
chance for innovation and an opportunity to discover problems that are out of the ordinary
(Goldin, 2004).
The problems associated with this field can also be used to build mathematical knowledge and to
model situations, which help the student understand and control the world around him.
Mathematical modeling can be viewed as a means for linking problem solving to the real world.
While the term mathematical modeling itself may be in dispute, within mathematics it can be
used to solve real problems and to include students in some of the phases involved with the
modeling tasks (Stillman et al., 2007). The trend or perspective of models for teaching
mathematics and solving problems is to teach real life situations in the classroom (Lesh and
English, 2005).
It is within this framework and considering the points mentioned (development of mathematical
and problem-solving skills and modeling of everyday problems) that we developed the decision-
making activity involving several alternatives in a set context.

DESCRIPTION OF THE STUDY AND RESULTS

Design of the study

This paper details the results of a study conducted with 72 secondary school (ESO) students
divided into four groups, with 35 third-year students in two groups, and 37 fourth-year students
in the other two. Most of the students were between 15 and 16 years old. Also included are the
results of an interview with a group of fourth-year students asked to justify their answers.
The activity used, “The best car,” was taken in part from PISA 2003 (MEC, 2005). A discrete
decision problem was formulated by means of the data table with four attributes and five
alternatives shown below. The students were asked to calculate the value of a linear function
which ranks the alternatives so as to then weight a linear function and obtain a specified
objective. The PISA activity was expanded by asking that the best car be chosen considering two
rules (rule 1: number of top rankings and, rule 2: eliminate highs and lows and add).

Fuel Exterior Interior

Safety
Car consumption design space Score
(S)
(C) (D) (H)
Ca 3 1 2 3
M2 2 2 2 2
Sp 3 1 3 2
N1 1 3 3 3
Xk 3 2 3 2

The scores are interpreted as follows:

3 points = Excellent 2 points = Good 1 point = Acceptable

For the student, the activity had three parts:

. Question (a) consisted of evaluating a given simple linear function and choosing the best car:
Total score = (3x S) + C + D + H

102
. Question (b) required creating a weighted linear function to achieve a specific objective:
Total score = …. S +…. C +…. D +…. H,

which yielded alternative Ca as the best car.

. Question (c) required the students to apply two rules to select the best car according to each
rule:
Rule 1: number of first place rankings,
Rule 2: sum of scores after eliminating highs and lows.

Analysis of the information

For the analysis of the results, the three questions in the activity were broken down into seven
coded items as follows:
T = Complete table, A = Value of a linear function,
B = Assign weights to a linear function to achieve the objective,
CR1 = “First place” rule, CR2 = “Eliminate highs and lows” rule,
CE1 = “First place” choice, CE2 = “Eliminate highs/lows and add” choice.

The results of the data analysis are shown in four sections. The first includes the results on the
success of each of the items. The second and third sections are dedicated to question (b) on the
design of a function, and detail the results and strategies used by the students. The third section
offers a comparison of the 72-student sample with the results of PISA for Spain and the OECD
for questions (a) and (b). Along with the results there is a qualitative analysis of the interviews
with one of the groups that took part in the study and whose students were asked to explain their
solutions to the different questions in the activity, to express their opinions about the rules, to cite
other situations where these rules are applicable and to name other rules of choice.

1. Success of each of the seven items

Table 1 shows the percentage of students who correctly solved each of the seven items
comprising the three questions.

Table 1: Percentage of students answering each item correctly

T A B CR1 CR2 CE1 CE2
Ítem
(a) (a) (b) (c) (c) (c) (c)
% 93 100 57 96 93 83 82

The results of the study with 72 students yielded a high percentage of correct answers. There
were no noticeable differences between the groups of third- and fourth-year students. Ninety-
three percent of the students completed the table they were given with the score for each car
(item T). All the students managed to correctly solve question (a) and apply the rule, i.e. to find
the value of a linear function (item A). Most, over 90%, applied rule 1 (CR1) and rule 2 (CR2).
Over 80% also chose the correct car by applying the rules. The biggest difficulty was posed by
question (c) on the weighting required to have a given car (Ca) be the best. Slightly over half of
the students answered this question correctly.

103
The activity shown here and consisting of 7 items was part of a larger questionnaire with 23
items on game theory and decision making, and which was intended to analyze the development
of critical thinking and reasoning skills in secondary school students. The validity of the activity
was determined using the Rasch methodology (Linacer, 2007), which provides a correlation
between students and items and which confirms that assigning or selecting weights so as to
achieve a desired result is difficult when compared to the other items formulated in the
questionnaire (www.ince.mec.es/pub/pisamanualdatos.pdf pp. 64-81).
We now present a detailed analysis of this question.

2. Results of the question to weight a linear function

To analyze item B (assign weights to a linear function), we established a coding system for
Blank, Wrong, Acceptable and Right answers, where “Acceptable” was used when the assigned
weights yielded an equal score between the target car and another car. Table 2 shows the number
of answers for each category.

Table 2: Results for questions (b)

Categories Blank Wrong Acceptable Right
N = 72 2 18 11 41

In the results in Table 2, note how the majority of students, 41, answered correctly. But one
fourth of them, 18, gave a wrong answer and were unable to assign weights to a linear function
so as to achieve the desired objective. If to these we add those students with blank or acceptable
responses, a total of 44% of the students failed this question.
As concerns question (b), the reasons the students interviewed gave for choosing the weights are:
. They assigned a higher weight to the higher scores
. They assigned weights at random
. They used trial and error.
We believe the failure of these students, who are nearing the end of their obligatory schooling, to
come up with a weighted linear function to yield a given objective could have negative
repercussions in their future lives as members of society.
Next we analyze the correct and incorrect strategies used by students when assigning weights in
this question.

3. Listing and analyses of strategies for linear weighting

Table 3 summarizes the strings of four weights provided by the students, and assigns them to the
Right, Acceptable and Wrong categories, along with the value of the object function for each
attribute or car and the number of students offering said choice. When a cell shows several
weights, the number of students choosing those weights is shown instead of the values.

Table 3: Answer patterns according to strategies for achieving an objective

Grade Weight Total Scores Frequency
Right 3113 21, 16, 19, 18, 20 9
3114 24, 18, 21, 21, 22 7
3124 26, 20, 24, 24, 25 4
3116 30, 22, 25, 27, 26 4

104
 4114 27, 20, 24, 22, 25 3
2113 18, 14, 16, 17, 17 2
4214 28, 22, 25, 25, 27 2
5125;5215 4
3003;2102
4113;5113 2
3126;4216 4
1002;4236
3127 23, 18, 22, 21, 23 5
2123 20, 16, 19, 20, 20 2
Acceptable
4123;6123 4
6 2 4 6 ;15 1 2 3
3233;3323 9
2342;3383
2152;2322
3223
Wrong
9223;9323 9
9332;4232
4221;3132
1113

Of the 72 students, over half, 41, chose the proper weights which, as shown in Table 3, are
needed to have the first value be the highest of the five. A total of 17 different weights were
given as answers. The most frequent was (3 1 1 3), given 9 times, which results in a score of 21
for car Ca, and (3 1 1 4), appearing 7 times and resulting in a score of 24.
The most frequent strategy was to assign high values to S and H and low values to C and D. Note
how some students changed the last 3 (weight of H) to a 4, possibly to obtain a larger difference
between the two highest scoring cars. This same strategy seems to have led to other weightings
such as (3 1 1 6) and (3 1 2 6).
The 11 cases graded as acceptable correspond to ties, and resulted from weights such as (3 1 2 3)
or (2 1 2 3). Some students assigned weights to the attributes only to check the results with the
alternatives they believed to be the most problematic, which may be why they failed to notice the
tie.
The 18 students who answered incorrectly assigned weights like (3 2 3 3), (9 2 2 3) and other
similarly nonsensical or absurd answers.
As already noted, it is troubling how so many students at or near the end of their mandatory
schooling were unable to assign weights to achieve an objective. Differently stated, they were
unable to analyze and work with data given to them so as to obtain the desired result.

4. Comparison with the PISA results

Table 4 shows the percentage of right answers for questions (a) and (b) given by the 72 students
in the study, along with the results from Spain and the OECD. For question (a) we considered the
result for item T, since in PISA the results of items T and A are combined.

105
 Table 4: Comparison with the PISA results
ESO Spain OECD
N=72 PISA PISA
Question (a) 93% 71.4% 72.9%
Question (b) 57% 22.2% 25.4%

For the 72 ESO students, the results are noticeably better than those for the rest of Spain and also
than those for other OECD countries. For question (b), however, on the assignment of weights or
the prioritization of alternatives to achieve a specified outcome, all three gave unsatisfactory
results. The task involves an important mathematical process and evidences a lack of conceptual
and algorithmic learning in those countries where the test was administered.
During the interview, the students were asked to give their opinions about the rules and to name
other situations where these selection rules are applicable. In their answers, several commented
that the rules applied to question (c) did not seem fair since both resulted in the least safe car
being given the highest score, although they do accept that, in the case of cars, it makes sense to
pick the car with the highest score. Moreover, they were able to identify these two rules with real
situations and to come up with contexts, such as sports competitions, specifically synchronized
swimming, ice skating, diving, dance contests, surfing and some television game shows, where
said rules apply. They had problems inventing rules however, and most simply proposed
summing the scores.
It was also apparent from the interviews that this was a new problem for them. They did not, for
example, associate the scenario with functions, i.e. they did not see the activity as involving
mathematics, such as finding the value of a function, and so they were not aware of any transfer
of knowledge (Santos, 1997). They approached the problem as a novelty that they had never
encountered in mathematics class.

CONCLUSIONS AND REFLECTIONS

The results from the study with third- and fourth-year secondary school students showed how
they all completed the table with the scores for each car. They carried out a routine procedure
given a direct instruction. We found that they understood the decision rules and, as expressed by
those interviewed, were willing to apply them to their lives, though they did not see a connection
to mathematics. Almost half of the students in our sample experienced difficulties in finding the
right weights, and also failed to use the proper notation. In brief, it may be said that the students
were able to read a table and to find the value of a function, but faced serious difficulties when
assigning weights to reach a stated objective and said they knew of no method that would
guarantee success.
The conclusions drawn from the data for the countries participating in PISA are a cause for
reflection, in that while the students know how to apply rules to a data table, they do not know
how to manipulate the data to reach a specified goal. It may be deduced from the OECD data,
which agree with our own, that certain aspects of the educational system as it now stands are
failing. Students do not have a procedure or a linear strategy they can apply, as they do when
calculating an average or an expected value. Other studies (Cobo and Batanero, 2004) have
shown that a large number of students experience difficulties when calculating and interpreting
weighted averages in statistics. The correct weight when calculating an average involves the
ability to apply the distributive property when adding a set of numeric values. Another valid
analogy to demonstrate this problem is the “knapsack problem” or the “subset sum problem,”

106
where given a weight or sum, the appropriate items have to be selected to fill the knapsack. This
problem is NP-complete, but there are good heuristics that could be taught (Espinel, 1995). It
seems students are not taught methods for selecting weights or values from among a given group
to reach a total sum or weight.
We are currently designing and improving this and other activities so as to achieve the stated
objectives of our research.
Of the three aspects considered in PISA (content, situation and competence), the “Best car” task
is considered a public situation. It requires that the students resort to their mathematical
understanding, knowledge and skills to evaluate the aspects of an external situation with
repercussions in public life. Those competences or processes involved in the problem
development are thinking, reasoning and argumentation (Niss, 2002). This is, in our opinion, a
rich activity to take into the classroom (DeBellis and Rosenstein, 2004) since it covers a wide
variety of mathematical fields, such as social choice, voting methods, social justice and the
search for fair rules, and decision-making theory and the search for an optimum decision from a
set of alternatives.

Note: Part of our research was funded by the SEJ2006-10290 research project (Ministry of
Science and Technology, Madrid, National Plan for Research & Development & Innovation).

Referentes:

Berumen, B. A., Llamazares, F. (2007). La utilidad de los métodos de decisión multicriterio

(como el ahp) en un entorno de competitividad creciente. Cuad. Adm. Bogotá
(Colombia), 20 (34): 65-87. On line:
http://www.scielo.org.co/pdf/cadm/v20n34/v20n34a04.pdf
Cobo, B., Batanero, C. (2004). Razonamiento numérico en problemas de promedios. Suma.
Revista sobre la enseñanza y el aprendizaje de las matemáticas. 45, 79-86.
COMAP (1988). For All Practical Purpose. S. Garfunkel (Project Director) et al. New York:
W.H. Freeman. (“Las matemáticas en la vida cotidiana”. Addison-Wesley/Universidad
Autónoma de Madrid, 1999).
DeBellis, V.A., Rosenstein, J. (2004). Discrete Mathematics in Primary and Secondary Schools
in the United States. ZDM, 36 (2) 46-55.
Doorman, M. Drijvers, P., Dekker, T, Van den Heuvel-Panhuizen, De Lange, J. Wijers, M.
(2007). Problem solving as a challenge for mathematics educations in The Neatherlands.
ZDM Mathematics Educations, 39, 405-418.
Espinel, M.C. (1995). Optimizar cantidades. Revista de Didáctica de las Matemáticas. UNO, 3,
116-122.
Goldin, G. A. (2004). Problem Solving Heuristics, Affect, and Discrete Mathematics. ZDM . 36
(2), 56-60. Holland.
Kenney, M. J., Hirsch, C.R. (Eds) (1991). Discrete Mathematics Across the Curriculum, K-12.
Inc, Reston, Virginia: NCTM.
Lesh, R., English, L.D. (2005). Trend in the Evolution of Models & Modelling Perspectives on
Mathematical Learning and Problem Solving. ZDM. 37 (6), 487-489.
Linacre, J. M. (2007). Realiability and Separations. A User´s Guide to Winsteps/ Ministep Rasch
– Model Computer Programs. Chicago: Winsteps. Com www.winsteps.com
MEC (2005). PISA 2003. Pruebas de Matemáticas y de Solución de Problemas. Inecse. Madrid.

107
Niss, M. (2002). Mathematical competencies and the learning of mathematics: The Danish Kon
Project. On Line:
http://www7.nationalacademies.org/mseb/mathematical_competencies_and_the_learning
_of_mathematics.pdf
Parks, H., Musser, G., Burton, R., Siebler, W. (2000). Mathematics in Life, Society, & the World.
New Jersey: Prentice Hall.
Polya, G. (1945). Cómo plantear y resolver problemas. México: Trillas
Romero, C. (1993). Teoría de la decisión multicriterio: Conceptos, técnicas y aplicaciones.
Madrid: Alianza Universidad Manuales.
Rosentein, J. G., Franzblau, D. S., Robert, F. S. (1997) (Eds). Discrete Mathematics in the
Schools, DIMACS Series in Discrete Mathematics Computer Science, Volume 36,
Providence, RI: American Mathematical Society (AMS).
http://dimacs.rutgers.edu/Volumes/Vol36.html
Santos, L. M. (1997). La transferencia del conocimiento y la formulación o rediseño de
problemas de aprendizaje de las matemáticas. Revista Mexicana de Investigación
Educativa, 2, 3, 11-30.
Schoenfeld, A.H. (1992). Learning to think mathematically: Problem solving, metacognition,
sense-making in mathematics. In: D. Grouws (Ed.) Handbook for Research on
Mathematics Teaching and Learning (pp.334-370). New York: Macmillan.
Stillman, G., Brown, J., Galbraith, P., Edwards, I. (2007). A Framework for Success in
Implementing Mathematical Modelling in the Secondary Classroom. Watson, J.,
Beswick, K. (Eds) Proceeding of the 30th annual conference of the Mathematics
Education Research Group of Australasia. 2, 688-697. Australia.

108
 COGNITIVE AND METACOGNITIVE PROCESSES OF PRE-SERVICE
MATHEMATICS TEACHERS WHILE SOLVING MATHEMATICAL PROBLEMS
By
Omar Hernández Rodríguez
Wanda Villafañe Cepeda
University of Puerto Rico, Río Piedras Campus

A phenomenological study about mathematical problem solving is described. Eight pre-service mathematics
teachers participated; six were studying to become teachers at elementary school -4th to 6th grades- and two at high
school -7th to 12th grades-. The data was obtained through long interviews, thinking out loud problem solving
sessions and retrospective interviews that took place immediately after the problem solving sessions. The objective
of the long interview was to determine the participants' beliefs and declarative knowledge about this topic. The
objective of the problem solving sessions was to determine the type of representation, strategies, and control
processes that the participants use when solving problems. During the retrospective interview, the participants had
the opportunity to reflect about their performance. These techniques allowed the investigators to obtain a
comprehensive description of the phenomenon.
INTRODUCTION
Problem solving is recognized by some experts as a fundamental process for the
mathematical development of students. Some of the most influential organizations in
mathematics education have recognized the importance of problem solving in school
mathematics (National Council of Teachers of Mathematics, 2000; American Association for the
Advancement of Science, 1993). The Mathematics Program of the Department of Education of
Puerto Rico, states in its mission that solving problems is important as an aim and a mean for the
learning of mathematics. (DEPR, 2003).
Regardless of the recommendations and the importance recognized by experts and
educational organizations, what happens in the mathematics classrooms is completely different.
After finishing their education, some teachers ignore what they learned, and their practice
reflects more their experience as students rather than what they were taught in their pedagogy
classes (Skott, 2001, as cited by Liljedahl, et al. 2007). Some authors ascribe this situation to the
teachers’ knowledge about the discipline (Ball, 1990; Leonard and Joergensen, 2002; Van
Dooren, et al., 2003); others, to teachers’ affective and metacognitive factors, including their
beliefs (Grows y Good, 2002; Liljedahl, et al. 2007; Mewborn y Cross, 2007).
Education schools prepare their students in three areas: discipline knowledge, knowledge
of the fundamentals of education, and the teaching of the discipline. However, the time dedicated
to the study and development of metacognitive strategies is minimal. We hope that this research
can provide the insight required to make curricular changes, so that the study and development of
the metacognitive processes can be included in the curriculum of prospective teachers.
REVIEW OF LITERATURE
Problem solving has been the object of many researches; however, there are still
unanswered questions about it (Lester, 1994). The study of problem solving started with the
publication of How to Solve It (Polya, 1945), which was followed by a series of research studies
about the effectiveness of the use of general problem-solving strategies to learn mathematics.
Later, research was based on the information processing theory with the objective of determining
how the experts solve problems (Schoenfeld, 1985; Schoenfeld & Herrmann, 1982; Silver &
Marshall, 1990). Recently, with the advent of the constructive learning theory, there has been a
new wave of research that focuses in other aspects, such as the role of the metacognition, the
students and teachers' beliefs, and affective and social influences when solving a problem
(Garafalo and Lester, 1982; Hernández Rodríguez, 2002; Maqsud, 1997; Santos Trigo, 1995;
109
Schoenfeld, 1987, 1989, 1992; Swanson, 1990, 1992). It has been established that, even if the
students possess the mathematical knowledge, it is very difficult for them to use that knowledge
in new situations (Santos Trigo, 1995; Schoenfeld, 1985). Moreover, in few occasions they are
able to use specific strategies to solve mathematical problems (Hernández Rodríguez, 2002). It
also has been concluded that, in addition to the specific knowledge, students need other strategies
to solve problems (Polya, 1945; Santos Trigo, 1995; Schoenfeld, 1985). More specifically, it has
been established that students can benefit more if they learn general problem-solving strategies
(Hembree, 1992; Lawson, 1990; Silver and Marshall, 1990).
Flavell (1976) defined metacognition as the knowledge that people have about their own
cognition and the self-regulation processes of the cognitive processes. Later, this definition was
expanded to include the students’ beliefs about themselves, about mathematics, about the task,
and about the strategies required by a given situation (De Corte, Greer and Verschaffel, 1996;
Garafalo and Lester, 1985; Greeno, Collins and Resnick, 1996; Lampert, 1990; Schoenfeld,
1987). Stemberg (1998) considered that metacognition is part of the human abilities and that it is
indispensable in the formation of the discipline.
Lamper (1990) found that students believe knowing mathematics is about remembering and
applying certain rules correctly in a problem, and that the only correct answer is the one the
teacher gives. These beliefs have a generally negative effect in the way the students perform
when solving mathematical problems. Schoenfeld (1987) stated that students’ mathematical
beliefs are important to help or interfere in the process of problem solving. For example, he
found that students thought that a mathematical problem has to be solved in less than ten minutes
and this belief lead them to abandon it if they do not get a solution quickly.
Problem solving in pre-service mathematics teachers had been studied in different
aspects. Bjuland (2004) did a study with 105 pre-service teachers in which they reflected on their
own learning process when solving geometric problems in a collaborative way. Chapman (2005)
conducted a qualitative study intended to determine the knowledge that pre-service teachers have
about problem solving and the effects of the incorporation of reflection and inquiry processes to
the improvement of this knowledge. Cadenas (2007) did a study intended to determine the lacks,
difficulties and errors that pre-service teachers have in their mathematical knowledge previous
the start of university studies.
Mathematics teachers’ beliefs and their relationships with the student learning process have
been studied by Mewborn and Cross (2007). Teachers’ cognitive and metacognitive preferences
have been studied by Leikin (2003), and Grouws and Good (2002). In regard to pre-service
teachers, Liljedahl, Rolka and Rösken (2007) studied the affective aspects of problem solving,
and Van Dooren, Verschaffel, and Onghena (2003) investigated the evolution of pre-service
mathematics teacher’s cognitive preferences.
Lampert (1990) said that the school has a major responsibility in the development of
student's beliefs about the meaning of knowing mathematics and how to do mathematics. These
beliefs are created after so much time of looking, listening and practicing mathematics in
schools. Since beliefs are mental constructions originated by previous experiences and social
interactions, we can argue that student’s beliefs are mostly influenced by their teacher's beliefs.
Part of the difficulty that students show with problem solving can be explained by their teacher’s
beliefs about it (Goss, 2006; Mewborn and Cross, 2007). At the same time, the mathematics
teachers’ beliefs are the result of their school experiences and the knowledge acquired as
education students. The dialectic interaction between beliefs and professional evidence are the
object of this investigation.

110
 With regard to the representations that are used by pre-service teachers when solving
problems, it has been found that the numeric-table representation dominates over the algebraic
and geometric ones (Presmeg and Nenduradu, 2005). Moreover, Mousolide and Gagatsis (2004)
found that teachers do have difficulties constructing geometric representations.
It is important to remark the importance of mathematic content in the problem solving
process. Some authors claim that a good knowledge of the discipline is a great factor when
dealing with problem solving (Ball, 1990; Cadenas, 2007; Leonard and Joergensen, 2002; Van
Dooren, et al., 2003). Ma (1999) concluded that teachers that have a deep understanding about
mathematical concepts can create a problematic situation related to the mathematical concept
more easily, which implies that a deeper understanding of the mathematical concepts redounds in
a better pedagogical knowledge.
To think about the beliefs and the way that pre-service teachers solve problems will allow
the researchers propose educational environments that promote the construction of beliefs and
knowledge that favor problem solving.
RESEARCH QUESTIONS
This investigation was guided by the following questions:
• What beliefs do pre-service mathematics teachers have about mathematical problem
solving?
• Which kind of external representations (iconic or symbolic) do pre-service mathematics
teachers use when solving a non typical problem?
• What kind of strategies (general or specific) do pre-service mathematics teachers use
when solving a non typical problem and in which circumstances do they use them?
• How does self-regulation intervene during the different stages of problem solving?
DEFINITION OF TERMS
A non typical mathematical problem is a situation that has to be modeled to find an
answer to a question that derives from the same situation and which solution is not
straightforward. (Parra, 1991).
Problem solving refers to the coordination of knowledge, previous experiences, and
intuition in an effort to find a solution that is unknown (Parra, 1991). Operationally, it is the set
of all written and verbal processes used by the student to find the answer to a problem.
The cognitive processes that will be studied in this investigation are the construction of
the problem's representation and the strategy selected and used to solve a problem.
An external representation is a stimulus to the sense, generally in the form of drawings,
diagrams, graphics, models or other formal symbolic systems (Janvier, Girandon and Morand,
1993).
A general strategy is a technique that can be applied to various knowledge domains and
that serves as a guide to solve a problem. Some general strategies are trial and error, finding a
pattern, constructing a table, using analogies, using auxiliary elements and backward working. A
specific strategy is a technique that can be used to solve a problem in a specific domain.
Metacognitive processes include the beliefs and the processes of self-regulation and
control. A belief is a made-up explanation that a person has about a specific field of knowledge
and that determines the way the person conceptualizes and fulfills on it (Schoenfled, 1992). A
belief can be about itself (De Corte, Greer and Verschaffel, 1996), the area of study -
mathematics, in this case- (Greeno, Collins and Resnick, 1996) or the task to be done (Garofalo
and Lester, 1985).

111
 The self-regulation or control is an ordered process used by a person to control its own
cognitive activity, and, in this way, ensure the accomplishment of the cognitive objective
(Schraw and Graham, 1997). A person that can control its cognitive activity can make
predictions, elaborate a plan before starting to solve a problem, pay attention to all the
components of a problem, question the process, value the products and the efficiency of the
execution, and review, change, and abandon unproductive strategies or plans (Garofalo and
Lester, 1985; Schraw and Graham, 1997).
METHODOLOGY
In the present study, a constructivism point of view is
assumed. This considers that human beings build their own knowledge, and that cognitive and
metacognitive processes take part inand that cognitive and metacognitive processes take part in
the construction of such knowledge (Flavell, 1976; Noddings, 1990; von Glasersfeld, 1990). In
addition, it is also recognized that some social and emotional factors intervene in the
construction of knowledge (Greeno, Collins, and Resnick, 1996). To have access to the cognitive
and metacognitive processes, different methodologies were used to describe what was happening
in the minds of the participants. The research design used responds to the necessity of gaining
access to the field of the perceptions to explore what is happening when non typical problems are
solved from the participant's perspective.
This is a phenomenological study about cognitive and metacognitive processes that pre-
service mathematics teachers exhibit about solving non typical mathematical problems. It was
intended to find deep meanings, understandings and attributes of the phenomenological target
that was studied. The meaning that various people ascribe to the concept or phenomenon is
described (Creswell, 1998; Moustakas, 1994), the experiences that people have had with the
phenomenon are explored, and the essential structure or invariant in which underlies the meaning
of the experience is outlined. In this way, the intentionality of the conscience is described where
the experiences contain the external appearance as well as the internal conscience (Moustakas,
1994). From this point of view, the phenomenon gains meaning through people experiences with
it. To have access to the essence, product of the range of interactions that the people have had
with the phenomenon, we analyzed the memory, the image and the significance that they
attribute to it (Creswell, 1998; Morse, 1994).
Participants
The participants were university students enrolled in the teachers formation program of a
public university in Puerto Rico, specifically, those who majored in mathematics education at
elementary level or secondary level. Participation was on a voluntary basis. A public announcing
was done to the university community, requesting the participation of volunteers; eight people
applied, all of them females. The candidates assisted to an orientation, in which the nature of
their participation was discussed, as well as the observance, by the investigators, of the
university rules concerning the protection of human rights when students participate in an
investigation.
Data collection
The procedures used to collect the information were descriptive and qualitative, designed
to describe a wide range of internal and external activities. The techniques used were: long
interview, thinking-out-loud problem solving session, and a retrospective interview right after
the problem solving session. These techniques let the participants reflect about the theme, which
will contribute to their formation as teachers.
The long interview allowed the access to the meanings that participants had about

112
problem solving and it was possible to describe the beliefs that they have about this process.
After that, the participants solved four non typical mathematical problems. The participants were
instructed to solve the problems thinking out loud. This technique allowed the access to the
cognitive and metacognitive processes the participants went through. Immediately after each
problem solving session, a retrospective interview was made in which the participant had the
opportunity to reflect on their execution in the problem solving process. Thus, not only was
explored what was happening in the participant's minds, but also the problem solving execution
and the participant's thoughts about it. In this way, there were three sources of information that
allowed the triangulation of the data, allowing the researches to conclude which kind of
representation, strategy, metacognitive process of control and beliefs the participants have about
solving non typical mathematical problems.
Problems
The problems used have the characteristic that they are sufficiently challenging to
produce metacognitive behavior, but at the same time could be solved by the students with the
mathematical knowledge they had from their mathematical classes (Goos and Galgraith, 1996).
In addition, the problems can be represented in different ways and solved using diverse
strategies. The problems that were used for this analysis were problem 2 and 4.
PROBLEM 2
A square and a rectangle have the same area. The square diagonal has a longitude of inches.
If the width of the rectangle is 4 inches, what is the length of the rectangle?
PROBLEM 4
A candy sale is organized with the purpose of raising funds for the Children Cancer Association.
Olga, who is engaged in the cause, wants to sell 27 chocolate bags. There are two different kinds
of chocolate: with almond and with strawberry. Each bag of chocolate with almond has 8 bars
and each bag of chocolate with strawberry has 9 bars. If Olga has a total of 232 chocolate bars,
how many bags of each kind of chocolate does Olga have?
Analysis
The analysis of the collected data was enriched by the comprehension reached by the
investigators, and complemented by the review of literature and their experience as professors
and investigators. All the interviews were made and transcribed by the investigators. The fidelity
of each transcription was corroborated with the audio records and simultaneous reading of what
was transcribed.
Particularly, the analysis that the investigators did of the long interviews let them find an
answer to the students’ beliefs question. The retrospective interview analysis let them determine
which self-regulation processes were used when the participants solved each problem and helped
them complement and contrast all the information obtained in the long interviews and problem
solving sessions.
On the other hand, the analysis of the problem solving session provides plenty
information about how the students construct the representation of each problem, the strategies
that they use and the self-regulation strategies they showed in the solution process. The
investigators used the audio record of each problem solving session, all the computations the
participants made during the process and the transcriptions of their observations as data to
analyze.
RESULTS AND DISCUSSION
The results obtained are organized following the research questions. The extensive
interviews were used to establish the students’ beliefs about mathematical problem solving, and

113
the analysis of the solutions of problems 2 and 4 to establish the representations, the strategies
and the processes of self-regulation used by the participants. The retrospective interviews were
used to triangulate the information.
Beliefs
In general terms, the participants consider themselves very good mathematical problem
solvers. This indicates great self-confidence although in many occasions they expressed having
difficulties solving the problems. They attribute their good disposition mainly to affective
reasons. The majority declared that problem solving represents a challenge and it motivates
them; other reasons for their good disposition were that they were interested in mathematics or
that they liked to solve problems.
The participants characterized a mathematical problem as an uncertain situation because
they do not know the subject matter or the method needed to solve it. A problem requires deeper
analysis; several pieces of knowledge take part and must be used simultaneously in the process
of solving a problem. In contrast, they already know what to do when it comes to solving an
exercise, since exercises are solved with some well-known algorithm.
Just as in the investigation of Chapman (2005), the majority point out that the steps used
to solve a problem are: reading the statement, identifying the given data, determining what is
asked and solving it. The participants assigned greater importance to the understanding of the
problem and less to the process of solving it. Still more, some participants assigned great
importance to the reading of the problem, because they think that the way to solve the problem is
ciphered in the statement.
The students declared preferring the arithmetical and algebraic strategies over the
graphical strategies. All the participants indicated that they verified the problem to know if the
answer were correct, however this was not observed when they solved the proposed problems.
Some indicated that, when they are solving a problem and do not know how to follow,
they returned to read and “reread” the statement. Others indicated they abandoned it temporarily
and returned to it later on. One student indicated that she reviewed the notes to see if she had
solved a similar problem previously, another one mentioned that she tried to get help. Half of the
students indicated that “they analyzed” the problem when they did not know how to follow
ahead. For them, a problem is difficult when: they cannot solve it in the first attempt, or they do
not understand it when they read it the first time. It is also difficult when it contains too much
information or when different operations are needed in the solution process.
When asking how they considered that problem solving should be taught at school, the
majority indicated that it must be integrated more frequently in class. They also point out that
problem solving should not be a separate or isolated topic; instead it must be related to daily life
events. Also they indicated that the teachers must give the students more problems to solve, that
is, more frequent practice, thing that many of them did not get when they were students. This
aspect agrees with the obtained by Grouws and Good (2002), which found that problem solving
was not very frequent subject in the classes of mathematics of the observed teachers.
One of the participants narrated a negative experience that she had had with a university
professor when she was solving a problem that involved roots. Nowadays it is very difficult for
her to solve problems that includes these, even more, she feels uneasy whenever she faces a
problem with that include roots. In fact, she could not solve problem 2, which included a squared
root.
Representations
Immediately after reading problem number 2, the students used graphs to represent it.

114
They drew a square, a rectangle and some marked the measurement of the diagonal of the square.
In addition, three students represented in iconic form the condition that the area of the square
was equal to the area of the rectangle. Nevertheless, only two could establish the connection
between the graphical representation and the algebraic one, which limited the use of this one.
This last point converges with the indicated by Gagatsis, Elia, and Kyriakides, 2003; as cited by
Mousoulides and Gagatsis (2004), in the sense that the pre-service teacher that participated in
their study could not make the connection between the graphical and the algebraic
representation.
Several difficulties in the representation of the problem appeared. These can be classified
as difficulties that came from the invention of conditions from the data, the omission of data that
the problem provided, and others whose origin is mathematical. First of all, three students used
as the diagonal of the rectangle. Second, a student did not take into account the piece of
information regarding the fact that the two figures had the same area; this prevented her from
solving the problem. Finally, the main mathematical error in the representation was that the
students thought that the measurement of the diagonal of the unit square is one. This made them
construct a square whose sides measured the same as the diagonal.
The representation of problem four occurred in different form. Half of the students
underscored the relevant information in the statement of the problem; the other half rewrote it in
the worksheet. The majority used numerical representations, only 2 made algebraic
representations. After reading the problem, the students began to conduct arithmetical operations
with the numbers that the statement provided, which indicates that there is no understanding of
the problem immediately after the reading.
The representation used is privileged by the situation of the problem. In problem number
two the graph prevailed, whereas in problem number four the numerical one, nevertheless, the
participants could not connect this first representation to the mathematical content that allowed
them to solve the problem. Once the initial representation was constructed, the participants did
not change it, which shows little flexibility to change plans, although in several occasions the
students returned to reread the statement.
Strategies
To solve the second problem, three participants tried to construct the figures from the
diagonal. None were successful because of the errors they had committed in the initial
representation of the problem. Two participants used the Pythagorean Theorem, one succeeded
and the other not, since she used it in a rectangle, in which she had labeled as the measure
of the diagonal. Two students did not make any attempt solve the problem, nevertheless, one
guessed an answer. The other student indicated the process that she would have used to solve the
problem; nevertheless, she indicated that she did not know how to calculate the side of the
square.
With respect to the fourth problem, seven of the participants used the strategy of trial and
error. From these, three arrived at the correct answer. They used “educated estimates”, that is,
they tried some set of possible values and adjusted them according to the results they were
obtaining, eventually arriving to the correct answer.
Five participants performed calculations with the numbers given in the problem. They
carried out operations such as 232 ÷ 2 = 116; 116 ÷ 8 = 14.5; 116 ÷ 9= 12.88 or 27 ÷ 2=13.5,
with the hope that the results would fit some of the given information. This agrees with the
findings of Kieran, 1992; and Linchevski and Herscovics, 1996; as cited by Van Dooren,
Verschaffel, and Onghena (2003), whom indicated that the students preferred to perform
115
arithmetical operations with the known numbers, the meaning of such operations remaining
invariably connected to what the students perceive to be the context of the original problem.
One participant used the specific strategy of system of equations; in particular, she set up
a system of 2 linear equations with 2 variables, obtaining the correct answer. This participant
was majoring in secondary education in mathematics. This finding agrees with the obtained one
Van Dooren, et al. (2003), whom indicated that secondary school pre-service teachers prefer to
use of algebra.
Several investigators have documented the difficulties that the students have when they
face algebra for the first time and, in specific, the solution of algebraic problems (Filloy &
Sutherland, 1996; Herscovics & Linchevski, 1994; Kieran, 1992; Lee & Wheeler, 1989; Sfard &
Linchevski, 1994; as cited by Van Dooren, et. al, 2003). Some educators and investigators have
suggested that a way to solve these difficulties is "to algebrazied" the elementary mathematics
curriculum (Ainley, 1999; Davis, 1985; Discussion Document for the Twelfth ICMI Study,
2000; Kaput, 1995; Swafford & Langrall, 2000; Vergnaud, 1988; as cited by the authors). They
argued that, early in the school mathematics education, the arithmetical activities can and must
gradually be attended with an algebraic meaning with the purpose of emphasizing the inherent
algebraic characteristics. Incorporating this recommendation to the curriculum of pre-service
elementary teachers will possibly help them extend their repertory of strategies to solve problems
successfully.
Self-regulation
During the problem solving sessions the participants showed little metacognitive activity,
for example, they did not express their familiarity with the problem, nor stated its level of
difficulty. In the case of the geometry problem, a student declared with disappointment “this is a
geometry problem” when reading the problem. In the retrospective interview it was possible to
verify that she had trouble with the subject. Another student stated that everything related to
radicals caused her anxiety due to an unpleasant experience she had when she was learning them.
The other students did not indicate their confidence (or lack of it) of solving the problem.
In few occasions the participants showed evaluation strategies of the representation or the
strategy they were using. For example, very few participants changed the initial representation
and even fewer changed the strategy that they selected initially even if it did not produce the
awaited results. The evaluation of the progress towards the solution occurred more in problem
four than in problem two, since the students used the values of the statement of the problem or
some that they considered important to the solution. Another element that was used to verify the
progress towards the solution of the problem was the appearance of decimal numbers in the
results of the operations. In this aspect, it is important to stress that although the initial performed
operations did not have sense, these helped them start making use of the trial and error strategy
with initial values closer to the actual solution.
Having taken a geometry course did not help them solve problem number two. Similarly,
the familiarity with problem number four did not help them arrive to the correct answer. This is
explained since it is difficult for students to apply their mathematical knowledge to a novel
situation (Hernandez Rodriguez, 2002; Santos Trigo, 1995; Schoenfeld, 1985; Selden, Selden,
Mason, 1994).
It was observed that some of the participants who had been more time working in the
problem tend to ignore the initial conditions. This can be explained in two ways: they are
approaching the time limit that they are supposed to expend on it or for them it is indispensable
to give an answer (Schoenfeld, 1989).

116
 CONCLUSIONS
1. The participants characterized the mathematical problem as an uncertain situation
because they do not know what it is about or the method in question to solve it. A
problem requires deeper analysis; several pieces of knowledge take part and must be
used simultaneously in the process of solving a problem. In contrast, they already know
what to do when it comes to solving an exercise, since exercises are solved with some
well-known algorithm.
2. The participants declared to prefer the arithmetical and algebraic strategies over the
graphical strategies.
3. All the participants indicated that they verified the problem to know if the answer was
correct, although this was not observed in the problem solving sessions.
4. Most of the participants indicated that problem solving must be integrated more
frequently in mathematics classes. Problems solving should not be studied as a separate
or isolated topic, and it must be related to daily life situations.
5. Participants assigned great importance to the reading of the problem because they think
that the way to solve it is ciphered in the statement.
6. Participants used a graphical representation for problem number two and a numeric
representation for problem number four.
7. There were several difficulties in the representation of the problems. These can be
classified as difficulties that came from the invention of conditions from the data, the
omission of data that the problem provided, and others whose origin is mathematical.
8. With regard to the geometry problem, three participants tried to construct the figures
from the diagonal. This strategy was not successful because of the errors they committed
in the initial representation of the problem.
9. Two participants used the Pythagorean Theorem to solve problem number two. One was
successful and the other not, since she used it in the rectangle and she had labeled its
diagonal as .
10. With regard to problem number four, seven of the participants used the strategy of trial
and error. From these, 3 arrived at the correct answer. The method used was “educated
rough estimates”, that is, they tried some set of possible values and adjusted them
according to the results they were obtaining, eventually arriving to the correct answer.
11. The participants showed little metacognitive activity. They did not express their
familiarity with the problem, nor the level of difficulty of the same. They did not
indicate their confidence to solve the problem.
12. In few occasions, the participants showed evaluation strategies.
13. It is difficult for the participants to apply the mathematical knowledge to solve
unfamiliar situations.
14. Participants possessed declarative knowledge about problem solving; however, it was
difficult for them to use it to solve the problems posed.
Educational implications
Pre-service mathematics teachers must be exposed frequently to problem solving in their
mathematics classes, so that they develop the necessary skills to solve and teach appropriately to
their students. The use of diverse representations should be stimulated to fortify the connection
between them, so that they can use the one that is suitable at the appropriate time.
Finally, it is essential to foment the use of algebraic strategies in the pre-service

117
elementary mathematics teacher, so that the arithmetical activities can be attended with an
algebraic meaning.
REFERENCES
American Association for the Advancement of Science. (1993). Benchmarks for Science
Literacy. NewYork, NY: Oxford University Press.
Ainley, J. (1999). Doing algebra-type stuff: emergent algebra in the primary school.
Proceedings of the 23rd annual conference of the International Group for the
Psychology of Mathematics Education (Vol. 2, pp. 9–16). Haifa, Israel.
Ball, D. L. (1990). The mathematical understandings that prospective teachers bring to teacher
education. The Elementary School Journal, 90, 449 – 467.
Bjuland, R. (2004). Student teachers’ reflections on their learning process through collaborative
problem solving in geometry. Educational Studies in Mathematics, 55, 199- 225.
Cadenas, R. (2007). Carencias, dificultades y errores en los conocimientos matemáticos en
alumnos del primer semestre de la escuela de educación de la Universidad de los Andes.
ORBIS, Revista Científica Ciencias Humanas, 2 (6), 68 – 84.
Chapman, O. (2005). Constructing pedagogical knowledge of problem solving: Pre-service
mathematics teacher. En H. L. Chick, y J. L. Vincent (Eds.), Proceedings of the 29th
conference of the international group for the psychology of mathematics education: Vol.
2 (pp.225 – 232). Melbourne: PME.
Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among five
traditions. Thousand Oaks, CA: Sage Publications.
Davis, R. (1985). ICME-5 Report: Algebraic thinking in the early grades. Journal of
Mathematical Behavior, 4, 195–208.
De Corte, E., Greer, B., & Verschaffel, L. (1996). Mathematics teaching and learning. In D. C.
Berliner, & R. C. Calfee (Eds.), Handbook of educational psychology (pp. 491-549).
New York: Macmillan.
Departamento de Educación de Puerto Rico. (2003). Marco curricular del programa de
matemáticas. San Juan, P. R.: Autor.
Discussion document for the twelfth ICMI study (2000). Educational Studies in Mathematics,
42, 215–224.
Flavell, J. H. (1976). Metacognitive aspects of problem solving. In L. Resnick (Ed.), The
nature of intelligence (pp. 231-236). Hillsdale, NJ: Erlbaum.
Filloy, E., & Sutherland, R. (1996). Designing curricula for teaching and learning algebra. In A.
J. Bishop, K. Clements, C. Keitel, J. Kilpatrick & C. Laborde (Eds.), International
handbook of mathematics education (pp. 139–160). Dordrecht: Kluwer.
Garofalo, J., & Lester, F. K. (1985). Metacognition, cognitive monitoring, and mathematical
performance. Journal for Research in Mathematics Education, 16 (3), 163-176.
Goos, M. (2006). Why teachers matter. Australian Mathematics Teacher, 62 (4), 8-13.
Greeno, J. G., Collins, A. M.,, & Resnick, L. B. (1996). Cognition and learning. In D. C.
Berliner, & R. C. Calfee (Eds.), Handbook of educational psychology (pp. 15-46). New
York: Macmillan Library Reference.
Grows, D., & Good, T. L. (2002). Issues in problem-solving instruction. In D. L. Chambers
(Ed.), Putting research into practice in the elementary grades: Readings from journals
of the National Council of Teachers of Mathematics (pp. 60-62). Reston, VA: NCTM.
Hembree, R. (1992). Experiments and relationals studies in problem solving: A meta-analysis.
Journal for Research in Mathematics Education, 23 (3), 242-273.

118
Hernández Rodríguez, O. (2002). Procesos cognoscitivos y metacognoscitivos en estudiantes
universitarios puertorriqueños en la solución de problema matemáticos no típicos.
Disertación doctoral.
Herscovics, N., & Linchevski, L. (1994). The cognitive gap between arithmetic and algebra.
Educational Studies in Mathematics, 27, 59–78.
Janvier, C., Girandon, C., & Morand, J. C. (1993). Mathematical symbols and representation.
In P. S. Wilson (Ed.), Research ideas for the classroom. New York: Macmillan
Publishing Company.
Kaput, J. (1995). Transforming algebra from an engine of inequity to an engine of mathematical
power by “algebrafying” t he K-12 curriculum. Paper presented at the 1995 NCTM
meeting.
Kieran, C. (1992). The learning and teaching of school algebra. In D. A. Grouws (Ed.),
Handbook of research on learning and teaching mathematics (pp. 390–419). New
York:Macmillan.
Lampert, M. (1990). When the problem is not the question and the solution is not the answer:
Mathematical knowing and teaching. American Educational Research Journal, 25 (1),
29-63.
Lawson, M. J. (1990). The case for instruction in the use of general problem solving strategies:
A comment on Owen and Sweller. Journal for Research in Mathematics Education, 21
(5), 403 -410.
Lee, L., & Wheeler, D. (1989). The arithmetic connection. Educational Studies in Mathematics,
20, 41–54.
Leikin, R. (2003). Problem-solving preferences of mathematics teachers: Focusing on
symmetry. Journal of Mathematics Teacher Education, 6, 297–329.
Leonard, J., & Joergensen, P. (2002). Empowering all elementary pre-service teachers to teach
children mathematics. (ERIC Document Reproduction Service No. ED469957).
Lester, F. K. Jr. (1994). Musings about mathematical problem-solving research: 1970-1994.
Journal for Research in Mathematics Education, 25 (6), 660-675.
Liljedahl, P., Rolka, K., and Rösken, B. (2007). Affecting affect: The reeducation of pre-
service tearchers’ beliefs about mathematics and mathematics learning and teaching. In
G. W. Martin, M. E. Strutchens, and P. C. Elliott (Eds.), The learning of mathematics,
(pp. 319-330). Reston, VA: NCTM.
Linchevski, L., & Herscovics, N. (1996). Crossing the cognitive gap between arithmetic and
algebra: operating on the unknown in the context of equations. Educational Studiesin
Mathematics, 30, 38–65.
Ma, L. (1999). Knowing and teaching mathematics: Teahers’ understanding of fundamental
mathematics in China and the United Stattes. Mahwah, New Jersey: Lawrence Erlbaum
Associates, Publishers.
Maqsud, M. (1997). Effects of metacognitive skills and nonverbal ability on academic
achievement of high school pupils. Educational Psychology, 17 (4), 387-398.
Mewborn, D. S., & Cross, D. I. (2007). Mathematics teachers’ beliefs about mathematics and
links to students’ learning. In W. G. Martin, M. E. Strutchens, & P. C. Elliot (Eds.), The
learning of mathematics (pp. 259-269). Reston, VA: NCTM.
Morse, J. M. (1994). "Emerging from the data": The cognitive processes of analysis in
qualitative inquiry. In J. M. Morse (Ed.), Critical issues in qualitative research methods
(pp. 23- 43). Thousand Oaks, CA: Sage.

119
Mousoulides, N., y Gagatsis, A. (2004). Algebraic and geometric approach in function
problem solving. En P. Gates (Ed.), Proceedings of the 28th conference of the
international group for the psychology of mathematics education: Vol. 3 (pp. 385 – 392).
Melbourne: PME.
Moustakas, C. E. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage.
National Council of Teachers of Mathematics (2000). Principles and standards for school
mathematics. Reston, VA: Author.
Noddings, N. (1990). Constructivism in mathematics educations. In R. B. Davis, C. A. Maher,
& N. Noddings (Eds.), Constructivists views of the teaching and learning of
mathematics: Journal for Research in Mathematics Education, Monograph Number 9
(pp. 65-78). Reston, VA: National Council of Teachers of Mathematics.
Parra, B. M. (1991). La resolución de problemas en la construcción de esquemas de
razonamiento. Educación Matemática, 3 (1), 58-61.
Polya, G. (1945). How to solve it. Princeton, NJ: Princeton University Press.
Presmeg, N., y Nenduradu, R. (2005). An investigation of a pre-service teachers’s use of
representations in solving algebraic problems involving exponential relationships. En H.
L. Chick, y J. L. Vincent (Eds.), Proceedings of the 29th conference of the international
group for the psychology of mathematics education: Vol. 4 (pp.105 – 112). Melbourne:
PME.
Santos Trigo, M. L. (1995). ¿Qué significa el aprender matemáticas? Una experiencia con
estudiantes de cálculo. Educación Matemática, 7 (1), 46-61.
Schoenfeld, A. H. (1985). Mathematical problem solving. New York: Academic Press.
Schoenfeld, A. H. (1987). What’s all the fuss about metacognition? In A. H. Schoenfeld (Ed.),
Cognitive science and mathematics education (pp. 189-215). New Jersey, Erlbaum.
Schoenfeld, A. H. (1989). Explorations of students’ mathematical belief and behavior. Journal
for Research in Mathematics Education, 20 (4), 338-355.
Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition
and sense making in mathematics. In D. A. Grouws (Ed.), Handbook of research on
mathematics teaching and learning (pp.334-370). NY: Macmillan.
Schoenfeld, A. H., & Herrmann, D. J. (1982). Problem perception and knowledge structure in
expert and novice mathematical problem solvers. Journal of Experimental Psychology:
Learning, Memory and Cognition, 8, 484-494.
Schraw, G., & Graham, T. (1997). Helping gifted students develop metacognitive awareness.
Rooper Review, 20(1), 4-8.
Selden, J., Selden, A., & Mason, A. (1994). Even good calculus students can’t solve nonrutine
problems. In J. J. Kaput, & E. Dubinsky (Eds.), Research issues in undergraduate
mathematics learning (pp 19-26). Washington: Mathematics Association of America.
Schraw, G., & Graham. T. (1997). Helping gifted students develop metacognitive awareness.
Roeper Review, 20 (1), 4-8.
Sfard, A., & Linchevski, L. (1994). The gains and pitfalls of reification: the case of algebra.
Educational Studies in Mathematics, 26, 191–228.
Silver, E. A., & Marshall, S. P. (1990). Mathematical and scientific problem solving: Findings,
issues, and instructional implications. In B. F. Jones, & L. Idol (Eds.), Dimensions of
thinking and cognitive instruction (pp. 265-290). Hillsdale, NJ: Lawrence Earlbaum
Associates.
Skott, J. (2001). The emerging practices of novice teachers: The role of his school mathematics

120
 images. Journal of Mathematics Teacher Education, 4(1), 3-28.
Sternberg, R. J. (1998). Metacognition, abilities and developing expertise: What makes an
expert student? Instructional Science, 26, 127-140.
Swafford, J. O., & Langrall, C. W. (2000). Grade 6 students’ preinstructional use of equations to
describe and represent problem situations. Journal for Research in Mathematics
Education, 31, 89–112.
Swanson, H. L. (1990). Influence of metacognitive knowledge and aptitude on problem
solving. Journal of Educational Psychology, 82 (2), 306-314.
Swanson, H. L. (1992). The relationship between metacognition and problem solving in gifted
children. Roeper Review, 15 (1), 43-49.
Van Dooren, W., Verschaffel, L., & Onghena, P. (2003). Pre-service teachers' preferred
strategies for solving Arithmetic and Algebra word problems. Journal of Mathematics
Teachers Education, 6 (1), 27 - 52.
Vergnaud, G. (1988). Long terme et court terme dans l’apprentissage de l’algèbre. In C.
Laborde (Ed.), Actes du premier colloque franco-allemand de didactique des
mathématiques et de l’informatique (pp. 189–199). Paris: La Pense Sauvage.
Von Glasersfeld, E. (1990). An exposition of constructivism: Why some like it radical? In R. B.
Davis, C. A. Maher, & N. Noddings (Eds.), Constructivists views of the teaching and
learning of mathematics: Journal for Research in Mathematics Education, Monograph
Number 9 (pp. 65-78). Reston, VA: National Council of Teachers in Mathematics.

121
 Teachers’ beliefs about mathematical problem solving,
their problem solving competence and the impact on instruction:
A case study of three Cypriot primary teachers

Constantinos Xenofontos and Paul Andrews

University of Cambridge

Abstract
In this paper we report a case study of three Cypriot primary teachers, with respect to their
mathematical problem solving beliefs and competence, and the impact of these on instruction.
Semi-structured interviews were carried out with the teachers. Each of them was invited to
solve a purely mathematical non-routine problem and explain simultaneously the solution
process. The teachers prepared a lesson based on the problem and taught it to their
classrooms. Our findings suggest that teachers’ mathematical problem solving beliefs,
competence and instructional practices are in a complicated relation that cannot be explained
in terms of cause-and-effect.

Introduction
Mathematical problems and problem solving
In starting this literature review it is important, particularly as the word problem, even within
the domain of mathematics education, frequently means different things to different people
(Borasi, 1986; Blum & Niss, 1991; Nesher, Hershkovitz & Novotne, 2003; Wilson,
Fernandez & Hadaway, 1993; Goos, Galbraith & Renshaw, 2000), to consider how it is
defined. A common definition is that a mathematical problem presents an objective or goal
with no immediate or obvious solution or solution process (Blum and Niss, 1991; Schrock,
2000, Polya, 1981; Nunokawa, 2005). In summarising the work of Schrock (2000) and
Wilson et al (1993) we suggest that a mathematical problem must meet at least three criteria;
individuals must accept an engagement with the problem, they must encounter a block and
see no immediate solution process, and they must actively explore a variety of approaches to
the problem.

According to Chapman (1997) problem solving means different things to different people,
having been viewed as a goal, process, basic skill, mode of inquiry, mathematical thinking
and teaching approach. However, most research in the area seems to regard problem solving
as the process of achieving a solution (Chapman, 1997; Blum & Niss, 1991; Boekaerts,
Seegers & Vermeer, 1995; Franke & Carey, 1997; Hart, 1993). Famously, Polya (1981:ix)
described it as a means of “finding a way out of difficulty, a way around an obstacle,
attaining an aim which was not immediately attainable” and it is on this conception that we
focus our work.

Various writers, including Polya (1945), have developed frameworks for analysing the
problem solving process. Polya’s model comprises the four phases of understanding the
problem, devising a plan, carrying out the plan, looking back. Other models, frequently based
on Polya’s, include Kapa’s (2001) six phase and Mason et al’s (1985) three phase. The latter
suggest that problem solving comprises entry, attack and review. However, space prevents a
lengthy discussion on the details of these models and their similarities and differences,
although it is our view that their resonance with Polya’s is close and not difficult to discern.
From the perspective of this study, we tend towards Mason et al’s (1985) model as their

122
attack phase appears not to necessitate a predetermined plan in the manner of Polya’s
devising and carrying out a plan.

In the context of Cyprus, the location of this study, much research on problem solving has
been undertaken over the last few years (Christou & Philippou, 1998; Elia & Philippou, 2004;
Gagatsis & Elia, 2004; Gagatsis & Shiakalli, 2004; Nicolaidou & Philippou, 2003). However,
the focus of every study has been on students’ understanding, beliefs, abilities and attainment.
Such work, when seen against the international trend for teacher-focused research on problem
solving to address instructional effectiveness rather than teacher competence (Chapman, 1997,
Thompson, 1985) or the impact of beliefs (Thompson, 1984) highlights well a field ripe for
development.

In this paper we report on the mathematical problem solving beliefs and competence, and the
impact of these on their instructional choices, of three Cypriot primary teachers. Beliefs, and
their impact on teachers’ instructional choices, have been the subject of extensive
investigation in mathematics education. However, despite this extensive research, ambiguity
regarding terminology has caused confusion to the area (McLeod, 1988; Pehkonen & Pietila,
2003; Törner, 2002). In this paper we draw on Raymond’s (1997: 552) definition in which
teachers’ mathematics beliefs, in relation to the nature of mathematics, its teaching and
learning, refer to their “personal judgments about mathematics formulated from experiences
in mathematics”.

Methodology
A case study investigation was undertaken in March, 2007. The participants, who were all
teaching 11 and 12 year-old children at the time, were given the pseudonyms Mrs Antigoni
(22 years of teaching experience), Ms Electra (newly appointed teacher) and Mr Orestis
(second year of teaching). Initial semi-structured interviews focused on four thematic areas
relating to how colleagues viewed themselves as teachers of mathematics, their espoused
beliefs about the nature of problem solving, their perceptions of themselves as problem
solvers and, finally, their beliefs about the management of problem solving in classrooms.
After the interviews, each teacher was invited to solve a mathematical problem and explain
simultaneously the solution process. The problem presented was the following:

“On the grid paper you have been given, each little square is equal to one square unit. How
many isosceles triangles can you make which will satisfy all of the following three criteria?
1. The area must be nine square units.
2. One of the vertices is at the given point.
3. The other two vertices are on grid points too.”

Problems of this nature, which are non-routine and purely mathematical (Blum & Niss, 1991)
are never found in the National Textbooks of Cyprus.

The final phase involved each teacher in preparing and delivering a lesson based on the same
problem. The lessons were observed with the ways in which the teachers presented the
problem, managed the classroom during problem solving and their approach to responding to
students’ questions being the primary foci.

The aim was to ‘sketch portraits’ of the three participants, regarding their beliefs on problem
solving, their problem solving competence and the impact of the former on their instructional
practice. The work of Alba Gonzalez Thompson (Thompson, 1984) was particularly

123
influential. Data analysis was framed by the three-phase problem solving model of Mason et
al. (1985). The framework used for analysing classroom observation data (teachers’
instructional practices during the teaching of the problem solving activity) was chosen a
posteriori since we did not want to apply predetermined categories. For this purpose, six out
of the ten mathematical didactics (table below), proposed by Andrews (2007), were used. The
mathematical didactics are teaching strategies used by teachers in order to facilitate students’
understanding.

Table 1: Mathematical Didactics

Sketching the three portraits

A. Mrs Antigoni
Beliefs about teaching/learning mathematics and about herself as a mathematics teacher
For Mrs Antigoni, mathematics is the most important lesson in the primary curriculum. A
‘good’ mathematics teacher should give students many examples so that “they can adopt
some procedures easily in order to solve similar problems”. Also, he/she should accept
critique, either good or bad, in order to improve. She believes that all of her mathematics
teaching aspects require improvement. However, she does not make any allusions to how she
could improve her mathematics teaching. She only refers to external factors that could change
and therefore improve the conditions under which she teaches (i.e wishes that children would
concentrate more during the lessons and that, pupils would have greater self-confidence). She
claims that the quality of her teaching would be improved if the number of pupils per
classroom decreases.

Beliefs about the nature of problem solving

A mathematical problem is “a procedure, which requires you to discover which information
is given, to rank the given points, to find what the problem asks you to do and then solve it”.
The given information should be clear and accurate so that all the children understand what
has to be done. Problem solving is the achievement of a goal, either set by the problem solver
or by others. Teaching via problem solving can be feasible only if children know the
mathematical procedures. Otherwise, if students do not know how to solve procedural
exercises, they will be very disappointed and they will not have the motivation to try and

124
solve more difficult problems. Both finding a problem’s solution and the journey towards the
solution are important. The correct answer matters a lot since “in the future, students will be
asked to take mathematics exams, for which they have to achieve a high mark. If the steps
they follow are correct but still the answer is wrong, then they will not attain a good grade”.

Beliefs about her competence as a problem solver

Mrs Antigoni views herself as an average problem solver and is bored of problems which
require a more advanced level of thinking because she has been tackling simple problems
with simple solutions throughout the years of her teaching experience. Sometimes she views
problems in a superficial way and often has to revisit a problem many times in order to
understand what is required for its solution. Practice is for her the only way of improving her
problem solving abilities. Difficult mathematical problems are a cause of anxiety to her.
When she has to solve an advanced problem she tries to relax and concentrate on the given
information. When she faces difficulties during the solution process, she either recalls
previous knowledge regarding the solution of similar problems or tries to solve the problem
algebraically. If the problem is to be taught to primary school children, she tries to simplify
the solution, because “children in primary schools do not know algebraic equations”.

The management of problem solving in her classroom

The school programme, according to her, does not allow any time for problem solving
activities in classroom. Students should not spend more than three to five minutes on a
problem. Her mathematics teaching is based on the national textbooks of the Ministry of
Education and Culture. She brings mathematical problems in her classroom she creates
herself. It is easier and more practical for her to write problems similar to those in the
national textbooks rather than to search for them on the internet or in books. Most of her
students feel nervous when they have to solve mathematical problems. Only few of them will
try to reach a solution while the rest quickly abandon every effort. When students face
difficulties during problem solving, she tries to give them some hints that will help them to
proceed. Students should be encouraged by the teachers with expressions like “well done, you
are on the right track. Now try to think what else should be done”.

Solving the problem

Mrs Antigoni’s Entry to the problem began by acknowledging the central goal of the activity
(finding triangles that satisfy all three criteria) and by introducing the formula for the area of
triangles. After finding four of the 12 triangles of the group of “base 6, height 3”, Mrs
Antigoni began examining another case instead of finding the rest of the triangles in the same
group. Soon after that, she demonstrated her boredom and was ready to abandon the problem.
She continued her effort, mostly because of my interventions. Eventually she found all 36
solutions, though with great difficulty. A wide range of emotions were observed; from her
initial enthusiasm to the desire to abandon the problem.

The lesson based on the problem

Mrs Antigoni’s lesson was a procedural explanation of the problem by the teacher. Children
were given very little opportunities to demonstrate their thinking since the teacher was
explicitly explaining every step during the teaching period. Two mathematical didactics were
observed, Activating Prior Knowledge and Explaining.

125
Activating Prior Knowledge
At some point in the middle of the lesson, she asked students to bring to mind a previous
lesson they had on symmetry. She suggested that students could use the rules of symmetry to
find the triangles.

Explaining
Generally speaking, the lesson was based on the teacher’s explanations and step-by-step
guidance. Selectively, I present the two following selected incidences which confirm and
justify my claim:
(a) When children were given the problem, the teacher told them to start by using the formula
for the area.
(b) A student suggested that there could be a triangle with base 4 units and height 4.5 units.
The teacher told him that they were to work only with integers and that decimal numbers
could not be used. She did not offer any further explanations on this.

B. Ms Electra
Beliefs about teaching/learning mathematics and about herself as a mathematics teacher
Mathematics is for Ms Electra a very important lesson, since it is in the primary and
secondary curricula of every educational system around the world. Its importance can be seen
in daily life due to its many applications, like money, measurements and so on. She feels able
to teach mathematics only if she is well prepared. She claims that she does not actually know
mathematics, since she lacks advanced knowledge in this field. A ‘bad’ mathematics teacher
is one who teaches procedurally and does not allow children to build their own conceptual
understanding. A ‘good’ mathematics teacher approaches mathematics teaching in an
interdisciplinary way, and tries to facilitate students’ learning and to make children appreciate
mathematics. She wishes she would feel more secure in teaching lessons based on non-
routine problems.

Beliefs about the nature of problem solving

A mathematical problem is like every other problem; it is a situation that requires a solution.
This solution can either be approached through a simple procedure or certain strategies have
to be employed. “There is a wide range of mathematical problems; at one end of the scale,
there are very simple ones and at the other end, there are some which are much more
complicated”. A mathematical problem has a verbal and a numerical part. The ways the given
information and the question of the problem are expressed comprise its verbal part. The
numerical part is included in the verbal, in the form of numbers and symbols. Problem
solving is the procedure during which the problem solver understands the problem, separates
the given information from what he/she is asked to achieve and uses all the elements of the
problem in order to reach a solution. Teaching via problem solving is a method of direct
approach of a mathematical concept and is a way of connecting a mathematical concept with
daily life situations. Teaching via problem solving can be used as a way for introducing new
concepts.The journey towards the solution of a problem is much more important than the
solution itself. The solver has to find her/his own way which will eventually lead her/him to
the solution. Only then has the problem solver really understood the problem and its solution
and can explain her/his journey to others.

Beliefs about her competence as a problem solver

Ms Electra labels herself as an insecure problem solver. As she claims, “solving problems is
not something I enjoy doing and furthermore I do not have the skills for it”. If she does not
have to solve a difficult mathematical problem then she is not interested in solving it. She

126
would only try to solve a difficult problem if she had to present it to students and that would
cause her anxiety. Mathematical problems in particular make her nervous, but “if I really had
to solve it, I would try and use every possible means, because I am a really stubborn and
proud person”. When she faces difficulties during the process of solving a problem, she
either asks for advice from someone more experienced or she temporarily abandons the
problem until she relaxes and then revisits it and keeps trying until she reaches a solution.

The management of problem solving in her the classroom

Time parameters set by the school programme do not allow much time to spend on problem
solving activities, states Ms Electra. Most of the problem solving activities she uses concern
the introduction of new mathematical concepts. She does not spend time on problems if they
are not related to a particular mathematical concept she wants to teach. The time a student
should spend on a problem depends on her/his abilities and on the degree of difficulty of the
problem. The majority of the students in her classroom do not like solving problems due to
the insecurity children feel. Her class do not respond with enthusiasm when she presents
them with mathematical problems. When her students face difficulties during the process of
solving a mathematical problem, she helps each student independently. She asks her students
questions that will help them to think rationally. She does not directly explain to children
what to do; she prefers that they discover the way by themselves. Suggesting her students
work in pairs is another method she uses. Students can learn a lot through discussion and
exchange of ideas while working with mathematical problems.

Solving the problem

Ms Electra seemed a little nervous when she was told she would be given a mathematical
problem. Nevertheless, during the process of solving the problem, she managed to control her
initial anxiety and successfully navigated through the problem’s difficulties. As soon as she
realised that the initial way she was following was not appropriate, because of the three
criteria, she revised her thinking and applied a new way. After discovering a pattern, she
easily generalised her ideas and found all 36 solutions.

The lesson based on the problem

Ms Electra’s lesson was mostly based on students’ sharing and explaining their ideas to their
classmates. Also, Ms Electra was particularly interested in offering individual and group
feedback and assessing her students’ understanding. This was achieved either through a
group discussion, where the teacher was managing rather than explaining, or through
questioning, where the teacher used to challenge her students with questions. Analytically,
selected incidents confirming the mathematical didactics employed by Ms Electra are
presented below.

Sharing
(a) A student argued that the height that begins from the vertex between the two equal sides
of an isosceles triangle divides the triangle into 2 equal parts. The teacher asked the student to
demonstrate his ideas on the board and explain it to his classmates.
(b) At a point where most of the students had found all of the triangles with base 2 and height
9 and base 6 and height 3, the teacher asked if there was anyone who would like to explain
analytically her or his way of thinking to the others. One student went to the board (where a
grid paper was presented though an overhead projector) and explained to her classmates how
she found all the triangles with base 6 and height 3. In the meantime, the other students asked
that student questions on some of her steps.

127
Questioning
The teacher asked the students to number the characteristics of isosceles triangles.
Student: The sum of their angles is 180˚.
Ms Electra: Is this a characteristic that only isosceles triangles have?
Student: No, it applies to all triangles.
Ms Electra: How important do you think this information is for our problem?
Student: We don’t need it.

Assessing and coaching

When students were working either individually or in groups, Ms Electra went round the
class and asked them to explain what they were doing. Her help and feedback was very
encouraging. When the students faced difficulties, she challenged them with hints that would
lead them to discover their own way.

C. Mr Orestis
Beliefs about teaching/learning mathematics and about himself as a mathematics teacher
Mr Orestis views mathematics as a way of thinking. Mathematics learning and the acquisition
of particular techniques/information are for him totally different. The learning of mathematics
can help students in their daily lives. More importantly, though, mathematics helps the
learner in developing a more organised way of thinking. He feels confident in teaching
primary mathematics but he would like to improve his mathematics teaching abilities as
regards the introduction of new concepts. He wishes he knew better ways of developing the
conceptual understanding of his students, because “a good mathematics teacher is one who
helps students in building a conceptual understanding of mathematics rather than finding
answers mechanistically”.

Beliefs about the nature of problem solving

A mathematical problem is a situation which leads the solver “on a mental adventure”. The
solver has to use the given information methodically in order to arrive at a certain solution.
There must not be a standard way to the solution and problem should be clearly posed so that
it does not lead the solver to misunderstandings. Problem solving is the successful
manipulation of the given information in order to find a solution. Teaching via problem
solving is not feasible for all mathematical topics, despite the fact that it should be at the
centre of teaching and learning mathematics. The journey towards the solution of a problem
is much more important than its solution. It does not actually matter if the solver did not
reach a correct answer because of some procedural mistakes when using the numbers and so
on. “The emphasis is on the mental processes required in order to arrive at the solution”.

Beliefs about his competence as a problem solver

Mr Orestis feels confident at solving mathematical problems, and particularly those which
refer to primary mathematics. Experience is for him a very important factor for someone to
improve her/his problem solving abilities. Difficult mathematical problems cause him some
anxiety, but if he had to solve a difficult problem he would overcome his anxiety and would
do his best to solve it. When he faces difficulties during the solving of a problem, he revisits
the problem and the given information and tries to connect them in a way that will lead him
to the solution.

The management of problem solving in his classroom

The school teaching program, as he claims, does not offer many opportunities for problem
solving activities. The time that should be spent on a problem depends on two factors, the

128
degree of the problem’s difficulty and the students’ abilities. A teacher should be flexible as
regards time because he has to avoid creating the impression that “the fastest problem solver
wins”. A whole teaching period (40 minutes) should not be spent on a problem, unless there
is one and only objective for that lesson. “A problem that requires an entire teaching period
in order to be solved must be very demanding and students will have to use advanced mental
processes”. His students do not like mathematical problems and they prefer simple
procedural exercises with only one particular answer and one particular way of arriving at it.
When his students face difficulties during the process of solving a problem, he gives some
hints so that the student can find their own way towards the solution. If he observes that some
difficulties are common among the students, then he gives them group feedback.

Solving the problem

Mr Orestis was very calm when he was given the problem. Nonetheless, during solving the
problem, he was methodical from the beginning. He identified all the pairs of numbers which
could be the dimensions of triangles satisfying the three criteria. After drawing all the 12
triangles of the group “base 2, height 6”, he generalised his idea and claimed that there were
36 triangles for the problem.

The lesson based on the problem

During the lesson based on the problem, Mr Orestis offered hints to his students on how they
could proceed with the solution. However, a big part of his lesson, especially at the end of it,
was mostly based on the teacher’s demonstration of the solution. Selected incidents that
justify the use of particular mathematical didactics are presented below.

Coaching
(a) The classroom was discussing the criterion of the problem which stated that all vertices
should be on grid points. In Greek, there is no single word for “grid point”, so the Greek
translation of the problem referred to this as the “points where the horizontal and vertical
lines of the paper intercept”. The teacher asked the students if they understood what was
meant. Some children argued that they were not sure. Then Mr Orestis drew the following
shape on the board and asked the children if a vertex could be like that.

A student explained to her classmates that this could not happen. All children agreed.
(b) Mr Orestis asked the children: “How are we going to use the first piece of information
that the problem gives us? 9 square units. Definitely, we are not going to draw triangles by
chance”.

Explaining
(a) A student asked if a triangle with base 4 and height 4.5 could be a solution. The teacher
told him that only integers should be used, without offering any further explanations.
(b) At the end of the lesson, the teacher told the students that they should look at all the cases
of triangles with certain dimensions. He chose to demonstrate the triangles with base 6 and
height 3, since he had already presented some of those cases earlier. When the bell rang, Mr
Orestis told the children that if anyone was interested in finding all the solutions, he/she
should try at home.

129
Conclusions
Many researchers have examined the relationship between beliefs and mathematics
achievement, and specifically the attainment regarding problem solving. However, as already
noted, the foci of their studies were on students and not teachers. For instance, the study of
Nicolaidou and Philippou (2003) has shown that there is a significant correlation between
Cypriot primary students’ attitudes towards mathematics, self-efficacy beliefs and
performance on problem solving activities. Similarly, Mason’s (2003) study, concerning
Italian students, revealed that the strongest predictors for mathematics achievement were two
problem solving beliefs (belief regarding perceived ability to solve time-consuming problems
and belief that not all problems can be solved by applying step-by-step procedures). We
assert, however, that the connection between teachers’ beliefs about their problem solving
competence and the observed competence is complicated and cannot be presented as a linear
function.

The findings of the research reveal the complexity of the relation between teachers’ problem
solving beliefs and competence. Moreover, they show that the interrelationships of beliefs
and competence on teachers’ instructional practice are also complex with no simple cause
and effect in much the same way as Thompson (1984) found with her three teachers. In
closing we refer to Thompson’s appeal for teachers (1) to experience mathematical problem
solving from the perspective of the problem solver before they can adequately deal with its
teaching, (2) to reflect upon the thought processes that they use in solving problems to gain
insights into the nature of the activity and (3) to become acquainted with the literature on
research on problem solving and instruction in problem solving. According to Cooney (1985),
studies suggested that teachers may not possess rich enough constructs to envision anything
other than limited curricular objectives or teaching styles and hence may be handicapped in
realising a problem solving orientation. The use of a problem-solving approach demands not
only extensive preparation but also the development of ways to maintain at least a modicum
of classroom control and, perhaps most importantly, the ability to envision goals of
mathematics teaching in light of such an orientation.

References
Andrews, P. (2007) Negotiating meaning in cross-national studies of mathematics teaching:
kissing frogs to find princes. Comparative Education, 43(4), 489–509
Blum, W. & Niss, M. (1991) Applied Mathematical Problem Solving, Modelling,
Applications, and Links to Other Subjects: State, Trends and Issues in Mathematics
Instruction. Educational Studies in Mathematics, 22 (1), 37-68.
Boekaerts, M., Seegers, G, & Vermeer, H. (1995) Solving Math Problems: Where and Why
Does the Solution Process Go Astray? Educational Studies in Mathematics, 28 (3), 241-
262.
Borasi, R. (1986) On the Nature of Problems. Educational Studies in Mathematics, 17 (2),
125-141.
Christou, C. & Philippou, G. (1998) The Developmental Nature of Ability to Solve One-Step
Word Problems. Journal for Research in Mathematics Education, 29, No. (4), 436-442.
Cooney, T. J. (1985) A Beginning Teacher's View of Problem Solving. Journal for Research
in Mathematics Education, 16 (5), 324-336
Elia, I. & Philippou, G. (2004) The functions of pictures in problem solving. Proceedings of
the 28th Conference of the International Group for the Psychology of Mathematics
Education, 2, 327–334.

130
Franke M. L. & Carey, D. A. (1997) Young Children's Perceptions of Mathematics in
Problem-Solving Environments. Journal for Research in Mathematics Education, 28 (1),
8-25
Gagatsis, A. & Elia. I. (2004) The effects of different modes of representation on
mathematical problem solving. Proceedings of the 28th Conference of the International
Group for the Psychology of Mathematics Education, 2, 447–454.
Gagatsis, A. & Shiakalli, M. (2004) Ability to Translate from One Representation of the
Concept of Function to Another and Mathematical Problem Solving. Educational
Psychology, 24 (5), 645-657
Goos, M., Galbraith, P. & Rensaw, P. (2000) A Money Problem: A Source of Insight into
Problem Solving Action. International Journal for Mathematics Teaching. (April, 13th).
Electronic Journal.
Hart, L. C. (1993) Some Factors That Impede or Enhance Performance in Mathematical
Problem Solving. Journal for Research in Mathematics Education, 24, (2), 167-171
Kapa, E. (2001) A Metacognitive Support during the Process of Problem Solving in a
Computerized Environment. Educational Studies in Mathematics, 47 (3), 317-336.
Kelly, C. A. (2006) Using Manipulatives in Mathematical Problem Solving: A Performance-
Based Analysis. The Montana Mathematics Enthusiast, ISSN 1551-3440, vol. 3 (2), 184-
193
Mason, L. (2003) High School Students’ Beliefs About Maths, Mathematical Problem
Solving, and Their Achievement in Maths: A cross-sectional study. Educational
Psychology, 23 (1), 73-84
Mason, J., Burton, L., & Stacey, K. (1982) Thinking mathematically. London: Addison-
Wesley.
Nesher, P., Hershkovitz, S. & Novotne, J. (2003) Situation Model, Text Base and What Else?
Factors Affecting Problem Solving. Educational Studies in Mathematics, 52 (2), 151-176.
Nicolaidou, M. & Philippou, G. (2003) Attitudes towards mathematics, self-efficacy and
achievement in problem solving. In: M. A. Mariotti (Ed), European Research in
Mathematics Education III. Pisa: University of Pisa.
Nunokawa, K. (2005) Mathematical problem solving and learning mathematics: What we
expect students to obtain. Journal of Mathematical Behavior, 24, 325–340
Polya, G. (1945) How to Solve It: A New Aspect of Mathematical Method. London: Penguin
Books Ltd.
Polya, G. (1981) Mathematical Discovery: On Understanding, Learning and Teaching
Problem Solving. New York: Wiley.
Raymond, A. M. (1997) Inconsistency between a Beginning Elementary School Teacher's
Mathematics Beliefs and Teaching Practice. Journal for Research in Mathematics
Education, 28, (5), 550-576.
Schrock, C. (2000) Problem Solving-What Is It? Journal of School Improvement, 1 (2).
Thompson, A. G. (1984). The relationship of teachers’ conceptions of mathematics and
mathematics teaching to instructional practice. Educational Studies in Mathematics, 15,
105-127.
Thompson, A. G. (1985). Teachers’ conceptions of mathematics and the teaching of problem
solving. In E. A. Silver (Ed.), Teaching and learning mathematical problem solving:
Multiple research perspectives (pp. 281-294). Hillsdale, NJ: Lawrence Erlbaum.
Wilson, J., Fernandez, M., & Hadaway, N. (1993) Mathematical problem solving. Retrieved
from University of Georgia, Department of Mathematics Education EMAT 4600/6600
Website: http://jwilson.coe.uga.edu/emt725/PSsyn/PSsyn.html

131
 Strategies for Solving Word Problems on Speed: A Comparative

Study between Chinese and Singapore Students

Jiang Chunlian
Department of Mathematics and Statistics
Central China Normal University, Wuhan, 430079, People’s Republic of China
cljiang@mail.ccnu.edu.cn

Abstract: The study was conducted to investigate strategies that Chinese and
Singapore students used for solving word problems on speed. A test comprising 14
word problems on speed was administered to 1002 Singapore and 1070 Chinese
students from Primary 6 to Secondary 2. A two-way country×grade ANOVA revealed
that there was a significant interaction. The strategy analysis indicated that the
Chinese students performed better than the Singapore students because they used
algebraic strategies more frequently than the Singapore students. The Singapore
students performed better than the Chinese students on one problem because the
Primary-6 Singapore students successfully used model drawing and unitary methods.
The Singapore students were found to use model drawing, unitary, guess-and-check,
etc. more frequently than the Chinese students. However, the success rates of the use
of these strategies were lower than those of the algebraic strategies that were used
more frequently by the Chinese students. The study has some implications for the
teaching and learning of speed, algebra, and problem solving in schools.
Keywords: Word problems, Speed, Cross-national comparison, Problem-solving
strategies

Introduction
Problem solving has been included in Singapore syllabi since 1992 (Ministry of
Education (MOE) (Singapore), 1990a, 1990b). It is explicitly postulated that “the
primary aim of the mathematics programme is to enable pupils to develop their ability
in mathematical problem solving” (MOE (Singapore), 2001a, 2001b, p.5). About
twelve heuristics such as “act it out”, “draw a model/diagram” etc. are suggested in
mathematics syllabi (MOE (Singapore), 2001a, 2001b). However, in China, the
teaching of mathematics put more emphases on the learning of basic knowledge and
the training of basic skills, also known as “Two Basics” (Zhang, Li, & Tang, 2004).
Problem solving is taught after the teaching of basic mathematical concepts and
techniques to illustrate their applications to the real world. The new curriculum
standard (BNUP (China), 2001) has included problem solving as one of the four
aspects of mathematics teaching and learning. A comparative study between the two
countries is meaningful for getting insight into what are the advantages and
disadvantages of the different ways of teaching problem solving in schools.
Word problems on speed are selected because they are application problems of
various mathematical concepts from primary to university levels. The mathematics of
change and variation, and in particular, the study of motion, is a fundamental concept

132
that underlies elementary algebra and calculus instruction (Bowers & Nickerson,
2000). Speed has been studied as conceptual problems (Acredolo & Schmid, 1981;
Piaget, 1970; Zhou, Peverly, Boehm, & Lin, 2000; Zhou, Peverly, & Lin, 2004) with
children at the age of 5 to 12 years. However, very little work has been done with
word problems on speed as those included in textbooks. Though several studies
include rate problems as a specific model of multiplication and division (Bell,
Fischbein, & Greer, 1984; Fischbein, Deri, Nello, & Marino, 1985; Greer, 1992), the
word problems on speed included in these studies are only the simplest one among the
13 categories of motion (speed) problems Mayer (1981) identified. Mayer analyzed
word problems on speed in secondary school mathematics textbooks, but Mayer did
not investigate how students actually solve the problems and what difficulties they
may have. This study seeks in part to fill these gaps. Another reason is that word
problems on speed are both included in mathematics textbooks in China (Jiang, 1998a,
1998b; People’s Education Press (PEP), 1992, 1993a, 1993b, 1994) and in Singapore
(MOE (Singapore), 2000a, 2000b; Teh & Looi, 2002a, 2002b).
The comparative study was conducted to answer the research question “Being
taught various problem solving heuristics, do Singapore students perform better than
Chinese students in solving word problems on speed?”
A problem-solving strategy model (Table 1) including nine strategies was
developed from a review of textbooks and syllabi in the two countries and the
Concepts in Secondary Mathematics and Science (CSMS) project (Hart, 1981). Here,
strategies refer to the methods or problem solving procedures that direct the search for
a solution (Krulik & Rudnick, 1988). The problem-solving strategy model was used to
identify strategies used by the students.

Table 1
The Problem-Solving Strategy Model
No. Strategy category Definition
1 Arithmetic method It is used where the subject writes down a mathematical
statement involving one or more operations on the numbers
given in the problem (Fong & Hsui, 1999).

2 Algebraic method It is used when one or more unknowns are chosen as

variables and equation(s) is set up.

3 Model drawing method It is used when the solution is suggested by or follows a

model or a diagram (Kho, 1987).

4 Guess-and-check It involves the following two steps:

(a) Make a guess of a certain answer or the unknown in the
problem based on an estimation;
(b) Check if the constraints given in the question or implied
from some of the question statements are satisfied. If all
the constraints are satisfied, the guess is correct; the

133
 answer has been obtained or can be worked out. All
processes will end at this point. If the constraints are not
satisfied, the guess will be refined or adjusted, and another
guess will be made, then another round of guess-and-check
will begin.

5 Looking for a pattern It involves the following three steps:

(a) Several specific instances/special cases/particular
examples of a problem are explored and listed;
(b) A pattern (or a conjecture, a generalization, a hypothesis,
or common property among those special cases) is
determined by investigating the special cases explored in
Step (a); and
(c) A solution to the entire problem is found by applying the
generalized result obtained in Step (b).

6 Unitary method Unitary method involves finding the value equivalent to one
unit of a quantity from an equating statement and obtaining
the value equivalent to more units of the quantities using the
value for one unit just found (Fong, 1999; Fong & Hsui,
1999; Yuen, 1995).

7 Proportion method A proportion method is used when proportional properties

(direct and inverse proportions) are used.

8 Logical reasoning Logical reasoning strategy is used when some forms of

“if-then” reasoning are used (van De Walle, 1993).

9 No strategy It refers to the absence of a written response and where only

pieces of information taken from the question are written
down but without any continuing working (Fong & Hsui,
1999).

Method
1070 Chinese students (361 Primary 6, 354 Secondary 1, & 355 Secondary 2) and
1002 Singapore students (345 Primary 6, 315 Secondary 1, & 342 Secondary 2)
participated in the study. The Chinese sample was from Wuhan City, China. A test
was developed from an analysis of various types of word problems on speed (Jiang,
2005) and administered in intact classes. No calculators were allowed. Prior to the
test, all the students have learned and completed the topic on speed.
Both quantitative and qualitative analyses were conducted. The responses were
scored using a 0-1-2 scale. Two points were given to each correct answer or an
incorrect answer where all the necessary steps are included but with only minor
computational errors. One point was given to each answer that solved part of the

134
problem. A zero ‘0’ point was given to answers that were completely wrong and to
cases with no solution offered. The problem-solving strategy model was used to
identify students’ responses to 11 problems where workings were required.

Results
This section has three parts. The first part is about the performance comparison
results. The second part is trying to explain the performance differences from
analyzing the strategies the students used. The third part presents detailed results
about the use of strategies for solving three problems.

Performance Comparison Results

Table 2 shows the mean scores and standard deviations of the Chinese and the
Singapore students in the test. A two-way country×grade ANOVA revealed that there
was a significant interaction (F(2, 2066) = 22.88, p < .001). Post hoc pairwise
comparisons indicate that there were no statistically significant differences between
the Chinese students at any two of the three grade levels. However, there were
statistically significant differences between the Primary-6 and Secondary-1 Singapore
students (Mean difference (MD) = 2.76, p < 0.001) and between the Primary-6 and
Secondary-2 Singapore students (MD = 3.38, p < 0.001). There were no significant
differences between the Secondary-1 and Secondary-2 Singapore students (MD = 0.62,
p = 0.37). The post hoc pairwise comparisons also showed that there were statistically
significant differences between the Chinese and the Singapore students at each of the
three grade levels (p < 0.001). Therefore, the main difference existed between the
Primary-6 and secondary Singapore students.

Table 2
Means and Standard Deviations of the Chinese and Singapore Students on the Test
China Singapore
Grade Mean Difference
M SD M SD
Primary 6 21.95 5.41 19.52 4.60 2.43
Secondary 1 21.74 5.10 16.76 5.09 4.99
Secondary 2 22.20 5.05 16.13 5.72 6.07
Overall 21.96 5.19 17.49 5.36 4.47

Differences in the Use of Strategies

Comparisons in the use of strategies between different groups of students are
made in terms of Strategy Percents (SP), which is the percent of a specific group of
students using a specific category of strategy. Table 3 shows the SPs of the students
for solving 11 problems.

Table 3
Strategy Percents (SP) of the Chinese and the Singapore Students in Solving Problems in the study
Strategy category China Singapore

135
 P-6 S-1 S-2 Total P-6 S-1 S-2 Total
(n=361) (n=354) (n=355) (n=1070) (n=345) (n=315) (n=342) (n=1002)

Arithmetic method 68.24 38.50 42.51 49.86 47.01 44.36 46.23 45.91
Algebraic method 15.66 50.13 48.14 37.84 0.18 0.26 4.89 1.81
Model drawing method 0.13 0 0.05 0.06 20.26 20.20 13.85 18.05
Unitary method 0 0 0 0 12.02 9.90 10.55 10.85
Guess-and-check 2.37 0.15 0.18 0.91 13.18 8.60 5.21 9.02
Logical reasoning 2.04 0.10 0.20 0.79 0.47 0.63 0.35 0.48
Proportion method 1.44 0.46 0.18 0.70 0.53 0.38 0.19 0.36
Looking for a pattern 0.05 0 0 0.02 0.34 0.23 0.13 0.24
No strategy 10.07 10.66 8.73 9.82 6.01 15.44 18.61 13.27
Note. P-6 = Primary 6; S-1 = Secondary 1; S-2 = Secondary 2.
a
The numbers are the percentages (% is omitted) of the cases (No. of problems (11) × No. of students)
where the group of students used the specific kind of strategies.

Data in Table 3 show that:

(1) The students from both countries used the arithmetic strategies in most of the
cases.
(2) For the Chinese students, the second most frequently used strategy is algebraic
strategies. For the Singapore students, the second most frequently used strategy is
model drawing methods followed by unitary and guess-and-check methods.
(3) The use of other three kinds of strategies (i.e., logical reasoning, proportion
method, and looking for a pattern) was very rare.
(4) The Singapore students had no strategy in more of the cases than their Chinese
peers.
(5) For the samples from the same country, the trends were different. For the Chinese
sample, the secondary students used algebraic strategies more frequently than the
arithmetic strategies. Though the Secondary-2 Singapore students did use the
algebraic strategies more frequently than the Primary-6 and Secondary-1
Singapore students, they used the algebraic strategies even much less frequently
than the Primary-6 Chinese students. The secondary Singapore students still used
the arithmetic, model drawing, unitary, and guess-and-check methods as the
Primary-6 Singapore students.

Therefore, the strategy analyses generally show that (a) the Singapore students
used a greater variety of strategies for solving word problems on speed; and (b) the
Chinese students used algebraic strategies more frequently than the Singapore
students, especially the secondary Singapore students.

Strategies for Solving Three Problems

As examples, this part will show the detailed results about the use of strategies for
solving two problems (Appendix A). Problem 1 is a typical algebraic word problem
like Cows and Chickens Problem in Kaur (1998). Problem 2 describes a round trip.
Knowledge of inverse proportion could be used for solving it. Problem 3 involves

136
fractions to represent the relationships among the distances. This kind of problems
were found in a popular workbook written by Fong (1998).
For solving individual problems, we do not only use SP for comparison, but also
use Success Rate (SR). SR is the proportion of the specific group who could use the
strategy to get the correct answers to a problem. It is used to measure how
successfully the strategy is used by different group of students. If the Success Rate
Difference (SRD) between different groups of students is more than 10%, it is taken
as high. If the success rate of a strategy is lower than 30%, the strategy is taken as
inappropriate for solving a specific problem because a majority (70%) of the students
using the strategy could not reach the correct answer. Table 4 shows the SPs and SRs
in the use of strategies for solving the three problems.

Table 4
Strategy Percents (SP) and Success Rates (SR) of Chinese and Singapore Students in Solving the
Three Problems
China Singapore
Strategies P-6 S-1 S-2 Total P-6 S-1 S-2 Total
(n=361) (n=354) (n=355) (n=1070) (n=345) (n=315) (n=342) (n=1002)
Problem 1
Arithmetic SP 8.31 3.11 4.51 5.33 21.16 25.71 31.87 26.25
Method SR 6.67 0 0 3.51 1.37 0 1.83 1.14
Algebraic SP 46.54 82.20 80.85 69.72  0.32 13.16 4.59
Method SR 92.26 91.75 89.20 90.88  100 80.00 80.43
Guess-and SP 18.84 1.69 1.69 7.48 66.38 41.90 27.49 45.41
-check SR 97.06 100 100 97.50 96.51 93.18 97.87 95.82
Logical SP 12.19 0.56 1.12 4.67 3.48 5.08 3.22 3.89
Reasoning SR 75.00 100 50 74.00 83.33 31.25 54.54 53.85
Model SP 0.28   0.09 3.48 8.25 3.51 4.99
drawing SR 100.00   100.00 25.00 30.77 16.67 26.00
Looking for SP     0.29   0.10
a pattern SR     100.00   100.00
No strategy 13.85 12.43 11.83 12.71 5.21 18.73 20.76 14.77

Problem 2
Arithmetic SP 30.75 22.88 21.97 25.33 54.20 46.35 43.27 48.00
method SR 0 3.66 0 1.11 0 0.68 0 0.21
Algebraic SP 31.86 53.39 62.82 49.25  0.32 11.70 4.09
method SR 87.83 67.20 63.23 70.02  100.00 65.00 65.85
Guess-and SP 2.77   0.93 21.16 11.11 6.43 12.97
-check SR 80.00   80.00 90.41 74.29 72.73 83.08
Proportion SP 15.79 5.08 1.97 7.66 5.22 3.81 1.46 3.49
method SR 82.46 94.44 100.00 86.59 83.33 75.00 40.00 71.43
Model SP     7.83 3.81 2.05 4.59

137
 drawing
SR     40.74 41.67 14.29 36.96
method
Unitary SP     2.90 0.95 1.17 1.70
method SR     40.00 100.00 100.00 64.70
Logical SP 4.43 0.56 1.13 2.06 1.45 1.90 0.58 1.30
reasoning SR 93.75 100 100 95.45 100 50.00 100 76.92
No strategy 14.40 17.80 12.11 14.77 7.25 31.75 33.33 23.85

Problem 3
Arithmetic SP 81.44 40.96 42.54 55.14 3.77 7.30 19.30 10.18
Method SR 61.90 40.69 53.64 54.58 23.08 8.70 12.12 24.51
Algebraic SP 8.86 49.44 48.73 35.51   0.88 0.30
Method SR 34.38 49.14 54.91 50.53   66.67 66.67
Model SP 0.28  0.28 0.19 52.75 51.75 38.01 47.41
drawing SR 0.00  100.00 50.00 65.93 44.79 47.69 53.68
Unitary SP     42.61 39.05 36.55 39.42
Method SR     68.03 56.91 52.00 59.49
No strategy 9.42 9.60 8.45 9.16 0.87 1.90 5.26 2.69
Note. P-6 = Primary 6; S-1 = Secondary 1; S-2 = Secondary 2.

For Problm 1, we can have the following abservations:

(1) A higher percentage of the Singapore students used the inappropriate arithmetic
strategies (SR < 5%) than that of the Chinese students (SPD = 21%). It is
surprising to find that the higher the grade, the higher percentage of the Singapore
students used this inappropriate strategies.
(2) A majority (70%) of the Chinese students used the effective algebraic strategies
(SR > 80%). However, only about 5% of the Singapore students did so. The
Chinese students used the algebraic strategies better than their Singapore peers
(SRD > 10%).
(3) Up to 45% of the Singapore students used the guess-and-check strategies.
However, only 7% of the Chinese students did so. The students from the two
countries used the guess-and-check strategies equally successfully (SRD = 2%).
(4) About 5% of the Chinese students and 4% of the Singapore students used logical
reasoning method. The Chinese students used the logical reasoning method more
successfully than their Singapore peers (SRD = 20%).
(5) A higher percentage of the Singapore students used the inappropriate model
drawing methods (SR < 30%) than the Chinese student (SPD = 5%).
(6) One Singapore student was found to use the “looking for a pattern” method to
solve this problem.
(7) Similar percentages of the Singapore students and the Chinese students had no
strategies (SPD = 2%). However, for the students at the three grade levels, the
trends are different. For Primary-6 samples, a higher percentage of the Chinese
students had no strategies than their Singapore peers (SPD = 9%). However, for
secondary samples, higher percentages of the Singapore students had no strategies

138
 than those of their Chinese peers (SPD = 6-9%).

For Problem 2, we can have the following abservations:

(1) A higher percentage of the Singapore students used the inappropriate arithmetic
method (SR < 2%) than that of the Chinese students (SPD = 23%). Most of them
used irrelevant procedures like “(120+40)÷2 = 80”, “120+40 = 160, 160×2 = 320”,
etc.
(2) A higher percentage of the Chinese students used the appropriate algebraic
strategies (SR > 65%) than that of the Singapore students (SPD = 45%). The
students from both countries could use the algebraic strategies equally well (SRD
= 4%).
(3) A higher percentage of the Singapore students used the guess-and-check methods
than that of the Chinese students (SPD = 12%). The Primary-6 Singapore students
used this method more successfully than the secondary Singapore students.
(4) A higher percentage of the Chinese students used the proportion methods than that
of the Singapore students (SPD = 5%). The Chinese students used the proportion
methods more successfully than the Singapore students (SRD = 15%). For the
Chinese students, the higher the grade, the more successfully they used proportion
method. However, for the Singapore students, it was reversed.
(5) About 6% of the Singapore students used the appropriate model drawing (SR =
37%) and unitary methods (SR = 65%). However, no Chinese students were found
to use these two kinds of methods.
(6) Very few students from the two countries used logical reasoning strategies (CN
2%; SG 1.3%) though their success rates are high.
(7) A higher percentage of the Singapore students had no strategies than that of their
Chinese peers (SPD = 9%).

For Problem 3, we can have the following abservations:

(1) A majority (55%) of the Chinese students used arithmetic strategies, especially for
the Primary-6 Chinese students. However, only 10% of the Singapore students
used this method. The Chinese students used this method more successfully than
the Singapore students (SRD = 30%).
(2) About 36% of the Chinese students used algebraic strategies. Higher percentages
of the secondary Chinese students used this method than that of the Primary-6
Chinese students (SPD > 40%). However, very few Singapore students were
found to use algebraic method. The secondary Chinese students used this method
more successfully than the Primary-6 Chinese students (SRD > 15%).
(3) About half of the Singapore students used model drawing method, but very few
Chinese students did so. The success rate of this method of the Singapore students
is not low (SR = 54%), especially for the Primary-6 students (SR = 66%).
However, the secondary Singapore students used this method much less
successfully than the Primary-6 Singapore students (SRD > 18%).
(4) Up to 40% of the Singapore students used unitary methods for solving this
problem. The secondary Singapore students used this method much less

139
 successfully than the Primary-6 Singapore students (SRD > 11%).
(5) A lower percentage of the Singapore students had no strategies than that of the
Chinese students (SPD = 6%). This is also true for the samples at each of the three
grade levels.

In conclusion, (a) the Singapore students used the inappropriate arithmetic

strategies more frequently than their Chinese peers; (b) the Chinese students used
algebraic strategies more frequently and more successfully than their Singapore peers;
(c) the Singapore students used more varied strategies including guess-and-check,
model drawing and unitary methods, etc. However, the secondary Singapore students
could not use these strategies as successfully as the Primary-6 Singapore students.

Summary and Discussion

Cross-national studies provide us with an opportunity to ascertain the strength
and weakness of educational systems (Robitaille & Travels, 1992), and consequently
provide information about how to improve teaching and learning of mathematics (Cai,
2000, 2004; Robitaille & Travers, 1992). This comparative study has serveral
implications for the teaching and learning of mathematics in schools.
First, the teaching of various heuristics in Singapore schools has equipped the
Singapore students with more ways to tackle problems. This kind of practice has also
made the Primary-6 Singapore students have no strategy in a lower percentage of the
cases than their Chinese peers. The use of model drawing and unitary methods has
converted the multiplication and division of fractions (Problem 3) to the
multiplication and division of whole numbers. This kind of conversion might have
removed the difficulties that the Chinese students experienced. This finding may
explain the Singapore students’ good performance in number and proportionality in
the TIMSS studies. In the TIMSS studies, the secondary Singapore students
performed the best in number and proportionality (Beaton, et al., 1996; Mullis, et al.,
2000; Mullis, et al., 2004). Therefore, it is suggested to teach the Chinese students the
problem-solving heuristics that are proved to be effective, such as guess-and-check,
model drawing and unitary methods.
Second, the strategy analyses show that the secondary Singapore students do not
use algebraic strategies as frequently as the Chinese students. Problem 1 is similar to
the Cows and Chickens Problem in the study of Kaur (1998) and the Correct Answer
Problem in the study of Loh (1991). The results obtained in the study are consistent
with the studies of Kaur and Loh in that (a) few Secondary-2 Singapore students were
found to use algebraic methods, (b) the Singapore students preferred to use
guess-and-check methods and could use them efficiently, and (c) about one-third of
them used irrelevant procedures. Consequently, effort needs to be made to help the
secondary students recognize the strength of algebraic methods for solving word
problems as well as to bridge the gap between primary and secondary mathematics
(Fong, 1994).
Third, the secondary Singapore students used the arithmetic, model drawing,
unitary, and guess-and-check methods less successfully than the Primary-6 Singapore

140
students. Therefore, effort needs to be made to help the secondary students
consolidate what they have learnt in primary schools.

Appendix A
1. On Sunday, Judy went to see her grandma who lives 150 km apart. After cycling
at an average speed of 15 km/h for a few hours, she got tired and took a lift from
the passing truck. The truck’s average travelling speed is 75 km/h. When she got
to her grandma's house, she checked the time and knew that the trip took her 6
hours. Find the time she cycled?
2. Sunday morning, Rebecca and her parents went out to enjoy the natural scenery.
On the way to the destination, they travelled at a slow speed of 40 km/h. On the
way back, they drove at a faster speed of 120 km/h. When they came back home,
they found that they had been out for 2 hours. Find the average speed for this
round trip (Ignoring time around the destination).
3. Mike made a journey from City P to City Q. In the first half an hour, he covered
1 1
of it. In the second half an hour he covered of the remaining journey.
7 3
Finally he took another half an hour to finish the journey at a speed of 72 km/h.
Calculate his average speed for the whole journey.

Reference:
Acredolo, C., & Schmid, J. (1981). The understanding of relative speeds, distances, and durations of
movement. Developmental Psychology, 17(4), 490-493.
Beaton, A. E., Mullis, I. V. S., Martin, M. O., Gonzalez, E. J., Kelly, D. L., & Smith, T. A. (1996).
Mathematics achievement in the middle school years: IEA's Third International Mathematics
and Science Study (TMSS). TIMSS International Study Center, Boston College.
Beijing Normal University Press (BNUP) (China). (2001). !
[Curriculum standards for school mathematics in the full-time nine-year compulsory education
(Experimental Version)]. China: Beijing Normal University Press.
Bell, A., Fischbein, E., & Greer, B. (1984). Choice of operation in verbal arithmetic problems: The
effects of number size, problem structure and context. Educational Studies in Mathematics,
15(2), 129-147.
Bowers, J. S., & Nickerson, S. N. (2000) Students’ changing views of rates and graphs when
working with a simulation microworld. Focus on Learning Problems in Mathematics, 22(3-4),
10-25.
Cai, J. (2000). Mathematical thinking involved in U.S. and Chinese students’ solving
process-constrained and process-open problems. Mathematical Thinking and Learning, 2(4),
309–340.
Cai, J. (2004). Why do U.S. and Chinese students think differently in mathematical problem

141
 solving? Impact of early algebra learning and teachers’ beliefs. Journal of Mathematical
Behavior, 23(2), 135-167.
Fischbein, E., Deri, M., Nello, M. S., & Marino, M. S. (1985). The role of implicit models in solving
verbal problems in multiplication and division. Journal for Research in Mathematics
Education, 16(1), 3-17.
Fong, H. K. (1994). Bridging the gap between secondary and primary mathematics. Teaching &
Learning, 14(2), 73-84.
Fong, H. K. (1998). Solving challenging mathematical problems: The heuristic approach for
primary schools. Singapore: Kingsford Educational Services.
Fong, H. K. (1999). Strategic model for solving ratio and proportion problems. The Mathematics
Educator, 4(1), 34-51.
Fong, H. K., & Hsui, V. (1999). Strategy preferences and their association with hierarchical
difficulties of fraction problems. Science, Mathematics and Technical Education, 5, 3-12.
Gallagher, A. M., Lisi, R. D., Holst, P. C., Lisi, A. V., McGillicuddy-De, Morely, M., & Cahalan, C.
(2000). Gender differences in advanced mathematical problem solving. Journal of
Experimental Child Psychology, 75(3), 165-190.
Greer, B. (1992). Multiplication and division as models of situations. In D. A. Grouws (Ed.),
Handbook of Research on Mathematics Teaching and Learning (pp.276-295). New York:
Macmillan.
Hart, K. M. (1981). Fractions. In K. M. Hart, D. Kerslake, M. L. Brown, G. Ruddock, D. E.
Küchemann, & M. McCartney (Eds.), Children’s understanding of mathematics: 11-16 (pp.
66-81). The CSMS Mathematics Team.
Jiang, C.L. (2005). Strategies and errors in solving speed problems: A comparative study between
Chinese and Singapore students. Unpublished PhD dissertation, National Institute of
Education, Nanyang Technological University, Singapore.
Jiang, L. (1998a). ! ! ! !
! . [Nine-year Compulsory Education Six-year Primary Textbook Series: Experimental
Mathematics 6A. China: Hubei Science and Technology Publishers].
Jiang, L. (1998b). ! ! ! !
! . [Nine-year Compulsory Education Six-year Primary Textbook Series: Experimental
Mathematics 6B. China: Hubei Science and Technology Publishers].
Kaur, B. (1998). Mathematical problem solving in Singapore schools. Teaching & Learning, 19(1),
67-78.
Kho, T. H. (1987). Mathematical models for solving arithmetic problems. In Institute of Education
(Singapore), Mathematics Education in the 1990’s. Proceedings of Fourth Southeast Asian
Conference on Mathematical Education (ICMI-SEAMS) (pp.345-351), Singapore: Institute of
Education, June 1-3, 1987.
Krulik, S., & Rudnick, J.A. (1988). Problem solving: A handbook for elementary school teachers.
Boston: Allyn and Bacon.
Loh, W. F. (1991). Secondary school students' understanding of algebraic expressions and
equations. Unpublished master’s thesis, National Institute of Education, Nanyang
Technological University, Singapore.
Mayer, R. E. (1981). Frequency norms and structural analysis of algebra word problems into
families, categories, and templates. Instructional Science, 10(2), 135-175.

142
Ministry of Education (MOE) (Singapore). (1990a). Mathematics syllabus: Primary 1 to 3, Primary
4 to 6 (Normal Course), Primary 4 to 8 (Extended Course). Ministry of Education (Singapore):
Curriculum Planning Division.
Ministry of Education (MOE) (Singapore). (1990b). Mathematics syllabus: Secondary 1 to 2
(Special/Express course), Secondary 1 to 2 (Normal course). Ministry of Education
(Singapore): Curriculum Planning Division.
Ministry of Education (MOE) (Singapore). (2000a). Primary Mathematics 6A (3rd ed.). Ministry of
Education (Singapore): Curriculum Planning and Development Division.
Ministry of Education (MOE) (Singapore). (2000a). Primary Mathematics 6A (3rd ed.). Ministry of
Education (Singapore): Curriculum Planning and Development Division.
Ministry of Education (MOE) (Singapore). (2001a). Lower Secondary Mathematics Syllabus,
Special/express, Normal academic, Normal technical. Ministry of Education (Singapore):
Curriculum Planning and Development Division.
Ministry of Education (MOE) (Singapore). (2001b). Primary Mathematics Syllabus. Ministry of
Education (Singapore): Curriculum Planning and Development Division.
Mullis, I. V. S., Martin, M. O., Gonzalez, E. J., & Chrostowski, S. J. (2004). TIMSS 2003
International Mathematics Report: Findings from IEA’s Trends in International Mathematics
and Science Study at the fourth and eighth grades. TIMSS & PIRLS International Study Center,
Lynch School of Education, Boston College.
Mullis, I. V. S., Martin, M. O., Gonzalez, E. J., Gregory, K. D., Garden, R. A., O’Connor, K. M., et
al. (2000). TIMSS 1999: Findings from IEA’s Repeat of the Third International Mathematics
and Science Study at the Eighth Grade, International Mathematics Report. Boston College,
Chestnut Hill, MA, USA.
People’s Education Press (PEP) (China). (1992). !
! [The nine-year compulsory education three-year secondary school textbook series:
Algebra 1A]. PEP: Department of Junior Middle School Mathematics.
People’s Education Press (PEP) (China). (1993a). !
! [The nine-year compulsory education three-year secondary school textbook
series: Algebra 1B]. PEP: Department of Junior Middle School Mathematics.
People’s Education Press (PEP) (China). (1993b). !
! [The nine-year compulsory education three-year secondary school textbook series:
Algebra 2]. PEP: Department of Junior Middle School Mathematics.
People’s Education Press (PEP) (China). (1994). ! :!
! [The nine-year compulsory education six-year primary school textbook series: Mathematics
3A]. PEP: Department of Primary Mathematics.
Piaget, J. (1970). The child’s conception of movement and speed. (G. E. T. Holloway & M. J.
Mackenzie, Trans.). NY: Routledge & K. Paul.
Robitaille, D. F., & Travers, K. J. (1992). International studies in achievement in mathematics. In D.
A. Grouws (Ed.), Handbook for research on mathematics and teaching (pp.687-709). NY:
Macmillan.
Teh, K. L., & Looi, C. K. (2002a). New Syllabus D Mathematics 1 (5th ed.). Singapore: Shing Lee
Publishers Pte Ltd.
Teh, K. L., & Looi, C. K. (2002b). New Syllabus D Mathematics 2 (5th ed.). Singapore: Shing Lee
Publishers Pte Ltd.

143
van de Walle, J. A. (1993). Elementary school mathematics: Teaching developmentally. NY:
Longman Publishing Group.
Yuen, C. L. (1995). Strategies and errors of children in solving fraction problems. Unpublished
master’s thesis, National Institute of Education, Nanyang Technological University, Singapore.
Zhang, D., Li, S., & Tang, R. (2004). The two basics: Mathematics teaching and learning in
mainland China. In L. Fan, N. Y. Wong, J. Cai, & S. Li (Eds.), How Chinese learn mathematics:
Perspectives from insiders (pp.189-207). World Scientific Publishing Co. Pte. Ltd.
Zhou, Z., Peverly, S. T., Boehm, A. E., & Lin, C. (2000). American and Chinese children’s
understanding of distance, time, and speed interrelations. Cognitive development, 15(2),
215-240.
Zhou, Z., Peverly, S. T., & Lin, C. (2004). Cross- and within-cultural variations in children’s
understanding of distance, time, and speed interrelationship: A follow-up study. The Journal of
Genetic Psychology, 165(1) 5-27.

144
TSG 19: Research and Development in Problem Solving in Mathematics Education

Beyond Show and Tell: Neriage for Teaching through Problem-Solving

- Ideas from Japanese Problem-Solving Approaches for Teaching Mathematics -

Akihiko Takahashi, Ph.D.

DePaul University

Abstract

Japanese teachers use problem solving as a powerful approach for teaching mathematics. There
are several notable characteristics of the Japanese approach to problem solving. One of the
characteristics is that Japanese problem solving lessons usually do not end even after each
student finds a solution to the problem. Japanese teachers and researchers believe that the heart
of the lesson begins after students come up with solutions. The teacher facilitates extensive
discussion with students, which is called Neriage, by comparing and highlighting the similarities
and differences among students’ solutions. In this paper, some of the characteristics of the
Japanese approach for teaching mathematics will be discussed by focusing on the heart of the
approach, Neriage.

145
 Introduction

Problem solving has been a major focus in Japanese mathematics curricula for nearly a
half century. Numerous teacher reference books and lesson plans using problem solving have
been published since the 1960s. Government authorized mathematics textbooks for elementary
grades, which are published by six private companies, have had more and more problem solving
over the years. As a result, almost every chapter in recent Japanese mathematics textbooks for
elementary grades begins with problem solving as a way to introduce new concepts and ideas to
students.
A few key publications have greatly influenced how problem solving is used in Japanese
mathematics education. Polya’s How to Solve It (Polya, 1945, p. 26) was translated and
published in Japanese in 1954, and had been studied by various researchers and educators in
Japan. Japanese researchers, teachers, and administrators worked collaboratively through Lesson
Study, a professional development approach that is popular in Japan, to develop mathematics
instruction by referring to Polya’s (1945) four phases of problem solving work (Takahashi,
2000). One of the results from the studies of problem solving, Open-ended Approach, was
published in 1977 by Shimada et al. The open-ended approach has been widely used in Japanese
classrooms since then. Moreover, the English translation of the book was published (J. P. Becker
& Shimada, 1997) and has been popular among educators in the U.S. too. The Ministry of
Education in Japan has recognized the importance of problem solving in school mathematics and
emphasized the need for students to develop problem-solving skills to learn and use mathematics
in various documents since the beginning of the 1980s. The position statement from the NCTM's
An agenda for action: Recommendations for school mathematics of the 1980s (1980) that
"problem solving must be the focus of school mathematics" was referenced in various research
articles and resource materials for teachers in Japan during the 1980s. Also, Teaching Problem
Solving: What, why & how (Charles & Lester, 1982) was translated into Japanese in 1983.
Stigler and Hiebert (1999) described Japanese mathematics lessons as “structured
problem solving”. Similar characteristics of Japanese mathematics lessons were also reported in
the proceedings of the U.S.-Japan Seminar of Mathematical Problem Solving (Jerry P. Becker &
Miwa, 1987; Jerry P Becker, Silver, Kantowski, Travers, & Wilson, 1990). Structured problem
solving is designed for students to acquire knowledge and skills through creative mathematical
activity by presenting challenging problems to students. Students are expected to solve a problem

146
using their own mathematical knowledge. Thus, Japanese teachers usually do not tell students
how to solve a problem before students try to solve the problem by themselves. Working with
problems, students bring several different approaches and solutions to the class. The teacher then
leads students in a whole-class discussion in order to compare individual approaches and
solutions. This whole-class activity provides students with opportunities to learn mathematics.
Through their extensive study of problem solving, Japanese teachers and educators have come to
recognize that this whole-class discussion is the heart of structured problem solving and have
named this discussion part Neriage.
In this paper, I will discuss the Japanese approach of using problem solving for teaching
mathematics, structured problem solving, by focusing on the heart of the approach, Neriage.
The Japanese Problem Solving Approach

One of the major goals of teaching mathematics is to help students become able to solve
problems. Thus, mathematics lessons employing problem solving are sometimes viewed as an
approach for students to develop problem-solving skills and strategies, and teachers sometimes
focus solely on the strategy of solving the problem and not necessarily on developing
mathematical concepts and skills. This interpretation of problem solving lessons usually ends
after each student comes up with a solution to the problem. The teachers’ role during students’
problem solving is to help students find the solution by providing an efficient strategy, because
the major goal of the lesson is for students to solve problems.
On the other hand, problem solving can also be viewed as a powerful approach for
developing mathematical concepts and skills. Thus, in this approach teachers use problem
solving not only for lessons that solely focus on developing problem-solving skills and strategies
but also on lessons that develop mathematical concepts, skills, and procedures. As a result the
lesson plans for this approach usually include content goals in addition to goals for developing
problem solving strategies and skills.
To highlight the difference between the two approaches, the latter approach is often
called "teaching though problem solving." Since Japanese structured problem solving uses
problem solving as a process for learning mathematical content, it could be considered a type of
teaching through problem solving. Looking at problem solving as a process for students to learn
mathematics is not unique to Japanese mathematics education. The National Council of Teachers
of Mathematics (NCTM) has also emphasized the importance of teaching mathematics through

147
problem solving (Mathematics, 2006; National Council of Teachers of Mathematics, 1989, 2000,
2006). These documents discuss the necessity of learning mathematical content through the
processes of problem solving, reasoning and proof, communication, connections, and
representation. Various reform documents have also suggested that mathematics lessons should
be designed to provide students learning opportunities through the processes of problem solving
and not simply by listening to teachers’ lectures.
Although teaching through problem solving has been suggested by NCTM and other
reform documents in the U.S., it is hard to find lessons that employ this idea in U.S. classrooms.
Stigler and Hiebert argue that Japanese mathematics lessons better exemplify current U.S. reform
ideas, such as teaching through problem solving, than do typical U.S. mathematics lessons, based
on the TIMSS videotape classroom study (1997). One of the reasons behind this phenomenon
might be that Japanese mathematics teaching already had a history of focusing on developing
mathematical thinking skills by using a variety of story problems even before the idea of
problem solving was introduced. Therefore Japanese educators looked at problem solving as an
ideal approach for learning mathematics rather than simply a way to promote problem solving
skills when the idea of problem solving was introduced.
There are several notable characteristics of the Japanese problem solving approach. First,
the Japanese problem solving approach can be found throughout the curriculum because it is
designed for learning mathematical content. Japanese teachers have tried to use the approach for
students not only to develop concepts and understanding of mathematics but also to acquire skills
to learn and use mathematics. Therefore problem solving is not viewed as an end-of-the-chapter
activity that is solely focused on developing problem-solving skills and strategies. Second,
Japanese problem solving lessons usually do not end even after each student finds a solution to
the problem. Problem solving lessons that solely focus on developing problem solving and skills
often end after students share their solutions with the class. Japanese teachers and researchers,
however, believe that the heart of the lesson begins after each student comes up with a
solution(s). Japanese teachers facilitate extensive discussion with students, which is called
Neriage, by comparing and highlighting the similarities and differences among students’ solution
approaches.

148
 Extensive discussion (Neriage)

The term Neriage has been widely used among Japanese teachers and researchers of
mathematics education as a technical term since 1980s. Neriage is a noun form of a verb
Neriageru, which means to polish up. The term Neriage is used among Japanese teachers for
describing the dynamic and collaborative nature of a whole-class discussion in the lesson
(Shimizu, 1999). The most important role of the teacher in Neriage is to orchestrate students’
ideas and approaches to solve the problem and to help them polish their solutions in order to
learn mathematical content. During the process, a teacher highlights important mathematical
ideas and concepts for students to reach the goals of the lesson. This is why Japanese teachers
see Neriage as the heart of teaching mathematics through problem solving. From the viewpoint
of Japanese teachers, the solving of the problem by each student at the beginning of the lesson is
a preparation for Neriage. Therefore it is important for students to struggle with the problem and
find their own way to solve the problem, because this experience will be the foundation for
students to make a connection between their previous learning and the content that they are
going to learn through Neriage.
The following is an example of a series of problem solving activities from the most

Figure 1. Which cabin is the most crowded?

Reprinted from Mathematics for Elementary School 5B. Tokyo, Japan: Tokyo
Shoseki Co., Ltd. p.23 (Hironaka & Sugiyama, 2006)

widely used Japanese mathematics textbook series. This example shows how Neriage leads
students to acquire a new idea in mathematics through a series of problem solving activities. The

149
unit on Per Unit Quantity in the 5th grade textbook begins with the following problem (Figure 1)
without showing any solution or hint to the students on the page. The textbook is designed for
students to see everyday situations as mathematical problems. Most 5th grade students may
realize that the crowdedness of Cabin A and Cabin B may be compared by looking at the number
of the people who share each cabin and because the areas of the rooms look the same. Students
may also be able to see that the crowdedness of Cabin B and Cabin C may be compared by
looking at their area because the number of the people who share each cabin are the same. The
textbook expects teachers to facilitate the above discussion at the beginning of the lesson so that
students understand the situation. Then, the next page of the textbook (Table 2) provides further
information for students to understand what is the mathematical problem in this everyday
situation.
Area (m2) Number of people
Cabin A 16 6
Cabin B 16 5
Cabin C 15 5

Table 2. Which cabin is the most crowded?

Reprinted from Mathematics for Elementary School 5B. Tokyo, Japan:
Tokyo Shoseki Co., Ltd. p.24 (Hironaka & Sugiyama, 2006)

By looking at the data, students are able to compare the crowdedness of Cabin A and
Cabin B without any calculation. They also may be able to see which room is more crowded
between Cabin B and Cabin C. It is, however, not easy for students to figure out which is more
crowded between Cabin A and Cabin C. Through this discussion, teachers are expected to lead
students to understand what is the problem to be solved. There are two data for each cabin, the
area of the cabin and the number of people who share the cabin. It is easy to compare the
crowdedness of the cabins if one of these two data are the same. Since neither the area nor the
number of people for Cabin A and Cabin C are the same, it might not be possible to compare the
crowdedness by using the data. If so, are there any ways to make one of the data the same --
either the area or the number of the people?
After the above discussion, each student is encouraged to figure out which is more
crowded, Cabin A and Cabin C. Since the numbers in the problem are carefully chosen, there are
several approaches for students to compare crowdedness. The following are four commonly seen
student solution methods.

150
 Method A:
Cabin A: 6 ÷16 = 0.375
Cabin B: 5 ÷15 = 0.33L
Division is used to find how many people occupy 1 m2. Because a larger
!
number of people would occupy 1 m2, Cabin A is more crowded.
!
Method B:
Cabin A: 16 ÷ 6 = 2.66L
Cabin B: 15 ÷ 5 = 3
Division is used to find how many square meters there are per person.
!
Because there is less area per person, Cabin A is more crowded.
!
Method C:
A common multiple of 16 and 15 is 240
Cabin A: 6 "15 = 90
Cabin B: 5 "16 = 80
This method looks at how many people would share each cabin if both
!
cabins have the same area. In order to use this method, a common multiple 240 is
!
found as the area of each cabin. Because more people would share 240 m2, Cabin
A is more crowded.
Method D:
A common multiple of 6 and 5 is 30
Cabin A: 16 " 5 = 80
Cabin B: 15 " 6 = 90
This method looks at how much area would be shared by a person if both
!
cabins have the same area. In order to use this method, a common multiple 30 has
!
been found as the number of people in each cabin. Because less area would be
occupied by a person, Cabin A is more crowded.
After students come up with solution methods that include the above four methods, the
teacher assigns students to share their solutions. Japanese teachers usually monitor students’
work during their individual or group problem-solving time and come up with a plan for the
discussion. For example, teachers often use a seating chart of the class to jot down how each
student approaches the problem as well as to plan how to lead the discussion. A teacher might

151
ask one of the students who used the most common method to share his/her method to begin the
discussion, and then ask another student to share different methods. At this time, Japanese
teachers usually do not say that the answers are right or wrong in order to provide students with
opportunity to think carefully about each solution method. Teachers carefully use blackboard
writing for students to see all the solution methods from their peers and to help them understand
each method.
The Neriage begins after the students present their various solution methods. Until
Neriage begins, the whole class activity is very similar to the children's favorite school activity,
Show and Tell. Neriage is an activity that goes beyond Show and Tell, however. If the goal of
the problem-solving lesson is just to find a solution to the problem, this could be the end of the
lesson. But because the Japanese problem-solving lesson is designed for students to learn new
mathematical knowledge, the Neriage is necessary.
Teachers might begin the Neriage by asking students to see if there are some common
ideas or approaches among the solution methods or some differences. For example, students
might notice that both Method A and Method B use division but in different ways. In contrast
Method C and Method D use multiplication instead of division. On the other hand, Method A
and Method C use the same idea that looks at what if both cabins had the same area. Method B
and Method C also share the same idea that looks at what if both cabins were shared by the same
number of people. During the discussion, teachers would provide opportunities to think about
why using addition and subtraction might not be the best approach to this problem. This
comparison would help students deepen their understanding of the concept and the use of
multiplication and division concerning ratio and rate. If there are some methods based on
students’ misunderstanding, teachers can use them to help students develop reasoning skills to
justify whether the solutions are right.
Then, teachers would lead students to see whether each approach has advantages and
limitations. Students might realize that the methods that use a common multiple might have a
limitation if the number in the problem becomes larger or if the situation requires them to
compare more than two rooms. Finding a common multiple with large numbers or with several
numbers might not be an easy task. Considering this limitation, teachers might be able to
conclude that the use of division to find a unit quantity might be a better way for the similar
problem with a more complex situation. In fact, the Japanese textbook series gives students the

152
following problem after “Which cabin is the most crowded?” for students to actually see which
method, multiplication or division, is most useful.
At Yoshiko’s farm, which is 600m2, 1968kg of potatoes were produced. At Tadashi’s
farm, which is 900m2, 2682kg of potatoes were produced. Which farm was better at
producing potatoes? (Hironaka & Sugiyama, 2006)
Teachers would also lead students to see whether there are any advantages to Method A
and Method B. Comparing these two methods requires students to think deeply about the
meaning of division. Moreover, this opportunity can help students to understand practical
application of the use of division, since the order of division requires students to use different
interpretations of the quotients. The quotients in Method A show that the larger quotient
indicates more crowdedness. The quotients in Method B show that the larger quotient indicates
less crowdedness. To use a number to represent crowdedness, the quotients in Method A are
more understandable, because the larger number means that it is more crowded. Later in the unit,
this Japanese textbook introduces population density, comparing crowdedness by using the
number of people who live in an area of 1km2, and asking students to solve the problem.
Neriage is a critical component of a Japanese problem-solving lesson because this is the
place where teachers can teach students new mathematical ideas and concept by using students’
solution methods. Because the Neriage is built upon the students’ solutions as a foundation of the

Figure 2. Population dencity

Reprinted from Mathematics for Elementary School 5B. Tokyo, Japan: Tokyo
Shoseki Co., Ltd. p.27 (Hironaka & Sugiyama, 2006)

dynamic and collaborative whole-class discussion, Japanese teachers put so much effort to
preparing the discussion.

153
 Lesson Planning for Problem Solving Lesson: Beyond Show and Tell

At the beginning of the development of the Japanese problem solving approach, the term
Neriage had not been widely used. Instead, educators used the Japanese translation of terms from
Polya’s four phases of problem solving—understanding the problem, devising a plan, carrying
out the plan, and looking back—as a framework to design lessons. For example, 1) the problem
solving lesson usually began with the presentation of the problem of the day by the teacher and
the teacher helping students to understand what the problem really is, thus, "understanding the
problem", 2) then the teacher asked each student to devise a plan to solve the problem, thus
"devise a plan", 3) next, students solved the problem by using their previously learned
knowledge and skills, thus "to solve the problem based on the plan," and finally, 4) students
examined whether their solution was correct and the method that they used was reasonable and
efficient, thus, "looking back." Although Polya’s four phases of problem solving had been used
as the foundation of Japanese problem solving lessons, Japanese teachers revised the framework
over the years through lesson study. There are two notable changes over the years. First, the 2nd
phase, devise a plan, was omitted from the flow of the lesson. Second, to describe the 4th phase,
looking back, the term Neriage began to be used. These changes did not happen in a top-down
manner. Through numerous research lessons and post-lesson discussions, teachers gradually
shifted their use of problem solving and reached the problem solving approach which Stigler and
Hiebert (1999) described as structured problem solving.
Teaching mathematics through problem solving is not an easy task for teachers,
especially facilitating a good discussion, Neriage. To develop problem-solving lessons, Japanese
teachers usually begin by considering the following three major issues—the curriculum, the
students, and the problem. By investigating the curriculum teachers should be able to identify
what content should be taught. Japanese teachers try to identify the contents as specific as
possible so that the lesson will be focused on a specific topic(s). Then, they examine students'
previous learning to identify the goal of the problem-solving lesson and to identify the problem
for the lesson. Japanese teachers seek meaningful problems for problem solving lessons by
looking at Japanese mathematics textbooks and resources materials. These resource materials
include lesson study reports and books published by experienced lesson study practitioners.
These three issues can be discussed in any order.

154
 Sometimes Japanese teachers develop lessons for lesson study from scratch but they often
develop lessons by using others' work and modify them to fit into their own students’ needs.
One of the challenges is to find a good problem that can lead students to accomplish the
goals of the lesson. There are a lot of interesting and engaging problems, including puzzles and
games; however, the problem should be able to foster students' ability to learn something new
after they have solved the problem by using their existing knowledge and skills. In other words,
when students solve the problem it is expected that the problem provides students with
opportunities for see a need for learning new knowledge and skills, which is the goal of the
lesson.
A group of teachers carefully examine problems for students. They always solve each
problem by themselves in several different ways in order to examine whether the problem is
mathematically meaningful for the students at the time of the lesson. Then, teachers begin to
anticipate students’ responses to the problem including ones based on misunderstandings and
misuses of previous learning. Then they start to design the flow of the lesson so that students will
be able to reach the goal of the lesson through problem-solving. Coming up with good
questioning for students to think deeply about mathematics is always a challenge for teachers
when designing lessons.
One of the major tasks for Japanese teachers is to facilitate meaningful mathematical
discussion during the whole-class activity to help students to achieve the goals of the lesson.
When a teacher presents a problem to students without giving a procedure, it is natural that
several different approaches to the solution will come from the students. Thus, the textbooks
include examples of students’ typical approaches and ideas. Because the goal of the structured
problem-solving approach is to develop students’ understanding of mathematical concepts and
skills, a teacher is expected to facilitate mathematical discussion for students to achieve this goal.
This discussion is often called Neriage in Japanese, which implies polishing ideas. In order to do
this, teachers need a clear plan for the discussion as a part of their lesson plans, which will
anticipate the variety of solution methods that their students might bring to the discussion. These
anticipated solution methods include not only the most efficient methods but also ones caused by
students’ misunderstandings. Thus, anticipating students’ solution methods is a major part of
lesson planning for Japanese teachers.

155
 Towards the end of a lesson, a teacher often leads the lesson in a way that pulls all the
different approaches and ideas together in order to see the connection. Then, he or she
summarizes the lesson to help students achieve the objective of the lesson. The teacher often asks
students to reflect on what they have learned during the lesson.
Japanese teachers and researchers believe that Neriage is the most important component
of teaching mathematics through problem solving. Therefore, teachers spend much time
investigating various resources through Kyozaikenkyu in lesson study (Watanabe, Takahashi, &
Yoshida, in press). The quality of Neriage depends upon how well the teacher(s) plan the lesson,
because this is the place where teachers have to use all their knowledge of mathematics, their
knowledge about teaching mathematics, their knowledge of students, and their skills to facilitate
the whole-class discussion. Therefore, Japanese educators believe that teaching, especially in
Neriage, is the proving ground of teachers’ knowledge and skills (Fujii, in press).
Conclusion

This paper focuses on the term Neriage to highlight the characteristics of a Japanese
teaching approach based on problem solving. There are several other technical terms in the
Japanese educational community. These technical terms are sometimes used among teachers to
describe and discuss specific events or techniques in teaching and learning. This means that a
term such as Neriage is used only among teachers, and people in other professions do not share
the meaning of the term. The existence of such technical terms tells us that Japanese teachers
have an opportunity to discuss teaching and learning with their colleagues regularly.
The Japanese teaching profession has a long history of collaboration among teachers to
discuss how to improve teaching and learning. The Japanese problem-solving approach is one of
the outcomes of this collaboration. Thus, the development of collegial communities among
teachers is an ideal way to make a shift in the teaching and learning of mathematics by using
suggestions from various documents.

156
 References

Becker, J. P., & Miwa, T. (1987). Proceedings of the U.S.-Japan Seminar on Mathematical
Problem Solving (Honolulu, Hawaii, July 14-18, 1986) (COLLECTED WORKS -
Conference Proceedings No. INT-8514988): Southern Illinois Univ.,
Carbondale.[JIM81075].
Becker, J. P., & Shimada, S. (1997). The open-ended approach: A new proposal for teaching
mathematics. Reston, Virginia: National Council of Teachers of Mathematics.
Becker, J. P., Silver, E. A., Kantowski, M. G., Travers, K. J., & Wilson, J. W. (1990). Some
Observations of Mathematics Teaching in Japanese Elementary and Junior High Schools.
Arithmetic Teacher, 38(2), 12-21.
Charles, R., & Lester, F. (1982). Teaching Problem Solving: What, why & how. Palo Alto, CA:
Dale Seymour Publications.
Hironaka, H., & Sugiyama, Y. (Eds.). (2006). Mathematics 5B for Elementary School. Tokyo,
Japan: Tokyo Shoseki Co., Ltd.
National Council of Teachers of Mathematics. (1980). An agenda for action: Recommendations
for school mathematics of the 1980s. Reston, Virginia: National Council of Teachers of
Mathematics.
National Council of Teachers of Mathematics. (1989). Curriculum and evaluation standards for
school mathematics. Reston, Virginia: National Council of Teachers of Mathematics.
National Council of Teachers of Mathematics. (2000). Principles and standards for school
mathematics. Reston, Virginia: National Council of Teachers of Mathematics.
National Council of Teachers of Mathematics. (2006). Curriculum Focal Points for
Prekindergarten through Grade 8 Mathematics: A Quest for Coherence.
Polya, G. (1945). How to solve it: A new aspect of mathematical method. Princeton, New Jersey:
Princeton University Press.
Shimizu, Y. (1999). Aspects of Mathematics Teacher Education in Japan: Focusing on Teachers'
Role. Journal of Mathematics Teacher Education, Vol.2(No.1), 107-116.
Stigler, J., & Hiebert, J. (1997). Understanding and imporving mathematics instruction: An
overview of the TIMSS video study. Phi Delta Kappan, 79(1), 14-21.
Stigler, J., & Hiebert, J. (1999). The teaching gap: Best ideas from the world's teachers for
improving education in the classroom. New York: Free Press.
Takahashi, A. (2000). Current trends and issues in lesson study in Japan and the United States.
Journal of Japan Society of Mathematical Education, 82(12), 15-21.
Watanabe, T., Takahashi, A., & Yoshida, M. (in press). Kyozaikenkyu: A critical step for
conducting effective lesson study and beyond. In F. Arbaugh & P. M. Taylor (Eds.),
Inquiry into Mathematics Teacher Education. Association of Mathematics Teacher
Educators (AMTE) Monograph Series, Volume 5.

157
 Instructional Practices to Facilitate Prospective Mathematics Teachers’
Learning of Problem Solving for Teaching

Olive Chapman
University of Calgary

Paper presented at ICME 11, Monterrey, Mexico, July 2008

Introduction
Problem solving is considered central to school mathematics. It is highlighted in reform
documents as a key factor of change in mathematics education (NCTM 1989, 1991, 2000). As
NCTM (2000, p. 52) states,
Instructional programs should enable all students to build new mathematical knowledge
through problem solving; solve problems that arise in mathematics and in other contexts;
apply and adapt a variety of appropriate strategies to solve problems; and monitor and reflect
on the process of mathematical problem solving.
Similarly, Kilpatrick, Swafford, and Findell (2001, p. 420) explained,
Studies in almost every domain of mathematics have demonstrated that problem solving
provides an important context in which students can learn about number and other
mathematical topics. Problem-solving ability is enhanced when students have opportunities to
solve problems themselves and to see problems being solved. Further, problem solving can
provide the site for learning new concepts and for practicing learned skills.
Thus, problem solving is important as a way of doing, learning and teaching mathematics.
However, whether or how such ways of viewing problem solving get implemented in the
classroom will depend on the teacher. In addition, Schoenfeld (1985) has identified four aspects of
students‟ problem solving that can be used to guide instruction – resources, heuristics, control, and
beliefs. This means that teachers have to play a central role in helping students choose resources,
implement heuristics, control their problem solving actions, and develop useful beliefs.
If problem solving should be taught to students, then it should be taught to prospective
teachers who are likely to enter teacher preparation programs without having been taught it in an
explicit way. If it is to form a basis of teaching mathematics, then prospective teachers should
understand it from a pedagogical perspective. There are studies, discussed later, that raise issues
about prospective teachers‟ understanding of problem solving and ability as problem solvers that
could affect what and how they implement problem solving in their teaching. It therefore seems to
be important that teacher education includes learning opportunities explicitly focused on problem
solving. In this paper, I draw on studies, including my own work, that include instructional
practices to facilitate prospective teachers‟ learning of problem solving and problem-solving
pedagogy in order to highlight the nature of these practices and the learning that results from them
and to discuss key characteristics of the practices that have implications for how we prepare
prospective teachers to use problem solving in their teaching. The paper, then, is based on a
review of research literature and a report of a study I conducted to identify (i) prospective
mathematics teachers‟ knowledge and ability of problem solving; (ii) instructional approaches to
facilitate their learning of problem solving; and (iii) implications for teacher education.

158
Prospective Teachers’ Knowledge of Problem Solving
Studies focusing explicitly on prospective teachers‟ knowledge of problem solving are a scarcity
in the research literature, regardless of whether routine or non-routine problems are considered.
What are available deals mainly with the prospective teachers‟ ability or strategies in solving word
problems, as in these examples. Schmidt and Bednarz (1995) examined modes of problem solving
that 131 prospective elementary and secondary teachers used in arithmetical and algebraic word
problems to identify the resistance and eventual difficulties that arose in the shift from one type of
approach to the other. The majority of them confined themselves to algebra even when dealing
with arithmetical problems. The prospective elementary teachers appeared to be the best prepared
for addressing both fields, i.e., used arithmetic for arithmetic problems and algebra for algebra
problems. van Dooren, Verschaffel, and Onghena (2003) investigated the arithmetic and algebraic
word problem-solving skills and strategies of 97 prospective elementary and secondary teachers,
both at the beginning and at the end of their teacher preparation. They found that the prospective
secondary teachers clearly preferred algebra, even for solving very easy problems for which
arithmetic was more appropriate. About half of the prospective elementary teachers adaptively
switched between arithmetic and algebra, while the other half experienced serious difficulties with
algebra. Contreras &Martínez-Cruz (2001) examined 68 prospective elementary teachers‟ solution
processes to a word problem involving division of fractions in which the numerical answer to the
division did not provide the appropriate solution to the problem when the realities of context of the
problem were considered. They found that the participants did not always base their responses on
realistic considerations of the context situation. Only 28% of their responses contained a realistic
solution to the given problem. Contreras and Martinez-Cruz (2003) also examined 139 prospective
elementary teachers‟ solution processes to additive word problems for which the solution was one
more or one less than that produced by simply adding or subtracting the given numbers. They
found that about 91% of the prospective teachers‟ responses contained incorrect solutions to the
problems based on their failure to interpret correctly the solution produced by addition or
subtraction of the two numbers given in each word problem. Simon (1990), in his investigation of
33 prospective elementary teachers‟ knowledge of division found that they failed to connect their
understandings of division and the semantic features of the word problems to the procedures that
they employed to divide. Finally, Verschaffel, De Corte, and Borghart (1996) investigated 332
prospective teachers' conceptions and beliefs about the role of real-world knowledge in arithmetic
word problem solving. For each of the 14 word problems, the prospective teachers were first asked
to solve the problem themselves, and then to evaluate four different answers given by students.
The results revealed a strong overall tendency among the participants to exclude real-world
knowledge and realistic considerations from their own spontaneous solutions of school word
problems as well as from their appreciations of the students‟ answers.
In addition to the above studies that focused on word problems, two studies, Taplin (1996)
and Leung (1994), dealt with problem-solving approaches, more generally, and problem-posing
processes, respectively. Taplin (1996) explored the approaches to problem solving used by 40
prospective elementary teachers and found that they preferred to work with a narrow range of
strategies, predominantly verbal and numerical. They tended to select a method of approach and
not change from that through the tutorial, implying inflexibility in their choice or management of
problem-solving strategies. Leung (1994) analyzed problem-posing processes (i.e., posing a
sequence of problems in each problem-solving activity) of eight prospective elementary teachers
with differing levels of mathematics knowledge. Findings showed that those with high
mathematics knowledge systematically manipulated given conditions to make problems and used
solutions to prior posed problems as new pieces of information to pose subsequent problems.

159
Those with low mathematics knowledge posed problems that might not be solved mathematically
and the mathematics problems posed were not necessarily related in structure. My work
(Chapman, 2005) also examined prospective secondary mathematics teachers‟ knowledge of
problems and problem solving as part of a study to investigate an instructional approach to
enhance participants‟ knowledge of problem solving. The study indicated that in relation to their
initial knowledge, most of the participants made sense of problems in terms of the traditional,
routine problems they had experienced, directly or indirectly, prior to entering the teacher
education program. They also understood these problems as genuine problems that required
thought and logic to arrive at a solution. They understood the problem-solving process in a way
consistent with the traditional classroom approach of dealing with routine problems.
These studies imply concerns about how prospective teachers of mathematics may
conceptualize problem solving and engage in it. They provide evidence that prospective teachers
are likely to need help in their development and understanding of problem solving from the
perspectives of a learner and a teacher.
Instructional Practices for Problem Solving in Teacher Education
Based on a review of current research in mathematics teacher education for this paper, there seems
to be also a scarcity of published studies that explicitly address instructional practices for problem
solving in teacher education. In this section, I focus on five studies that include some form of open
or non-routine problems that were intended to play a role in the development of prospective
teachers‟ knowledge of non-routine problem solving and problem-solving pedagogy. Two of these
studies are explicitly about problem solving in terms of their stated goals. None defined problem
solving, but the implication is that it is associated with a way of thinking involved in solving non-
routine problems in the learning of mathematics. These studies provide some basis for instruction-
al practices in helping prospective teachers to grow in their knowledge of problem solving for
teaching. This aspect of these studies will be highlighted here in order to make explicit possible
ways of engaging prospective teachers in problem solving. In particular, the focus will be on the
approaches used in these studies to facilitate learning of problem solving; the goals relating to
problem solving of these approaches; and the effect of the approaches on the prospective teachers‟
learning of problem solving and its pedagogy. This sample of studies described next deals with
approaches to help prospective teachers understand: the nature of problems (Arbaugh & Brown,
2004); the problem solver and problem-solving pedagogy (Lee, 2005); the problem-solving
process (Ebby, 2000; Roddick, Becker & Pence, 2000; Szydlik, Szydlik, & Benson, 2003). All of
these studies directly or indirectly dealt with problem-solving pedagogy, e.g., how to facilitate
students‟ learning of problem solving. But there is less focus on teaching through problem solving.
(a) Nature of Problems (Arbaugh & Brown, 2004)
Although this study is not explicitly about problem solving, the focus on tasks in the form of
problems makes it relevant to problem solving, which begins with the selection of worthwhile
problems. The goal of this study was to help prospective teachers understand the relationship
between a task and the kind of thinking that task required of students. The tasks consisted of
mathematical problems, both routine and non-routine, categorized based on “level of cognitive
demand,” i.e., the type of thinking that the task required of students. For example, the category
that can be associated with genuine problem solving is “higher-level demands (doing
mathematics)” with the following characteristics: require complex and non-algorithmic thinking;
demand self-monitoring or self-regulation of one‟s own cognitive processes; require students to
access relevant knowledge and experiences and make appropriate use of them in working through
the task; require students to analyze the task and actively examine task constraints that may limit
160
possible solution strategies and solutions; require considerable cognitive effort and may involve
some level of anxiety for the student because of the unpredictable nature of the solution process
required (Arbaugh & Brown, 2004, p. 30).
The instructional approach consisted of a task-sorting activity. The researchers developed a set
of high school-based mathematical tasks to us in the task-sorting activity in order to help the
teachers to learn about the levels of cognitive demands. They used the task-sorting activity with
six different groups of prospective high school mathematics teachers.
The effect of the approach was that it was successful in helping the prospective teachers learn
the levels of cognitive demand criteria. Each group left their methods class with the ability to
categorize tasks according to the criteria, and overall the prospective teachers had been able to
communicate the importance of considering the cognitive level demanded of the tasks. They
learned to look past surface characteristics of the individual problems and analyze them on the
basis of the types of thinking required by students. They built knowledge about the activities that
had enabled them to reach a deeper understanding of worthwhile mathematical tasks and the
relationship between those tasks and students‟ learning.
(b) Problem-solver and Problem-Solving Pedagogy (Lee, 2005)
This is one of the studies that focused explicitly on problem solving with an approach that allowed
participants to learn about students as problem solvers and about problem-solving pedagogy. The
goal of the approach was to help the prospective teachers interpret and develop in their role of
facilitating students‟ mathematical problem solving with a technology tool.
The approach consisted of a cycle of “planning–experience–reflection” repeated twice during
an undergraduate course to allow the prospective teachers to change their strategies when working
with two different groups of students. The prospective teachers enacted the six phases of the cycle
by: (1) Individually solving the open-ended problem using a java applet and discussing the
problem with peers and the teacher educator/researcher. (2) Developing anticipatory ideas and
planning a learning trajectory for students. (3) Interacting with two students as they solved the
same problem with a java applet. (4) Discussing the experience with peers, reflecting, and
planning a revised learning trajectory for different students. (5) Interacting with two different
students as they solved the same problem with a java applet. (6) Reflecting on their role in
facilitating students‟ problem solving with technology and their understanding of what the
students understood about the problem. Several prompts in Phase 2 of the cycle helped the
prospective teachers think about students‟ learning trajectory by considering possible solution
strategies, difficulties students may have, and questions that might be asked to help students
overcome those difficulties. In Phase 4, the prospective teachers were asked to reflect on their
interactions with students, students‟ understanding and problem solving, and changes or
improvements desirable for the next group of students. In Phase 6, they were prompted to compare
the two experiences and to reflect on what may have caused any similarities or differences in their
interactions with the students and how the students solved the problem.
The effects of the approach, based on a study of three prospective teachers, was that the
planning–experience–reflection cycle provided opportunities for them to begin to struggle with
issues of facilitating students‟ problem solving and to make their struggle an open and reflective
activity used as an opportunity to improve their practice. Six themes emerged from the cases, i.e.,
the prospective teachers: (1) Used their own mathematical problem-solving approaches to
influence their pedagogical decisions. (2) Desired to ask questions that can guide students in their
solution strategies without „„giving it all away.‟‟ (3) Recognized their own struggle in facilitating
students‟ problem solving and seem focused on improving aspects of their interactions with
students. (4) Assumed the role of an explainer for part of each facilitation phase. (5) Made
161
pedagogical decisions to use representations in the java applet to promote students‟ mathematical
thinking or focus their attention on specific aspects of the problem. (6) Used the technology tools
in ways consistent with the nature of their interactions and perceived role with students.
(c) Problem-Solving Process (Roddick, Becker & Pence, 2000)
This is another study that explicitly focused on problem solving with an approach that allowed
participants to learn about the problem-solving process and some related pedagogical processes.
The goal of the approach was to influence prospective teachers' problem solving, problem posing,
modelling, and beliefs about the role of problem solving in teaching mathematics. The authors
organized two courses aimed at: improving prospective teachers‟ problem-solving abilities, their
learning of ways to assess problem solving, broadening their views of problem solving and
mathematics, and enhancing their understanding of equity issues in teaching mathematics.
In the approaches used in two courses, the prospective secondary teachers were provided with
rich and varied problem-solving experiences. They spent significant time on topics such as: what
is a problem; examination of problem solving in traditional and innovative curricula; equity issues
in problem solving and its assessment; assessment of problem solving and use of technology.
Students used a model for reflecting on one‟s problem solving (i.e., that of Mason, Burton, &
Stacey, 1985) and concentrated on specializing, generalizing, and justifying their work. Both
courses included substantial in-class time working in groups on problems and giving presentations
and justifications to the class.
The effect of the approach was to impact changes of the participants‟ beliefs and practice to
various degrees. The participants fell on a continuum ranging from not much discernible
implementation to substantial integration of problem solving in their teaching, as in the case of
one participant, described in detail. She experienced considerable growth in her views of problem
solving and its role in instruction and incorporated such learning into her teaching. The case study
demonstrated the changes that can occur in beliefs and instruction as a result of an intensive year-
long course that immerses prospective teachers in being reflective problem solvers themselves.
(d) Problem-Solving Process (Szydlik, Szydlik, & Benson, 2003)
This study involves problem solving in an indirect way as part of a mathematics content course for
prospective elementary teachers that was designed to provide participants with authentic
mathematical experiences and to foster autonomous mathematical behaviors, i.e., behaviors that
involve sense-making rather than memorization or appeals to authority. The implied goal of the
approach in relation to problem solving was to help the prospective teachers to become autono-
mous problem solvers by promoting community autonomy rather than autonomy of individuals.
The approach involved engaging the prospective teachers in authentic mathematical behaviors
arising out of community work on a set of demanding problems, each of which had an underlying
mathematical structure that formed a part of the course content. The problems generally allowed
for a variety of problem-solving strategies with a mathematical structure that can be discovered by
collecting data, solving a smaller version of the problem, considering several specific cases, or by
logical considerations. During a typical class meeting, the prospective teachers worked for 20 to
30 minutes on such a problem in small groups of three or four. The class then convened in a large
semicircle for a discussion of their findings, strategies, solutions and arguments. In these
discussions, the course instructor emphasized the necessity of mathematical justification; complete
solutions required logical arguments. The class was designated as the mathematical authority. The
instructor declined to give the final word on the correctness or completeness of any solution and
there was no text. The instructor also provided almost no assistance in the problem solving aspect
of the course and no answers were provided for problems. The only way for the class to
162
understand a problem was to figure it out. The only way to know they were correct was to find a
convincing argument. However the spirit of the class was consistently one of community inquiry.
On six occasions throughout the semester, the prospective teachers were asked to produce written
reports that focused on their mathematical thinking by describing the problem, discussing the
strategies they used to work on the problem (including those that did and did not lead to a
solution), providing a solution, and, finally, arguing that their solution was complete and valid. In
some cases these reports were produced as group projects. These reports provided opportunities
for further reflection and discussion.
The effect of the approach was that a classroom focusing on problem solving using a variety
of strategies, reflection on the process of problem solving, and engagement in the process of
exploration, conjecture, and argument can help prospective teacher develop mathematical beliefs
that are consistent with autonomous behavior. The community work on the problems made the
process less frustrating for the prospective teachers, allowed them to see the ways in which their
peers did mathematics, and showed them that problems could be solved in more than one way, i.e.,
a broadening in the acceptable methods of solving problems. The participants‟ beliefs became
more supportive of autonomous behaviors during the course.
(e) Problem-Solving Process (Ebby, 2000)
This study involves problem solving in an indirect way as part of a methods course in a teacher
education program that aimed to integrate fieldwork and coursework. One goal of the approach
studied that relates to problem solving is helping prospective teachers to draw on their experiences
as learners of problemsolving in developing a conception of their teaching role. Thus the approach
provided opportunities for them to learn about problem-solving and mathematics pedagogy.
The approach involved inviting the prospective teachers to participate in mathematical inquiry
through individual, cooperative, and whole-class problem solving. Each week the professor
assigned a non-routine problem-of-the-week and encouraged the prospective teachers to work
together on the solutions outside of class. In class, the professor asked for volunteers to share their
solutions and strategies with the rest of the class. These problem-solving activities were designed
to invite prospective teachers to learn mathematics in a non-threatening environment and to
reconsider what it means to learn, teach, and know mathematics. By focusing on solution process
over answer, the professor was also modeling an alternative pedagogy for them.
The effect of the approach based on the three participants studied was to provide them with a
different experience. One participant experienced what it was like to be an active agent in her own
mathematical learning through her engagement in problem solving. Another discovered that others
had a diversity of approaches to mathematical problems and that their understandings were often
different from hers. Another developed a new definition of the nature of mathematics as a result of
her engagement in mathematical problem solving in the methods course.
The Author’s Multi-goal Approach
This section deals with a study I have conducted that explicitly deals with problem solving as its
primary focus. It is an extension of the instructional approach reported in Chapman, 2005. It
incorporated what was learned then to further enhance the approach. Thus the goal of this follow-
up study was to investigate the extended approach to help prospective teachers to understand
problem solving as a mathematical and pedagogical process, i.e., as mathematical thinking and a
method of instruction, respectively. This is a multi-goal approach because it explicitly deals with:
development of understanding of problems, problem solver, problem-solving process, problem-
solving pedagogy and problem solving as inquiry-based teaching. What follows is an abbreviated
report on the study to highlight the nature and effect of the instructional approach.
163
Theoretical Perspective of Approach: The approach is framed in the work of Dewey (1933) with
a focus on inquiry, reflection and social interactions; cognitive guided instructions (Carpenter et
al., 1999) with a focus on understanding students‟ thinking and strategies to inform instruction;
and narrative inquiry (Polkinghorne, 1988), with a focus of using personal experiences to
understand self. An inquiry perspective framed in social constructivism formed a basis of the
learning activities used in this study.
Description of the Approach: The activities were organized in three stages: individual reflection,
inquiry activities and final reflections.
Individual-reflection: The first stage focused on individual self-reflection on problems and
problem solving in order to create awareness of each of the prospective teacher‟s initial
conceptions and knowledge. The participants were required to respond to a list of questions and
prompts in sequence that included: What is a problem? Choose a grade and make a mathematics
problem that would be a problem for those students. What did you think of to make the problem?
Why is it a problem? Is it a „good‟ math problem? Why? What process do you go through when
you solve a problem? If possible, represent the process with a flowchart.
Inquiry activities: The second stage consisted of inquiry activities intended to extend the
prospective teachers‟ initial conceptions and knowledge. The prospective teachers worked on all
problems in these activities without the facilitator‟s intervention. These activities included:
(1) Comparing different types of problems without solving them in order to explore the nature of
problems used in teaching mathematics and the goal of these problems in learning mathematics.
The prospective teachers were provided with a list of different categories of problems influenced
by Charles and Lester (1982). They were asked to compare and contrast the problems and to draw
conclusions about problems in learning/teaching mathematics.
(2) Writing and unpacking narratives of their experiences in order to examine the cognitive and
affective components of the behaviors involved in solving a non-routine problem. The prospective
teachers were required to write narratives of their experiences solving a problem that was assigned
to them. The narrative had to be a temporal account not only of the mental and physical activities
they engaged in to solve the problem, but the emotional aspects of the experience. They later
analyzed the narrative in terms of the affective aspect of the experience.
(3) Investigating others (e.g., peers and secondary school students) solving non-routine problems
to explore the thinking of others compared to their own. The prospective teachers were required to
solve an assigned problem and make notes of the thought process. They then worked in pairs, took
turns to observe each other solve the problem while thinking aloud, and compared their thought
processes. They selected a non-routine problem appropriate for a secondary school student, first
solved it, and then used it to observe the student solving it while thinking aloud. They also solved
a problem as a group in order to explore the collaborative and cooperative problem-solving
experience.
(4) Developing a model for non-routine problem solving, representing it as a flowchart, and
applying it to solving a non-routine problem in order to evaluate it.
Final reflection: The third stage included activities that required the prospective teachers to
engage in a final reflection by comparing their post-Stage 2 thinking with their pre-Stage 2
thinking; comparing their understanding of problems to theory (e.g., Charles & Lester, 1982;
NCTM, 1989); comparing their problem-solving models and flowcharts with those from theory
provided to them (e.g., Mason, Burton, & Stacey, 1982; Polya, 1954; Verschaffel, Greer, & de
Corte, 2000); relating their problem-solving models to an inquiry instructional model for teaching
secondary school mathematics; applying their knowledge to critique a current secondary school
mathematics textbook approved for use in the Province; and preparing a lesson plan based on their
inquiry instructional model.
164
Group reflection: Each of the three stages also required small group and whole-class interactions.
This included: participants sharing and comparing their individual reflections in Stages 1 and 3
and their findings from the inquiry activities in Stage 2; preparing a model of problem-solving or
inquiry-based teaching in small groups; and sharing and discussing small-groups' findings in a
whole-class setting.
Research Method: The participants were 29 preservice secondary mathematics teachers in the
second semester of their 2-year post-degree education program. This was their first course in
mathematics education, so they had no instruction or theory on problem solving prior to this
experience. They also were not taking any other mathematics education course in this semester.
The reflection and inquiry activities served both research and learning purposes. Thus data
consisted of copies of all of the participants‟ written work for all of the activities. There were also
field notes of their group discussions and whole-class discussions. The analysis began with open-
ended coding (Strauss & Corbin, 1998) of the data. The researcher and research assistants,
working independently, coded the data. Coding included identifying significant statements about
the participants‟ thinking of problems and problem solving and the changes in thinking resulting
from the activities. The coded information was categorized based on common themes and
frequency of occurrence to form the findings.
Findings of Effects of Approach: The approach was effective in expanding and deepening the
participants‟ understanding of problems, problem solving, problem-solving pedagogy and inquiry-
based teaching. Their thinking of problems shifted from predominantly traditional exercises or
word problems to an understanding of characteristics that constitute worthwhile mathematics
problems. Some of their descriptors of good problems are: they are of many different forms and
types, challenging, needs one to use deductive or inductive reasoning to come to a solution, have
one or more solutions and many approaches, interesting, not procedural and memorization of facts,
challenging but solvable for grade level, applicable to real life problem, often requires more than
one attempt to find a solution.
The participants‟ description of the problem-solving process was also enhanced, as reflected
in their flowcharts of it, which showed the need to move back and forth, as opposed to their initial
thinking of a linear path, to get to a solution. Following is a participant‟s problem-solving model.
Any Assumptions Understand the problem Represent what you know
that need to be made Reread, write in own words
Do the assumptions List what you know/is given
remove a barrier? What is being sought as
Try to do with and solution?
without barrier

Plan – establish relationships,

How does it fit patterns, other representations
with assumptions Follow through to see if it yields
made? an answer
Rethink/replan if doesn‟t answer
the question

Check solution/Reason it out

Does it make sense? Is it an answer?
Have you learned anything from the
plan/solution?
Can you apply what you learned
to other problems?

165
The theoretical problem-solving models the participants examined in Stage 3 of the approach
allowed them to validate, and in some cases refine parts of, their models, but not to change it,
showing preference for their experience as a basis of validating what was more meaningful to
them. Most of the participants viewed Polya‟s model to be closer to theirs and easier to follow and
use in their teaching. One explained, “It flows more smoothly, it is cyclic, and non-exclusive in its
processes. Although very generic in describing the steps, it reflects my cognitive steps in problem
solving. As well, the steps I take in developing a plan varies from problem to problem.” In
general, the inquiry activities of Stage 2 of the approach allowed the participants to construct
knowledge compatible with formal theory of problem solving. This allowed them to relate to the
theoretical approaches they examined in Stage 3 of the approach in a more meaningful way.
In terms of problem-solving pedagogy, the thinking of the prospective teachers shifted to a
student-centered approach. For example, they explained that they will have students work on a
problem first, then share and discuss it in a whole-class discussion, or will have a whole-class
discussion first to understand the problem, have students work in groups or individually, then
share and discuss. They will pose questions or prompt the students when they (students) are stuck
or in response to their (teachers) questions. They would use their problem-solving model to help to
frame prompts and guide the whole-class discussions. This approach they viewed as being similar
to inquiry-based teaching where they will start with an applied situation that embodied the
mathematics concept being taught and allow students to unpack it with their (the teacher) guidance
to understand the concept.
Overall, the study suggests that this three-stage approach framed in a social constructivist
perspective and providing inquiry-based and self-reflective opportunities involving non-routine
problem solving, followed by comparison to theory can help to deepen prospective secondary
teachers‟ understanding and knowledge of problem solving in their teaching. However, whether
they are able to enact this knowledge in their teaching has not been researched as yet. But holding
such knowledge seems like an important first step in getting there.
Implications for Teacher Education
Based on the literature review and my study in the preceding sections of this paper, following are
eight key characteristics of instructional practices identified as important to form a basis of
preservice teachers‟ learning about problem solving. (1) Exploring others as problem solver: e.g.,
the prospective teacher works with a child or peer to observe, interview, and document
information about this child or peer as a problem solver. (2) Exploring self as problem solver: i.e.,
inquiring into one‟s thinking, learning and instructional practices and developing ability to
monitor and to control one‟s activities when solving problems. (3) Exploring nature/structure of
problems. (4) Solving challenging problems individually and in small groups without external
assistance, e.g., to develop awareness of strategies and skills for solving problems. (5) Posing
problems. (6) Comparing self with others, e.g., peers, students, theorists. (7) Formulating
instructional model for problem solving. (8) Exploring self as facilitator of problem solving, i.e.,
to develop an understanding of the teacher‟s role in facilitating students‟ problem solving.
If teachers are to hold knowledge to help them to teach about and through problem solving,
they should be provided with experiences not only in solving problems, but also with all of these
eight characteristics. It may also be necessary for these characteristics to be embodied in an
integrated experience as opposed to being in isolation.
References
Arbaugh, F. & Brown, C.A. (2004). What makes a mathematical task worthwhile? Designing a learning tool for high
school mathematics teachers. In R. R. Rubenstein & G. W. Bright (Eds.), Perspectives on the teaching of
mathematics (pp. 27-41). Reston, VA: National Council of Teachers of Mathematics.

166
Carpenter, T., Fennema, E., Franke, M., Levi, L., & Empson, S. (1999). Children’s mathematics: Cognitively Guided
Instruction. Portsmouth, NH: Heinemann.
Charles, R. & Lester, F. (1982). Teaching problem solving: What why & How. Palo Alto, CA: Dale Seymour
Publications.
Chapman, O. (2005). Constructing pedagogical knowledge of problems solving: Preservice mathematics teachers. In
H. L. Chick & J. L. Vincent (Eds.), Proceedings of the 29th PME International Conference, 2, 225-232.
Contreras, J.N. & Martínez-Cruz, A.M. (2001). An investigation of preservice elementary teachers' solution processes
to problematic story problems, In Van den Heuvel-Panhuizen, M. (Ed.), Proceedings of the 25th PME
International Conference, 2, 289-296.
Contreras, J.N. & Martinez-Cruz, A.M. (2003). Preservice elementary teachers‟ solution processes to problematic
addition and subtraction word problems involving ordinal numbers and their interpretations of solutions. In N. A.
Pateman, B. J. Dougherty & J. T. Zillox (Eds.), Proceedings of The 27th Conference of the International Group
for the Psychology of Mathematics Education. Honolulu: University of Hawaii, 2, 197-204.
Dewey, J. (1933). How we think. Boston: Heath.
Ebby, C.B. (2000). Learning to teach mathematics differently: The interaction between coursework and fieldwork for
preservice teachers. Journal of Mathematics Teacher Education, 3(1), 69-97.
Kilpatrick, J., Swafford, J., & Findell, B. (Eds.). (2001). Adding it up: Helping children learn mathematics.
Washington, DC: National Academy Press.
Lee, H. S. (2005). Facilitating students‟ problem solving in a technological context: prospective teachers‟ learning
trajectory. Journal of Mathematics Teacher Education, 8:223–254.
Leung, S.S.(1994) On analysing problem-posing processes: A study of prospective elementary teachers differing in
mathematics knowledge, In Ponte, J.P. & Matos, J.F. (Eds.), Proceedings of the 18th PME International
Conference 3, 168-175.
Mason, J., Burton, L., & Stacey, K. (1982). Thinking mathematically. New York: Addison-Wesley
National Council of Teachers of Mathematics (2000). Principles and standards for school mathematics. Reston, VA:
Author.
National Council of Teachers for Mathematics (1991). Professional standards for teaching mathematics. Reston, VA:
Author.
National Council of Teachers of Mathematics. (1989). Curriculum and evaluation standards for school mathematics.
Reston, VA: Author.
Polkinghorne, D. (1988). Narrative knowing and the human science. Albany: State University of New York Press.
Polya, G. (1954). How to solve it. New York: Anchor Books.
Roddick, C., Becker, J. R., & Pence, B. J. (2000). Capstone courses in problem solving for prospective secondary
teachers: Effects of beliefs and teaching practices, Proceedings of the 24th PME Conference, Vol. 4, 97-104.
Schmidt, S. & Bednarz, N. (1995). The gap between arithmetical and algebraic reasonings in problem-solving among
pre-service teachers, In Meira, L. & Carraher, D. (Eds.), Proceedings of the 19th PME International Conference,
2, 82-89.
Schoenfeld, A.H. (1985). Mathematical problem solving. Orlando: Academic Press.
Simon, M.A. (1990). Prospective elementary teachers‟ knowledge of division. In Booker, G.; Cobb, P.; Mendicuti, T.
(Eds.), Proceedings of the 14th PME International Conference, 3, 313-320.
Stein, M.K., Smith, M.S., Henningsen, M.A., & Silver, E.A. (2000). Implementing standard-based mathematics
instruction: a casebook for professional development. New York: Teachers College Press.
Strauss, A. & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded
theory. Thousand Oaks, California: Sage.
Szydlik, J.E., Szydlik, S.D., & Benson, S. R. (2003). Exploring changes in preservice elementary teachers‟
mathematics beliefs. Journal of Mathematics Teacher Education, 6(3), 253-279.
Taplin, M. (1996). Pre-service teachers' problem solving strategies, In Puig, L. & Gutiérrez, A. (Eds.), Proceedings of
the 20th PME International Conference, 4, 299-306.
Van Dooren, W., Verschaffel, L., & Onghena, P. (2003). Preservice teachers‟ preferred strategies for solving
Arithmetic and Algebra word problems. Journal of Mathematics Teacher Education, 6(1), 27-52.
Verschaffel, L., Greer, B. & de Corte, E. (2000). Making sense of word problems. Lisse, The Netherlands: Swets and
Zietlinger Publishers.
Verschaffel, L., De Corte, E. & Borghart, I. (1996). Pre-service teachers' conceptions and beliefs about the role of
real-world knowledge in arithmetic word problem solving, In Puig, L. & Gutiérrez, A. (Eds.), Proceedings of the
20th PME International Conference, 4, 387-394.

167