Educ Stud Math (2013) 83:103–116

DOI 10.1007/s10649-012-9436-4

Teachers implementing mathematical problem posing

in the classroom: challenges and strategies

Shuk-kwan S. Leung

Published online: 14 October 2012

# Springer Science+Business Media B.V. 2012

Abstract This paper reports a study about how a teacher educator shared knowledge with
teachers when they worked together to implement mathematical problem posing (MPP) in
the classroom. It includes feasible methods for getting practitioners to use research-based
tasks aligned to the curriculum in order to encourage children to pose mathematical problems.
Techniques, challenges, and strategies of implementing an MPP focus were also reported.

Keywords Mathematical problem posing . Enacting research-based tasks .

Teachers as co-investigators . Implementation . Elementary math curriculum

1 Introduction

Effective mathematical problem solving is a primary goal for mathematics education. At the
heart of problem solving often lies a great problem. When mathematicians meet, they share
how they find problems, pose new ones, and reformulate novel problems from known
problems. The problem-posing aspect of this process has been viewed as important (Polya,
1954; Freudenthal, 1973; Brown & Walter, 1983; Silver, 1994) and emphasized in curriculum
standards and instruction (NCTM, 2000; Ministry of Education of Taiwan, 2003). However in
practice, the attention paid to problem posing has not been proportionate to this importance.
Since research already exists that indicates children’s success in mathematics learning after
problem-posing instruction (Ellerton, 1986; English, Fox, & Watters, 2005), a natural next step
is to introduce problem posing to practitioners. However, a survey in Taiwan indicated that

The research reported was part of a larger research project funded by the National Science Council of Taiwan
ROC. Opinions in this report are those of the author and not representing those from the funding agency. The
author wishes to thank the editor of Educational Studies in Mathematics and the guest editors of this special
issue for the time and effort in the whole reviewing process; also the few anonymous reviewers for
constructive comments on earlier versions of this manuscript; and finally, the schoolteachers for participation
in this research.
S.-k. S. Leung (*)
National Sun Yat-sen University, Kaohsiung, Taiwan, Republic of China
e-mail: [email protected]
104 S.-k.S. Leung

children barely ever posed problems, teachers were inexperienced in posing activities, and such
activities were difficult to implement (Leung, 1994a, b). In addition, teachers expressed a need
for skills to enact tasks in class and specific ideas for handling posed problems from children.
Without this knowhow, high-level tasks would not function in the way intended (Stein, Smith,
Henningsen, & Silver, 2009).
The primary purpose of this paper is a close examination of teachers’ experiences in
implementing problem posing by using tasks and a coding method that I provide. The
research question is, “Why and how do teachers enact problem-posing task materials in an
elementary mathematics curriculum?” In addition to studying the reasons for and natures of
teachers’ actions, I also study how teachers enact research-based mathematical problem-
posing (MPP) tasks, how teachers classify and use children’s work in teaching, and what
instructional strategies teachers use when issues arise in teaching. In my attempt to answer
these questions, I also attend to a neglected component of research—the professional
development of the teacher educator and teachers (Jaworski, 2008). I explore, analyze,
and discuss my own learning and that of teachers by recruiting active teachers to connect
research to practice.

2 Literature review

In this literature review, I include mathematical problem posing and teachers’ participation
in curriculum reform.

2.1 Mathematical problem posing

In this study, I refer to problem posing as the formulation of new problems or the
reformulation of problems into novel problems (Silver, 1994) where the problem poser is
a provider of information (Simon, 1973) and creates problems by attending to problem
structure (English, 2003). Leung (1994a, b) discussed four characteristics of this kind of
problem posing. First, MPP can be idiosyncratic. When one is considering some givens and
poses a problem, one is trying to connect the various givens to a goal. For example, suppose
the given information in a problem is, “There are 10 boys and 20 girls in a class.” A person
may pose, “How many children are there altogether?” but another person may pose, “What
is the ratio of boys to girls?” Second, the act of problem posing involves plausible reasoning
(e.g., “Consider the change in this ratio if the total number of children is the same but the
number of boys is 11 instead?”). Third, problem posing can happen before, during, or after
problem solving. For example, after posing and solving for the total number of children, one
may ask, “What is the percentage of boys in the class?” Fourth, a posed problem that is not
yet solved can be insufficiently specified or even impossible, such as “If 25 bottles were
distributed to 10 boys and 20 girls and each one gets one, how many bottles were left?”
In this review, problem posing is viewed as related to problem solving. There are four
phases in problem solving (Polya, 1945): understanding the problem, devising a plan,
carrying out the plan, and looking back. These aspects can be seen in Fig. 1, with the
clockwise arrows representing the simplest case.
However, problem solving means more than a straightforward procedure of passing
through these four phases. Going back and forth is common and each switch to a different
phase signals a person’s evaluation of what he or she did in the current phase. When a step is
not successful the solver will backtrack, which is represented by the counter-clockwise
arrows in Fig. 1. If a person is solving his or her own problem, rather than a problem that has
Teachers implementing mathematical problem posing in classroom 105

Fig. 1 Four phases in problem Understand

posing and problem solving
(Pose)

Look Back Plan

Carry Out

been given, the initial Understand phase is considered a Pose phase. Subsequently, problem
posing can occur at many points too, before or after solving (Leung, 2009). The decisions
and actions of posing and solving can be inter-related.
In the studies of Kilpatrick (1978) on problem solving, he classified research on problem
solving by task and subject variables. Studies on problem posing can also be classified into
task and subject variables. One example of a task variable refers to mathematics content and
format, such as writing up a word problem on multiplication and division (Greer & McCann,
1991) or making a difficult and an easy problem on percent (Van den Heuvel-Panhuizen,
Middleton, & Streefland, 1995). Tsubota (1987) reported six types of problem posing tasks
used successfully after systematic implementation. These six types contain one of the
following: an algorithm, text, a figure/table, a math topic, an answer, or a mathematics
problem. Tsubota did not report on personal reflections and solutions to difficulties in
designing posing tasks. For instance, a crosstab on content and format is a feasible way to
establish MPP tasks.
Subject variables (Kilpatrick, 1978) include gender, ability, cultural groups, and grade
level. In practice, the grade-level variable is often attended to by classroom teachers.
Therefore, it is important to tell teachers which tasks to use for which grade, both for
alignment to curricular goals in teaching a piece of mathematics content and for teachers to
implement problem posing.

2.2 Teachers’ participation in curriculum reform

In this review, we use the term “teachers” broadly: “[Since] the teacher educator can be
regarded as the teacher of prospective or practicing teachers… it makes sense to regard
student teachers, teachers, and teacher educators as teachers.” (Krainer, 2008, p. 1) In the
third volume of the International Handbook of Mathematics Teacher Education (Krainer &
Wood, 2008), participants in mathematics teacher education are viewed as individuals,
teams, communities, and networks, and very often the growth of teachers is associated with
the curriculum reform climate.
In the mathematics curriculum reform occurring in Taiwan, teachers are facing unprec-
edented challenges to change the way they teach, including incorporating problem solving
and posing by children. A call for instructional change often involves teachers needing to set
up different types of tasks and to classify students’ work according to rubrics. In this regard,
sample tasks and children’s work are helpful to communicate to teachers the necessary
knowhow for instructional change (see Parke, Lane, Silver, & Magone, 2003). Therefore,
106 S.-k.S. Leung

one might expect that when asked to introduce problem posing in their instruction, teachers will
need similar research-based resources.
Under curriculum reform, teacher educators have brought interventions to classrooms,
have studied the growth of teachers engaging in action research (Kemmis, 1991), and have
interviewed teachers about how they teach a topic, as in Ma (1999). However, teachers often
find programs offered by teacher educators too theoretical to be applied to their own settings
(Hastie, 1992; Sowder, 2008). As a way to make programs more practical, teacher educators
should show teachers how tasks can be enacted. Additionally, both parties can work together
to solve problems arising in the new teaching methods, thus acting as co-learners (see
Jaworski, 1999). For instance, in a study on enacting proof-related tasks (Bieda, 2010),
the co-learners were teacher educators and veteran teachers. After a partnership was estab-
lished between the two groups, both teachers and teacher educators analyzed curriculum and
classified and selected problems in the implementation of the new teaching methods.
Jaworski (2008) reflected on her work as editor of the Journal of Mathematics Teacher
Education (JMTE) and commented that JMTE’s work mostly falls into the category of a
mathematics teacher educator (MTE) passing knowledge to teachers. She drew a Venn
diagram to represent how working together can share knowledge. One circle carries MTEs’
knowledge of research and theory while the other circle holds teachers’ knowledge of
students and schools. The sharing of knowledge is represented by the intersection of the
two circles, which means that the passing of knowledge is bidirectional. That is to say,
teachers also pass knowledge to MTEs.
In a comprehensive chapter on teachers as researchers, Henson (1996) discussed teacher
educators working with teachers and included three levels of teachers’ involvement: teacher
as helper (level 1), teacher as junior partner (level 2), and teacher as lone researcher or
collaborator (equal partner, level 3). In most cases, teachers’ level of involvement is at level
1—providing a classroom or students to be studied by an outsider (the teacher educator).
Very few teachers are able to conduct action research—identifying the problem, conducting
the study, and using data and findings. It is reasonable to suggest that research at involvement
level 2 is a way to prepare teachers to become lone researchers. A study on tracing the paths of
teachers’ increase in involvement levels is needed.
Based on the literature reviewed above, it seems that my being a teacher educator can
start by developing research-based MPP tasks, then introducing teachers to consider using
those tasks to teach particular math content listed in the curriculum standards document for
grades1 through 6. I take the orientation that teacher educators and teachers are co-learners,
and that learning is a process of changing one’s participation in a particular community
(Lave & Wenger, 1991; Wenger, 1998).

3 Method

3.1 Case background

This study is a case study (Yin, 1994) on an intervention and the context in which it
occurred. The case involved a teacher educator and two research assistants (the research
team); together with 60 elementary schoolteachers in the first term of year 2 of a 3-year
project. The teacher educator is the investigator, who will be referred as “I” in this paper. In
that project, the teacher educator recruited teachers to participate from 1 to 3 years. The first
year of this study involved the development of an MPP task inventory by grade, content
strand, and type according to Tsubota (1987); 52 tasks were the result. The case is situated in
Teachers implementing mathematical problem posing in classroom 107

a district in the south of Taiwan, where 105 elementary schoolteachers assisted me in collecting
children’s responses and piloted tasks. In the second year of the project, 60 selected teachers
participated and used the tasks to teach mathematics while I acted as a facilitator who described
and coded the implementation results of teachers enacting the tasks. There were 10 teachers
from each grade (grades1 through 6).
In the final year, three teachers conducted action research and continued their journeys
(reported separately). Note that the number of teachers participating in this phase was
controlled by me. I chose to allow only a smaller number of teachers as the level of
involvement (Henson, 1996) increased over time.
In Taiwan, every Wednesday afternoon, elementary schoolteachers do not have class; the
time is reserved for attending in-service training and accumulating credit hours. In this
project, this series of Wednesday teacher seminars were split into two sections for each term.
The first section was conducted for provision of tasks and coding methods before the
implementation period (seminar1). The second section was used to share results of imple-
mentation (seminar2). The implementation period occurred between the two seminars. As
issues arose that required attention, teachers communicated with peers or the teacher
educator through emails, phone calls, or by additional meetings in small groups.
Three sessions of seminar1 were conducted on the first three Wednesdays of the term
with 20 teachers each from the lower (grades1–2), middle (grades3–4), and upper grades
(grades5–6). The first feature of seminar1 was for teachers to experience MPP in pairs. They
acted as children while I acted as an elementary schoolteacher. These “elementary school
children” posed problems and when they did not understand, they asked questions and the
“teacher” showed how to answer these questions. Frequently asked questions were recorded
to compare with questions asked by children during the implementation period. The second
feature of seminar1 was categorizing children’s given problems. This exercise in coding
consisted of problems posed by teachers during the seminar as well as those posed by
children during the first year of the study. I introduced a categorization scheme developed by
Leung and Silver (1997) for this coding component.
Children’s posed problems were classified into five categories. The codes were: 10not a
problem, 20non-Math, 30impossible, 40insufficient, 50sufficient or extraneous. Figure 2
below includes one real example of children’s given problems in each category. The scale
was categorical, not ratio; a problem coded as “4” is not considered twice as good as a
problem coded as “2”.
As seen in Fig. 2, the first type is not-a-problem; such “problems” are actually only a
description or a phrase. The second example, “Who ate the cake?”, is a problem, but not a
mathematical problem. The third is an impossible math problem; no answer can be found
using the given information; even when more information is supplied. The fourth category,
insufficient, is different from the third in that insufficient problems can be solved if missing
information is added. Finally, the fifth category is sufficient (this includes those problems
with extraneous information).
The implementation period started right after seminar1. Teachers were free to decide on
when to use which task and how to use the teaching materials they received at the first
seminar. They acted as my junior partners, taking on involvement at level2 (Henson, 1996).
During the implementation period, teachers collected examples of children’s work in each
category and used the provided self-addressed envelopes to share results. At the end of the
term, teachers attended seminar2, where they reviewed examples of children’s work and
discussed how they used children’s own problems to teach. Common difficulties and
successes were also reported. All teachers who attended seminar1 returned information
about actual implementation (tasks used, grade level, children’s scripts). Due to time
108 S.-k.S. Leung

35. A piece of cake is cut into 8 equal parts. Wah ate 4/8 piece, Ming ate 2/8 piece,
Kong ate 2/8 piece.

Codes/Description Examples
1. Not a Problem A can of soda is poured into 8 equal cups, Chin drank
4/8 litres, Lip drank 2/8 litres, Jay drank 2/8 litres.

<This is a description, not a problem>

2. Non Math Who ate the cake?

<This is not a mathematics problem.>

3. Impossible A watermelon is cut into 10 parts. Ming ate 4/8, Ying
ate 2/8, Wah ate 3/8, how much is left?

<The problem is impossible as there’s no left over.>

4. Insufficient A piece of cake is cut into 8 equal parts. Wah ate 4/8 of
the piece of cake, Ming at 2/8 of the piece of cake.
How much of the piece of cake does Kong eat?

<This mathematics problem has insufficient information to provide an answer.>

5. Sufficient After my two sisters ate 2/4 of the piece of cake and
1/4 of the piece of cake, how much can my brother eat?

Fig. 2 Codes, descriptions, and examples of children’s work

conflicts that 14 teachers had with end-of-term routines, a total of 46 teachers attended
seminar2 and completed questionnaires.

3.2 Data sources

Various types of data from the seminars above were collected in the following ways:

Teacher educator’s reflections (TER) I kept a journal throughout the implementation. At

each of the meetings, I recorded the types of problems that teachers posed as well as the
Teachers implementing mathematical problem posing in classroom 109

types of questions that teachers asked as they attempted to categorize posed problems.
During seminar2, I collected data pertaining to tasks used by teachers, selected children’s posed
problems, and problems that teachers encountered as well as how they solved these problems. I
also recorded the emails or phone contacts of individual teachers or small groups.

Children’s scripts (CS) I received by post the children’s scripts sent by 60 teachers after the
implementation period. A script contained a collection of children’s work pasted onto a single
sheet of paper (A4 size, vertical; see Fig. 2). The paper format was specified in seminar1. The top
of the script contained the item itself. Below the item were five horizontal blank boxes marked “1”
to “5”. They were designed for teachers to collect sample work from each of the five categories.
Children wrote at least one posed problem to each given item. The teacher categorized problems
as “1” through “5”. When all posed problems to each item from the whole class were coded, the
teacher selected and pasted at least one representative work onto the script for that item. The
research team double checked the results of the categorization after receiving the CS in the mail.

Teachers’ questionnaires (TQ) Finally, during seminar2, we collected questionnaire data

that examined how teachers and children felt and performed when MPP was implemented.
We also asked if they were interested in doing action research in the coming year. The
questionnaire consisted of the following probes:

1. What type of tasks will teachers consider in MPP?

When deciding on a task, what concern(s) do you have?
In implementing the task, how did you get your children to do MPP?

2. What types of problems are produced by children? Can teachers categorize them?
Is there any difficulty in using the categorization scheme?
Is there any recommendation for revising the scheme?
3. What are the successes and difficulties of teachers during implementation?
Did your children understand the task? Did they ask questions? What did they ask?
Please describe the feelings of the children. Please describe how the teacher felt
when seeing children pose problems. Write about any fun or pain about imple-
menting posing problems.
4. Are you interested in doing action research in the next school year?

3.3 Data analysis

Each of the methods for collecting data was analyzed separately, using either quantitative or
qualitative analyses. For children’s scripts, teachers coded responses according to the
problem types they practiced during seminar1. The investigator and graduate research
assistants double checked that children’s work on the scripts was placed by the 60 teachers
into the correct categories (Leung & Silver, 1997). Descriptive statistics showed counts for
each of the five categories from tasks 1 to 52. I began with a particular analytic framework
with expected categories of responses (Leung & Silver, 1997) and examined how teachers
used the framework. The teachers’ journals and teacher questionnaires were analyzed
qualitatively, using the methods described in Creswell (2009); using a more inductive,
grounded-theory kind of approach.
110 S.-k.S. Leung

4 Results and discussion

In this study, teachers were actively involved in research when they chose to do problem
posing with the tasks that they picked, coded children’s work, and used results of posing in
teaching. They were no longer helpers in data collection (level 1) but they were also not yet
action researchers (level 3). Therefore, teachers were at involvement level2 (Henson, 1996),
where both teacher educators and teachers function as learners in a mutually reciprocal
influence of knowledge (Jaworski, 2008); where both are teachers in a broad sense as
described by Krainer and Wood (2008).
Instead of teacher educators designing programs to influence the methods of classroom
teachers, this study indicted that teacher educators can work and learn together with teachers.
Below, I report and discuss what I learned from walking through the implementation journey
with a group of teachers. I used the three data sources (TER, CS, and TQ) to improve my
knowledge of how my teachers (a) considered and used tasks, (b) used the coded children’s
work and results in teaching, and (c) responded to matters arising in teaching. In each case, I
used at least two of the three data sources to inform my understanding.
What did I learn from working closely with teachers? I consider why and how teachers
enact research-based tasks, how teachers use the coding scheme to analyze children’s posed
problems, and the techniques, challenges, and strategies they employed in enacting MPP.

4.1 Enactment of research-based tasks

MTE shared knowledge of research-based tasks As mentioned in detail above, seminar1

consisted of a briefing on problem posing tasks. Teachers were divided into groups of two
and were asked to pose problems as elementary school children. I purposely asked teachers
to watch how I responded to questions and led the discussion when the teachers orally
presented the problems they posed. When the learners asked for a problem-posing example,
I did not give an example but said, “No, no. Try to think it over. If I give you an example,
you will pose a problem similar to my example!” All the problems posed by teachers during
seminar1 belonged to the sufficient category.

Teachers shared knowledge from enacting MPP tasks After seminar1, I made notes on how
teachers posed problems, raised questions, or gave suggestions. In my reflections on
curriculum integration, I realized that the teachers were concerned with the alignment of
the tasks to their curriculum. They also wondered if they could prompt children to pose
problems that were suitable for specific topics (e.g., “What if children pose a problem that
does not belong to that chapter?”, “What if the problems are too difficult for them to
solve?”). The teachers also asked questions about the format of students’ answers to the
tasks (e.g., “Can children draw or read out loud instead of writing?”, “Do I need to give
examples?”). Finally, teachers made suggestions on the wording and format of the tasks
(e.g., “Algorithm” is a difficult word for lower grades. Question number “30.” and the
beginning of item “15” should be separated to avoid children reading it as a decimal
“30.15”).
Lastly, I analyzed the teachers’ questionnaire responses by focusing on the reasons why
these schoolteachers considered tasks and how they used each task. From the analysis, I
learned that when teachers considered a task, they would check the difficulty level (e.g., “the
item with a picture of few eggs and two broken is great for grade1; children automatically
made up interesting math problems after getting this item”), the children’s familiarity with
the task (“the item showing TV program and time is one I chose because children are
Teachers implementing mathematical problem posing in classroom 111

familiar with TV timetables.”), and if the mathematical topic was appropriate (“I
chose the Egg item because children were learning Number composition and decom-
position on that day”).
I also learned more about how the teachers set up tasks. Out of 46 teachers, 38 asked
children to handwrite the problems and two others asked the students to discuss in groups
before orally reporting the problems. One teacher even asked children to rate the task or
problems that their peers made. This teacher reported that the children loved difficult
problems or problems related to real-life situations.

4.2 Analyzing children’s work using the coding method

MTE shared knowledge on the coding method In seminar1, the second part of the session
focused on coding student work samples. In order to facilitate the transition of roles from
experiencing posing as children to categorizing children’s problems as teachers, I an-
nounced, “Let us read children’s posed problems. Remember that you are teachers now,
not children.” During the categorization exercise, the teachers were amazed to see real
examples of children’s problems in the five different categories. Teachers asked me to clarify
the distinction between the not a problem and Non-Math categories as well as the distinction
between Impossible and Insufficient.

Teachers’ reports on coding children’s work After seminar1, teachers introduced problem
posing in their classrooms whenever they found it appropriate. Though teachers worked
independently and chose their own tasks, they could discuss with colleagues in the same
school. At times, they called me and asked questions like “Can I use grade five tasks for
grade six?” Some of them collected children’s diaries on an irregular basis and kept journals
themselves. When students posed problems that fell under the Not a Problem or Non-Math
categories, teachers directed children’s attention to the instructions and asked them to “make
up a problem” instead of using the word “pose.”
After students posed their own problems, teachers used the children’s posed problems to
teach in the classroom. They invited children to be teachers and to fix the sample problems
that were in the other four categories. They found that this type of discussion made other
students more aware of the structure needed in order for a problem to be considered
sufficient. One teacher noted the success of these discussions in her journal when she wrote,
“After we do problem posing, children spot out problematic items on a test and do not even
solve them” (TJ).

Sharing knowledge by reviewing children’s scripts The final data analysis was conducted on
the work given by children. Teachers returned a total of 2,204 children’s posed problems.
Three findings from the analysis of these scripts are of note. First, all 52 tasks were
implemented; there was no task that the 60 teachers did not attempt to use in instruction.
Second, 84 % of all problems posed by children were plausible mathematics problems
(75 % with sufficient information, 9 % without). When the teacher educator was
working closely with teachers, having them experience problem posing before enact-
ing such tasks, children generated a high percentage of plausible mathematics problems. In
contrast, when researchers have presented problem posing tasks directly to children, the
finding was that “children could not pose reasonable math problems” (see Leung,
2009). Third, 11 tasks contained children’s work in all five categories. For the
remaining 41 tasks, there are missing examples in at least one of the five categories. In
future training sessions, the researcher can use this result to display actual examples
112 S.-k.S. Leung

(esp. to the 11 tasks) in order to explain the scheme of five different types of children’s
posed problems.
In seminar2, teachers attended small group meetings to discuss children’s work and
noticed that the posed problems revealed the children’s understanding or misunderstanding
of mathematical concepts. For example, in one item there is a picture of a mug of hot liquid
next to a thermometer showing 49 °C. A teacher reported what her children’s responses to
this item revealed. The sufficient math problem (category 5), “Mom heated the tea to 5°
higher, what is the final temperature?” indicated a second grader’s consideration of
addition of numbers up to two digits (part of the grade2 curriculum). The problem,
“How much tea will 7 mugs contain altogether?” was an insufficient problem (cate-
gory 4) as the capacity of the mug is missing. In category 3, the teacher’s selected
children’s posed problem was, “If my brother had a fever, and his temperature in the
morning was 25° and also 25 in the afternoon, what is his total temperature for the day?”
This is an impossible problem for two reasons. First, the temperature, 25°, is too low
for human beings. Second, the adding of the two temperatures does not make sense as
an indicator of the total seriousness of the fever for the whole day. For category 2 (Non-
Math), the example was, “Do you like soup with steam?” Finally, the example for not-a-
problem was, “The temperature is 49°, the capacity of the mug is 4 kg”. It belongs to not-a-
problem category.
In addition to learning about students’ understandings and misunderstandings, the exer-
cise of classifying responses enabled the teacher educator to see if teachers understood the
categorization scheme. In the above item, one teacher regarded “What is the price for a mug
of coffee?” as impossible while another teacher considered it as a problem with insufficient
data. In reading teachers’ reports such as this one, the teacher educator learned that although
the categorization scheme was explained, there were still problems with the distinction
between Impossible and Insufficient.
I also explored teachers’ comments from the questionnaire on classifying children’s
responses. A total of 36 teachers were able to use the scheme, and 20 said that they
needed to ask colleagues for help. Concerning the coding scheme, 17 found it hard to
use; nine said, “No opinion;” and 20 said, “Great, no problem.” Those who found the
coding scheme hard to use needed more examples to understand the categories of
Non-Math, Impossible, and Insufficient. Teachers also suggested rejecting or demoting
problems that did not match the given situations and problems that were not realistic
(e.g., problems where a boy’s age is greater than his father’s age or the height of a
grade1 child is 23 cm). Some teachers gave extra credit for wonderful problems.
Finally, one teacher wrote, “I prefer only two categories: a problem or not a
problem.” Indeed, this simple coding system is user friendly and may be best for teachers
at an early stage of attempting to integrate MPP into instruction. However, the more complex
coding scale (five categories) is useful for teachers who are serious about adequately integrating
MPP into their teaching.

4.3 Responding to matters arising during teaching MPP

My motivation for an implementation program was to combine teachers’ needs (e.g., tasks,
coding) with my solution of providing examples. Naturally, issues concerning the imple-
mentation program did not arise until we began working together since it is difficult to
combine an educator’s view of theory and research results with teachers’ knowledge of
students and schools (Jaworski, 2008). Five issues that arose as we worked together are
described below in detail.
Teachers implementing mathematical problem posing in classroom 113

First, teachers discussed the challenge they came across when children asked questions
about problem posing. It is interesting to note that the questions that children asked teachers
during the implementation period were similar to the questions the teachers asked the
investigator during seminar1 (e.g., What problems do you expect me to pose? Do I need
to solve them? How many shall I pose? Can you give me examples?). Teachers liked to
discuss how they dealt with such questions in their classrooms.
Second, teachers were also concerned with children’s ability to understand the process of
MPP since it was their new way of learning. The group of 60 teachers was given the opportunity
to reflect upon their practice and report success or failure when completing questionnaires in
seminar2. At the beginning of the implementation, children did not understand what they were
supposed to do. Only 16 teachers wrote “my children understand,” whereas others expressed
that the children were having a hard time. Students’ understanding was assessed by the
statements they made and the types of questions that they asked during the implementation,
such as: Should I use all information? Do I need to supply an answer? Can I pose this problem?
Is this problem a wrong problem to pose? What shall I do if I cannot pose? I do not
know what I am supposed to pose. I have not done a similar task like this before.
During seminar2, teachers reported that students’ frustrations had subsided. Children
became used to the activity and loved it.
Third, in addition to the above concern on understanding of MPP, teachers were also
attending to children’s feelings toward MPP. Since there were six participating teachers in
each school, a pair for each of the grade levels, teachers were free to talk about children’s
feelings and also how they felt in teaching MPP. More than half of the teachers replied that
their children were positive towards MPP (e.g., “Very excited with being creative.” “There’s
a sense of achievement.” “Children loved to share problems in class.”). Ten teachers reported
mixed feelings from their classes (e.g., students were excited about MPP but it was still painful
during implementation; children were excited that they were asked to pose, but they felt pained
if they could not pose a legitimate problem according to the given information; students who
were able to pose problems easily were excited whereas their less successful counter-
parts expressed frustration). Finally, seven teachers conveyed negative reactions (e.g.,
children disliked doing it again and again, not feasible for grade1, exam-oriented
children asked how many points they get from posing one problem).
Fourth, I captured teachers’ feelings, 33 were positive towards this implementation. Some
of the positive comments made by teachers are listed here:

& I feel easy when I go to class

& I look forward to the day when I can do posing
& I do not differentiate high and low achieving children… in posing
& All can pose problems using their own perspectives
& Unlike problem solving, children do problem posing without pressure
& I am touched by what they can do in posing
& I realized that some of my children have hidden potential
& Now I see their levels and can adjust my teaching to give appropriate challenge at their
levels
& MPP makes my lesson open and free
& The boys and girls are energetic and are in good moods.

Other teachers had more negative feelings towards MPP. These teachers conveyed feel-
ings of frustration when children could not pose problems. Some also found difficulty in
implementing MPP when the class was heterogeneous in abilities.
114 S.-k.S. Leung

Fifth, the MTE made sure that teachers’ effort were recognized and shared in group
settings. At the end of the implementation, 46 teachers met at seminar2. A glossary of real
examples or responses to the 52 tasks was distributed to all teachers at the seminar. As
teachers turned the pages, read, and pointed to examples they said, “This is the one that I sent
back.” Sharing the results of the teachers’ work proved to be a feasible way to empower
teachers.

5 Conclusion

In this study, I explored how 60 teachers who were participating in research at level1
the preceding year could work at level2, and prepared them to participate as action
researchers in year3. I adopted the three involvement levels of Henson (1996) in
research design. Given that the three parts of this MPP research project were devel-
oping research-based tasks [I], enacting tasks [II], and classifying children’s work and
using results in teaching [III], then the three levels of teacher participation in research
map onto the three parts as follows:

Level 1: Teachers assist in developing task by data collection [Helpers; I only].

Level 2: Teachers decide on when to use which tasks, use coding scheme then suggest
how to revise task and coding. They also suggest ways to use children’s work in
teaching [Junior partners; II, III only].
Level 3: Teachers conduct action research [Equal partners; I, II, III].

In walking through this journey with teachers, I found that a feasible initial step
for the implementation of MPP is the use of tasks to integrate with the curriculum.
The implementation of a set of tasks based on 24 categories obtained by pairing four
curriculum strands with six problem-posing types (Tsubota, 1987) in a program for
teachers can provide important data for developing a book of MPP tasks. This book
of tasks, with real sample work from children for the five possible posed problem
categories (see Leung & Silver, 1997) and a section of tips for teachers who try
problem posing in teaching, will allow more depth for teacher training sessions and
research on teacher development. An example can be found in Parke et al. (2003)
concerning tasks and scoring rubrics for classroom teachers’ instruction and assess-
ment. The findings from this study regarding challenges and strategies will inform the
teachers’ action research in year3; three teachers continued to explore additional tasks
for MPP for different grade levels and for specific topics, and cases of individuals or
teaching can be documented as exemplars (see Stein et al., 2009).
Building on the findings of this study, future research is needed on how to
implement problem posing and the conditions that allow students to perform well
on MPP, including teachers’ inexperience with implementation (Leung, 2009).
Researchers should strive to find actual examples of children’s work in year3 rather
than made-up examples. One way to do this would be by allowing teachers to invite
children to critique and rewrite the problems so that they can be understood and
solved by peers. This attempt–report–share in a group setting model provided support
in the community and was salient in teachers’ professional development. In sharing,
teachers obtained ideas on problem-posing tasks while the teacher educator acquired
information about classroom constraints. In this study, teachers voiced their queries
Teachers implementing mathematical problem posing in classroom 115

and suggestions boldly and the investigator listened patiently. Listening is clearly an
important part of in-service programs (Cooney & Krainer, 1996).

References

Bieda, K. N. (2010). Enacting proof-related tasks in middle school mathematics: Challenges and opportuni-
ties. Journal for Research in Mathematics Education, 41(4), 351–382.
Brown, S. I., & Walter, M. I. (1983). The art of problem posing. Philadelphia: Franklin Institute Press.
Cooney, T. J., & Krainer, K. (1996). In-service mathematics teacher education: The importance of
listening. In A. Bishop et al. (Eds.), International handbook of mathematics education (pp. 1155–1185). The
Netherlands: Kluwer.
Creswell, J. (2009). Research design: Qualitative, quantitative, and mixed methods approaches. New York:
Sage.
Ellerton, N. F. (1986). Children’s made up mathematics problems—a new perspective on talented mathema-
ticians. Educational Studies in Mathematics, 17, 261–271.
English, L. D. (2003). Engaging students in problem posing in an inquiry-oriented mathematics classroom. In
F. Lester & R. Charles (Eds.), Teaching mathematics through problem solving: Prekindergarten–grade 6
(pp. 187–198). Reston: National Council of Teachers of Mathematics.
English, L. D., Fox, J. L., & Watters, J. J. (2005). Problem posing and solving with mathematical modeling.
Teaching Children Mathematics, 12(3), 156–163.
Freudenthal, H. (1973). Mathematics as an educational task. Dordrecht: Reidel.
Greer, G. B., & McCann, M. (1991). Children’s word problems matching multiplication and division
calculations. In: F. Furinghetti (ed.) Proceedings of the Fifteenth International Conference for the
Psychology of Mathematics Education (vol. 2, pp. 80–87). Assisi, Italy
Hastie, P. A. (1992). Prospects for collaborations between teachers and researchers. Clearing House, 65(6),
371–372.
Henson, K. T. (1996). Teachers as researchers. In J. Sikula et al. (Eds.), Handbook of research on teacher
education (pp. 55–64). New York: Macmillan.
Jaworski, B. (1999). What does it mean to promote development in teaching? A response to Ron Tzur’s paper:
becoming a mathematics teacher-educator: Conceptualising the terrain through self-reflective analysis. In
O. Zaslavsky (ed.), Proceedings of the 23rd PME international conference, 1. pp. 183–193.
Jaworski, B. (2008). Mathematics teacher educator learning and development. In B. Jaworski & T. Wood
(Eds.), International handbook of mathematics teacher education, volume 4: The mathematics teacher
educator as a developing professional (pp. 1–13). Rotterdam: Sense.
Kemmis, J. (1991). Action research for educational change. Philadelphia: Open University Press.
Kilpatrick, J. (1978). Variables and methodologies in research on problem solving. In L. L. Hatfield (ed.).
Mathematical problem solving (pp. 7–20). ERIC/SMEAC.
Krainer, K. (2008). Individuals, teams, communities and networks: Participants and ways of participation in
mathematics teacher education. In K. Krainer & T. Wood (Eds.), International handbook of mathematics
teacher education, volume 3: Participants in mathematics teacher education (pp. 1–10). New York:
Sense.
Krainer, K., & Wood, T. (Eds). (2008) International handbook of mathematics teacher education, Volume 3:
Participants in mathematics teacher education. Sense Publishers.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. New York: Cambridge
University Press.
Leung, S. S. (1994a). On analyzing problem-posing processes: A study of prospective elementary teachers
differing in mathematics knowledge. In J. P. da Ponte & J. F. Matos (Eds.), Proceedings of the 18th
international conference of the International Group for the Psychology of Mathematics Education, vol. 3
(pp. 168–175). Lisbon: University of Lisbon.
Leung, S. S. (1994b). Research on problem posing and its role in the curriculum. In: Elementary mathematics
curriculum overview for grade 1–2 (pp. 152–169). National Academy of Educational Research, Taiwan
(in Chinese). http://www.naer.edu.tw/naerResource/study/217/12.htm
Leung, S. S. (2009). Research efforts on probing students’ conceptions in mathematics and reality: Structuring
problems, solving problems, and justifying solutions. In L. Verschaffel, B. Greer, W. Van Dooren, & S.
Mukhopadhyay (Eds.), Words and worlds: Modeling verbal descriptions of situations (pp. 213–225).
New York: Sense.
116 S.-k.S. Leung

Leung, S. S., & Silver, E. A. (1997). The role of task format, mathematics knowledge, and creative thinking on
the arithmetic problem posing of prospective elementary school teachers. Mathematics Education Re-
search Journal, 9(1), 5–24.
Ma, L. (1999). Knowing and teaching elementary mathematics. New Jersey: Erlbaum.
Ministry of Education of Taiwan (2003). Curriculum standards for national elementary schools in Taiwan
(pp. 91–132). (In Chinese)
National Council of Teachers of Mathematics. (2000). The principles and standards for school mathematics
[PSSM]. Reston: National Council of Teachers of Mathematics.
Parke, C. S., Lane, S., Silver, E. A., & Magone, M. (2003). Using assessment to improve middle-grade
mathematics teaching and learning: Suggested activities using QUASAR tasks, scoring criteria, and
student’s work. Reston, VA: National Council of Teachers of Mathematics.
Polya, G. (1945). How to solve it? Princeton, NJ: Princeton University Press.
Polya, G. (1954). Mathematics and plausible reasoning (vol. 1–2). Princeton, NJ: Princeton University Press.
Simon, H. A. (1973). The structure of ill-structured problems. Artificial Intelligence, 4, 181–201.
Sowder, J. T. (2008). The mathematics education and development of teachers. In F. K. Lester (Ed.), Second
handbook of research on mathematics teaching and learning (pp. 157–223). Virginia: National Council
of Teachers of Mathematics.
Silver, E. A. (1994). On mathematical problem posing. For the Learning of Mathematics, 14(1), 19–28.
Stein, M., Smith, M. S., Henningsen, M. A., & Silver, E. A. (2009). Implementing standards-based
mathematics instruction: A casebook for professional development (2nd ed.). VA: NCTM.
Tsubota, E. (1987). On children’s problem posing (grade 1 to 3). Japan (in Japanese).
Van den Heuvel-Panhuizen, M., Middleton, J. A., & Streefland, L. (1995). Student-generated problems: Easy
and difficult problems on percentage. For the Learning of Mathematics, 15(3), 21–27.
Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. New York: Cambridge
University Press.
Yin, R. K. (1994). Case study research: Design and methods. London: Sage.