Content uploaded by Peter Liljedahl

Author content

All content in this area was uploaded by Peter Liljedahl on Oct 15, 2018

Content may be subject to copyright.

361

© Springer International Publishing Switzerland 2016

P. Felmer et al. (eds.), Posing and Solving Mathematical Problems,

Research in Mathematics Education, DOI 10.1007/978-3-319-28023-3_21

Building Thinking Classrooms: Conditions

for Problem-Solving

Peter Liljedahl

In this chapter, I ﬁ rst introduce the notion of a thinking classroom and then present

the results of over 10 years of research done on the development and maintenance

of thinking classrooms. Using a narrative style, I tell the story of how a series of

failed experiences in promoting problem-solving in the classroom led ﬁ rst to the

notion of a thinking classroom and then to a research project designed to ﬁ nd ways

to help teachers build such a classroom. Results indicate that there are a number of

relatively easy-to-implement teaching practices that can bypass the normative

behaviours of almost any classroom and begin the process of developing a thinking

classroom.

Motivation

My work on this paper began over 10 years ago with my research on the AHA!

experience and the profound effects that these experiences have on students’ beliefs

and self-efﬁ cacy about mathematics (Liljedahl, 2005 ). That research showed that

even one AHA! experience, on the heels of extended efforts at solving a problem or

trying to learn some mathematics, was able to transform the way a student felt about

mathematics as well as his or her ability to do mathematics. These were descriptive

results. My inclination, however, was to try to ﬁ nd a way to make them prescriptive.

The most obvious way to do this was to ﬁ nd a collection of problems that provided

enough of a challenge that students would get stuck, and then have a solution, or

solution path, appear in a ﬂ ash of illumination. In hindsight, this approach was

overly simplistic. Nonetheless, I implemented a number of these problems in a

grade 7 (12–13 year olds) class.

P. Liljedahl (*)

Simon Fraser University , Burnaby , BC , Canada

e-mail: liljedahl@sfu.ca

362

The teacher I was working with, Ms. Ahn, did the teaching and delivery of prob-

lems and I observed. Despite her best intentions the results were abysmal. The stu-

dents did get stuck, but not, as I had hoped, after a prolonged effort. Instead, they

gave up almost as soon as the problem was presented to them and they resisted any

effort and encouragement to persist. After three days of constant struggle, Ms. Ahn

and I both agreed that it was time to abandon these efforts. Wanting to better under-

stand why our well-intentioned efforts had failed, I decided to observe Ms. Ahn

teach her class using her regular style of instruction.

That the students were lacking in effort was immediately obvious, but what took

time to manifest was the realization that what was missing in this classroom was

that the students were not thinking. More alarming was that Ms. Ahn’s teaching was

predicated on an assumption that the students either could not or would not think.

The classroom norms (Yackel & Rasmussen, 2002 ) that had been established had

resulted in, what I now refer to as, a non-thinking classroom. Once I realized this, I

proceeded to visit other mathematics classes—ﬁ rst in the same school and then in

other schools. In each class, I saw the same basic behaviour—an assumption,

implicit in the teaching, that the students either could not or would not think. Under

such conditions, it was unreasonable to expect that students were going to spontane-

ously engage in problem-solving enough to get stuck and then persist through being

stuck enough to have an AHA! experience.

What was missing for these students, and their teachers, was a central focus in

mathematics on thinking. The realization that this was absent in so many class-

rooms that I visited motivated me to ﬁ nd a way to build, within these same class-

rooms, a culture of thinking, both for the student and the teachers. I wanted to build,

what I now call, a thinking classroom —a classroom that is not only conducive to

thinking but also occasions thinking, a space that is inhabited by thinking individu-

als as well as individuals thinking collectively, learning together and constructing

knowledge and understanding through activity and discussion.

Early Efforts

A thinking classroom must have something to think about. In mathematics, the

obvious choice for this is a problem-solving task. Thus, my early efforts to build

thinking classrooms were oriented around problem-solving. This is a subtle depar-

ture from my earlier efforts in Ms. Ahn’s classroom. Illumination-inducing tasks

were, as I had learned, too ambitious a step. I needed to begin with students simply

engaging in problem-solving. So, I designed and delivered a three session workshop

for middle school teachers (ages 10–14) interested in bringing problem-solving into

their classrooms. This was not a difﬁ cult thing to attract teachers to. At that time,

there was increasing focus on problem-solving in both the curriculum and the text-

books. The research on the role of problem-solving as both an end unto itself and as

a tool for learning was beginning to creep into the professional discourse of teachers

in the region.

P. Liljed a hl

363

The three workshops, each 2 h long, walked teachers through three different

aspects of problem-solving. The ﬁ rst session was focused around initiating problem-

solving work in the classroom. In this session, teachers experienced a number of

easy-to-start problem-solving activities that they could implement in their class-

rooms—problems that I knew from my own experiences were engaging to students.

There were a number of mathematical card tricks to explain, some problems with

dice, and a few engaging word problems. This session was called Just do It , and the

expectation was that teachers did just that—that they brought these tasks into their

classrooms and had students just do them. There was to be no assessment and no

submission of student work.

The second session was called Teaching Problem-Solving and was designed to

help teachers emerge from their students’ experience a set of heuristics for problem-

solving. This was a signiﬁ cant departure from the way teachers were used to teach-

ing heuristics at this grade level. The district had purchased a set of resources built

on the principles of Pólya’s How to Solve It ( 1957 ). These resources were pedantic

in nature, relying on the direct instruction of these heuristics, one each day, fol-

lowed by some exercises for students to go through practicing the heuristic of the

day. This second workshop was designed to do the opposite. The goal was to help

teachers pull from the students the problem-solving strategies that they had used

quite naturally in solving the set of problems they had been given since the ﬁ rst

workshop, to give names to these strategies and to build a poster of these named

strategies as a tool for future problem-solving work. This poster also formed an

effective vocabulary for students to use in their group or whole class discussions as

well as any mathematical writing assignments.

The third workshop was focused on leveraging the recently acquired skills

towards the learning of mathematics and to begin to use problem-solving as a tool

for the daily engagement in, and learning of, mathematics. This workshop involved

the demonstration of how these new skills could intersect with the curriculum in

general and the textbook in particular.

The series of three workshops was offered multiple times and was always well

attended. Teachers who came to the ﬁ rst tended, for the most part, to follow through

with all three sessions. From all accounts, the teachers followed through with their

‘homework’ and engaged their students in the activities they had experienced within

the workshops. However, initial data collected from interviews and ﬁ eld notes were

mixed. Teachers reported things like:

“Some were able to do it.”

“They needed a lot of help.”

“They loved it.”

“They don’t know how to work together.”

“They got it quickly and didn’t want to do anymore.”

“They gave up early.”

Further probing revealed that teachers who reported that their students loved

what I was offering tended to have practices that already involved some level of

problem-solving. If there was already a culture of thinking and problem-solving in

the classroom, then this was aided by the vocabulary of the problem-solving posters,

Building Thinking Classrooms: Conditions for Problem-Solving

364

and the teachers got ideas about how to teach with problem-solving. It also revealed

that those teachers who reported that their student gave up or didn’t know how to

work together mostly had practices devoid of problem-solving and group work. In

these classrooms, although some students were able to rise to the task, the majority

of the class was unable to do much with the problems—recreating, in essence, what

I had seen in Ms. Ahn’s class. In short, the experiences that the teachers were having

implementing problem-solving in the classroom were being ﬁ ltered through their

already existing classroom norms (Yackel & Rasmussen, 2002 ).

Classroom norms are a difﬁ cult thing to bypass (Yackel & Rasmussen, 2002 ),

even when a teacher is motivated to do so. The teachers that attended these work-

shops wanted to change their practice, but their initial efforts to do so were not

rewarded by comparable changes in their students’ problem-solving behaviour.

Quite the opposite, many of the teachers I was working with were met with resis-

tance and complaints when they tried to make changes to their practice.

From these experiences, I realized that if I wanted to build thinking classrooms—

to help teachers to change their classrooms into thinking classrooms—I needed a set

of tools that would allow me, and participating teachers, to bypass any existing

classroom norms. These tools needed to be easy to adopt and have the ability to

provide the space for students to engage in problem-solving unencumbered by their

rehearsed tendencies and approaches when in their mathematics classroom.

This realization moved me to begin a program of research that would explore

both the elements of thinking classrooms and the traditional elements of classroom

practice that block the development and sustainability of thinking classrooms. I

wanted to ﬁ nd a collection of teacher practices that had the ability to break students

out of their classroom normative behaviour—practices that could be used not only

by myself as a visiting teacher but also by the classroom teacher that had previously

entrenched the classroom norms that now needed to be broken.

Thinking Classroom

As mentioned, a thinking classroom is a classroom that is not only conducive to

thinking but also occasions thinking, a space that is inhabited by thinking individu-

als as well as individuals thinking collectively, learning together and constructing

knowledge and understanding through activity and discussion. It is a space wherein

the teacher not only fosters thinking but also expects it, both implicitly and explic-

itly. As such, a thinking classroom, as I conceive it, will intersect with research on

mathematical thinking (Mason, Burton, & Stacey, 1982 ) and classroom norms

(Yackel & Rasmussen, 2002 ). It will also intersect with notions of a didactic con-

tract (Brousseau, 1984 ), the emerging understandings of studenting (Fenstermacher,

1986 , 1994 ; Liljedahl & Allan, 2013a , 2013b ), knowledge for teaching (Hill, Ball,

& Schilling, 2008 ; Shulman, 1986 ) and activity theory (Engeström, Miettinen, &

Punamäki,

1999 ).

P. Liljed a hl

365

In fact, the notion of a thinking classroom intersects with all aspects of research

on teaching and learning, both within mathematics education and in general. All of

these theories can be used to explain aspects of an already thinking classroom, and

some of them can even be used to inform us how to begin the process of build a

thinking classroom. Many of these theories have been around a long time, and yet

non-thinking classrooms abound. As such, I made the decision early on to approach

my work not from the perspective of a priori theory but from existing teaching

practices.

General Methodology

The research to ﬁ nd the elements and teaching practices that foster, sustain and

impede thinking classrooms has been going on for over 10 years. Using a frame-

work of noticing (Mason, 2002 ), 1 I initially explored my own teaching, as well as

the practices of more than 40 classroom mathematics teachers. From this emerged

a set of nine elements that permeate mathematics classroom practice—elements that

account for most of whether or not a classroom is a thinking or a non-thinking class-

room. These nine elements of mathematics teaching became the focus of my

research. They are:

1. the type of tasks used and when and how they are used

2. the way in which tasks are given to students

3. how groups are formed, both in general and when students work on tasks

4. student workspace while they work on tasks

5. room organization, both in general and when students work on tasks

6. how questions are answered when students are working on tasks

7. the ways in which hints and extensions are used, while students work on tasks

8. w hen and how a teacher levels 2 their classroom during or after tasks

9. assessm ent, both in general and when students work on tasks

Ms. Ahn’s class, for example, was one in which:

1. practice tasks were given after she had done a number of worked examples

2. students either copied these from the textbook or from a question written on the

board

3. students had the option to self-group to work on the homework assignment when

the lesson portion of the class was done

1 At the time, I was only informed by Mason ( 2002 ). Since then, I have been informed by an

increasing body of literature on noticing (Fernandez, Llinares, & Valls,

2012 ; Jacobs, Lamb, &

Philipp,

2010 ; Mason, 2011 ; Sherin, Jacobs, & Philipp, 2011 ; van Es, 2011 ).

2 Levelling (Schoenfeld, 1985 ) is a term given to the act of closing of, or interrupting, students’

work on tasks for the purposes of bringing the whole of the class (usually) up to certain level of

understanding. It is most commonly seen when a teacher ends students work on a task by showing

how to solve the task.

Building Thinking Classrooms: Conditions for Problem-Solving

366

4. students worked at their desks, writing in their notebooks

5. students sat in rows with the students’ desk facing the board at the front of the

classroom

6. students who struggled were helped individually through the solution process,

either part way or all the way

7. there were no hints, only answers, and an extension was merely the next practice

question on the list

8. when ‘enough time’ time had passed, Ms. Ahn would demonstrate the solution

on the board, sometimes calling on ‘the class’ to tell her how to proceed

9. assessment was always through individual quizzes and tests

This was not, as determined earlier, a thinking classroom. Each of these elements

was something that needed exploring and experimenting with. Many were steeped

in tradition and classroom norms (Yackel & Rasmussen, 2002 ).

Research into each of these was done using design-based methods (Cobb,

Confrey, diSessa, Lehrer, & Schauble, 2003 ; Design-Based Research Collective,

2003 ) 3 within both my own teaching practice as well as the practices of a number of

teachers participating in a variety of professional development opportunities. This

approach allowed me to vary the teaching around each of the elements, either inde-

pendently or jointly, and to measure the effectiveness of that method for building

and/or maintaining a thinking classroom. Results fed recursively back into teaching

practice , each time leading either to reﬁ ning or abandoning what was done in the

previous iteration.

This method, although fruitful in the end, presented two challenges. The ﬁ rst had

to do with the measurement of effectiveness . To do this, I used what I came to call

proxies for engagement —observable and measurable (either qualitatively or quan-

titatively) student behaviours. At ﬁ rst, this included only behaviours that ﬁ t the a

priori deﬁ nition of a thinking classroom. As the research progressed, however, the

list of these proxies grew and changed depending on the element being studied and

teaching method being used.

The second challenge had to do with the shift in practice need ed when it was

determined that a particular teaching method needed to be abandoned. Early results

indicated that small shifts in practice did little to shift the behaviours of the class as

a whole. Larger, more substantial shifts were needed. These were sometimes difﬁ -

cult to conceptualize. In the end, a contrarian approach was adopted. That is, when

a teaching method around a speciﬁ c element needed to be abandoned, the new

approach to be adopted was, as much as possible, the exact opposite to the practice

that had shown to be ineffective for building or maintaining a thinking classroom.

When sitting showed to be ineffective, we tried making the students stand. When

levelling to the top failed, we tried levelling to the bottom. When answering ques-

tions proved to be ineffective, we stopped answering questions. Each of these

3 This research is now informed also by Norton and McCloskey ( 2008 ) and Anderson and Shattuck

(

2012 ).

P. Liljed a hl

367

approaches needed further reﬁ nement through the iterative design-based research

approach, but it gave good starting points for this process .

In what follows, I will ﬁ rst present the results of the research done on two of

these elements—student workspace and how groups are formed—both indepen-

dently and jointly. I then present, in brief, the results of the research done on the

remaining seven elements and discuss how all nine elements hold together as a

framework to build and maintain thinking classrooms. All of this research is

informed dually by data and analysis that looks both on the effect on students and

the uptake by teachers.

Student Workspace

The research on student workspace began by looking at the default—students sit-

ting in their desks. It became obvious early in this work that this was not conducive

to the building of a thinking classroom. As such, almost immediately, a new space

was explored. Following the contrarian approach established early on, the next

space to test was to have students standing and working somewhere other than at

their desks. The shift to having students work on whiteboards and blackboards was

then an obvious extension.

In many classrooms where the research was being done, however, there were not

enough whiteboards and blackboards available for all groups to work at. Some stu-

dents would have to still be seated in their desks. This led to a phase of experimenta-

tion with alternative work surfaces, including poster board or ﬂ ipchart paper

attached to the walls and smaller whiteboards laying on desks—with some class-

rooms using all three at the same time. Whenever this occurred, there was a general

sense shared between whatever teachers were in the room, as well as myself, that

the vertical whiteboards were superior to any of the other options available to stu-

dents. These observations led to the following pseudo-quantitative study focusing

on this phenomenon.

Participants

The participants for this study were the students in ﬁ ve high school classrooms; two

grade 12 ( n = 31, 30), two grade 11 ( n = 32, 31) and one grade 10 ( n = 31). 4 In each

of these classes, students were put into groups of two to four and assigned to one of

ﬁ ve work surfaces to work on while solving a given problem-solving task.

4 In Canada, grade 12 students are typically 16–18 years of age, grade 11 students 15–18 and grade

10 students 14–17. The age variance is due to a combination of some students fast-tracking to be a

year ahead of their peers and some students repeating or delaying their grade 11 mathematics

course.

Building Thinking Classrooms: Conditions for Problem-Solving

368

Participating in this phase of the research were also the ﬁ ve teachers whose classes

the research took place in. Most high school mathematics teachers teach anywhere

from three to seven different classes. As such, it would have been possible to have

gathered all of the data from the classes of a single teacher. In order to diversify the

data, however, it was decided that data would be gathered from classes belonging to

ﬁ ve different teachers.

These teachers were all participating in one of several learning teams which ran

in the fall of 2006 and the spring of 2007. Teachers participated in these teams vol-

untarily with the hope of improving their practice and their students’ level of

engagement. Each of these learning teams consisted of between 4 and 6, a 2-h meet-

ing spread over half a school year. Sessions took teachers through a series of activi-

ties modelled on my most current knowledge about building and maintaining

thinking classrooms. Teachers were asked to implement the activities and teaching

methods in their own classrooms between meetings and report back to the team how

it went.

The teachers, whose classrooms this data was collected in, were all new to the

ideas being presented and, other than having individual students occasionally dem-

onstrate work on the whiteboard at the front of the room, had never used them for

whole class activity.

Data

As mentioned, the students, in groups of 2–4, worked on one of ﬁ ve assigned work

surfaces: a wall-mounted whiteboard, a whiteboard laying on top of their desks or

table, a sheet of ﬂ ipchart paper taped to the wall, a sheet of ﬂ ipchart paper laying on

top of their desk or table, and their own notebooks at their desks or table. To increase

the likelihood that they would work as a group, each group was provided with only

one felt or, in the case of working in a notebook, one pen. To measure the effective-

ness of each of these surfaces, a series of proxies for engagement were established.

It is not possible to measure how much a student is thinking during any activity,

or how that thinking is individual or predicated on and with the other members of

his or her group. However, there are a variety of proxies for this level of engage-

ment that can be established— proxies for engagement . For the research presented

here, a variety of objective and subjective proxies were established.

1 . Time to task

This was an objective measure of how much time passed between the task being

given and the ﬁ rst discernable discussion as a group about the task.

2 . Time to ﬁ rst mathematical notation

This was an objective measure of how much time passed between the task being

given and the ﬁ rst mathematical notation was made on the work surface.

3 . Eagerness to start

This is a subjective measure of how eager a group was to start working on a

task. A score of 0, 1, 2 or 3 was assigned with 0 being assigned for no enthusiasm

P. Liljed a hl

369

to begin and a 3 being assigned if every member of the group were wanting to

start.

4 . Discussion

This is a subjective measure of how much group discussion there was while

working on a task. A score of 0, 1, 2 or 3 was assigned with 0 being assigned for

no discussion and a 3 being assigned for lots of discussion involving all mem-

bers of the group.

5 . Participation

This is a subjective measure of how much participation there was from the group

members while working on a task. A score of 0, 1, 2 or 3 was assigned with 0

being assigned if no members of the group were active in working on the task

and a 3 being assigned if all members of the group were participating in the

work.

6 . Persistence

This is a subjective measure of how persistent a group was while working on a

task. A score of 0, 1, 2 or 3 was assigned with 0 being assigned if the group gave

up immediately when a challenge was encountered and a 3 being assigned if the

group persisted through multiple challenges.

7 . Non-linearity of work

This is a subjective measure of how non-linear groups work was. A score of 0, 1,

2 or 3 was assigned with 0 being assigned if the work was orderly and linear and

a 3 being assigned if the work was scattered.

8 . Knowledge mobility

This is a subjective measure of how much interaction there was between groups.

A score of 0, 1, 2 or 3 was assigned with 0 being assigned if there was no interac-

tion with another group and a 3 being assigned if there were lots of interaction

with another group or with many other groups.

These measures, like all measures, are value laden. Some proxies (1, 2, 3, 6)

were selected partially from what was observed informally when being in a setting

where multiple work surfaces were being utilized. Others proxies (4, 5, 7, 8) were

selected speciﬁ cally because they embody some of what deﬁ nes a thinking class-

room—discussion, participation, non-linear work, and knowledge mobility.

As mentioned, these data were collected in the ﬁ ve aforementioned classes dur-

ing a group problem-solving activity . Each class was working on a different task.

Across the ﬁ ve classes, there were ten groups that worked on a wall-mounted white-

board, ten that worked on a whiteboard laying on top of their desks or table, nine

that worked on ﬂ ipchart paper taped to the wall, nine that worked on ﬂ ipchart paper

laying on top of their desk or table, and eight that worked in their own notebooks at

their desks or table. For each group, the aforementioned measures were collected by

a team of 3–5 people: the teacher whose class it was, the researcher (me), as well a

number of observing teachers. The data were recorded on a visual representation of

the classroom and where the groups were located with no group being measured by

more than one person.

Building Thinking Classrooms: Conditions for Problem-Solving

370

Results and Discussion

For the purposes of this chapter, it is sufﬁ cient to show only the average scores of

this analysis (see Table 1 ).

Th e data conﬁ rmed the informal observations. Groups are more eager to start and

there is more discussion, participation, persistence and non-linearity when they

work on the whiteboards. However, there are nuances that deserve further attention.

First, although there is no signiﬁ cant difference in the time it takes for the groups to

start discussing the problem, there is a big difference between whiteboards and

ﬂ ipchart paper in the time it takes before groups make their ﬁ rst mathematical nota-

tion. This is equally true whether groups are standin g or sitting. This can be attrib-

uted to the non-permanent nature of the whiteboards. With the ease of erasing

available to them, students risk more and risk sooner. The contrast to this is the very

permanent nature of a felt pen on ﬂ ipchart paper. For students working on these

surfaces, it took a very long time and much discussion before they were willing to

risk writing anything down. The notebooks are a familiar surface to students, so this

can be discounted with respect to willingness to risk starting.

Although the measures for the whiteboards are far superior to that of the ﬂ ipchart

paper and notebook for the measures of eagerness to start, discussion, and participa-

tion, it is worth noting that in each of these cases, the vertical surface scores higher

than the horizontal one. Given that the maximum score for any of these measures is

3, it is also worth noting that eagerness scored a perfect 3 for those that were stand-

ing. That is, for all ten cases of groups working at a vertical whiteboard, ten inde-

pendent evaluators gave each of these groups the maximum score. For discussion

and participation, eight out of the ten groups received the maximum score. On the

same measures, the horizontal whiteboard groups received 3, 3, and 2 maximum

scores, respectively. This can be attributed to the fact that sitting, even while work-

ing at a whiteboard, still gives students the opportunity to become anonymous, to

hide and to not participate. Standing doesn’t afford this.

Table 1 Average times and scores on the eight measures

Vertical

whiteboard

Horizontal

whiteboard

Vertical

paper

Horizontal

paper Notebook

N (groups) 10 10 9 9 8

1. Time to task 12.8 s 13.2 s 12.1 s 14.1 s 13.0 s

2. Time to ﬁ rst

notation

20.3 s 23.5 s 2.4 min 2.1 min 18.2 s

3. Eagerness 3.0 2.3 1.2 1.0 0.9

4. Discussion 2.8 2.2 1.5 1.1 0.6

5. Participation 2.8 2.1 1.8 1.6 0.9

6. Persistence 2. 6 2.6 1.8 1.9 1.9

7. Non-linearity 2.7 2.9 1.0 1.1 0.8

8. Mobility 2. 5 1.2 2.0 1.3 1.2

P. Liljed a hl

371

With respect to non-linearity, it is clear that the whiteboards, either vertical or

horizontal, allow a greater freedom to explore the problem across the entirety of the

surface. Although the whiteboards provide an ease of erasing that is not afforded on

the ﬂ ipchart, work is rarely erased by the students working on whiteboard surfaces.

It seems that rather than erasing to make room for more work, the workspace

migrates around the whiteboard surface, representing the chronological nature of

problem-solving. In contrast, the groups working on ﬂ ipchart paper tended to not

write any work down until they were clear it would contribute to the logical devel-

opment of a solution.

Finally, it is worth noting that groups that were standing also were more likely to

engage with other groups that were standing close by. Although not measured, it

was clear that this was more true for the vertical whiteboard groups. There are a

number of reasons for this. Most obvious, vertical surfaces are more visible.

However, there were very few observed instances of groups that were sitting down

looking up to see what the groups that were standing were doing. Likewise, there

were no instances of the students standing, looking at the work of the groups that

were sitting. Amongst those that were standing, there was a lot of interaction

between those working on whiteboards, and almost none between those working on

ﬂ ipchart paper. Finally, there was very little interaction between those working on

ﬂ ipchart paper and those working on whiteboards. Part of this can be explained by

proximity—the whiteboard groups were clustered on one or two whiteboards, while

the ﬂ ipchart people were clustered elsewhere. But it also is the case that the white-

board groups had little reason to look to the ﬂ ipchart groups. They worked slower

and had little written on their work surfaces. This was also true between the ﬂ ipchart

groups—there was little to look at.

In short, groups that worked on vertical whiteboards demonstrated more think-

ing classroom behaviour—persistence, discussion, participation and knowledge

mobility—than any of the other types of work surfaces. The next most conducive

was a horizontal whiteboard. The remaining three were not only not conducive to

promoting thinking classroom behaviour but they may actually have inhibited it.

From this it is clear that the non-permanence of surfaces is critical for decreasing

time to task, as well as improving enthusiasm, discussion, participation, and persis-

tence. It also increases the non-linearity of work which mirrors the actual work of

thinking groups. Making these non-permanent surfaces vertical further enhances all

of these qualities, as well as fostering inter-group collaboration, something that is

needed to move the class from a collection of thinking groups to being a thinking

classroom.

Vertical Non-permanent Surfaces: Teacher Uptake

Having this evidence that vertical non-permanent surfaces (VNPS) are so instru-

mental in the fostering of thinking classroom behaviour, a follow-up study was

done with teachers vis-à-vis the use of this work surface. The goal of this follow-up

Building Thinking Classrooms: Conditions for Problem-Solving

372

study was to see the degree to which teachers, when presented with the idea of non-

permanent vertical surfaces, were keen to implement it within their teaching, actu-

ally tried it, and continued to use it in their teaching.

Participants

Participants for this portion of the study were 300 in-service teachers of mathemat-

ics—elementary, middle and secondary school. They were drawn from three

sources over a four-year period (2007–2011): participants in variety of single work-

shops, participants in a number of multi-session workshops, and participants in

learning teams. The breakdown of participants, according to grade levels, and form

of contact is represented in Table 2 .

There were a number of teachers who attended a combination of learning teams,

multi-session workshops and single workshops. In these cases, their data was regis-

tered as belonging to the group with the most contact. That is, if they attended a

single workshop, as well as being a member of a learning team, their participation

was registered as being a member of a learning team.

These participants are only a subset of all the teachers that participated in these

learning teams, multi-session workshops, and single workshops. They were selected

at random from each group I worked with by approaching them at the end of the ﬁ rst

(and sometimes only) session and asking them if they would be willing to have me

contact them and, potentially, visit their classrooms.

Data

Data consists primarily of interview data. Each participant was interviewed imme-

diately after a session where they were ﬁ rst introduced to the idea of vertical non-

permanent surfaces, 1 week later, and 6 weeks later. These interviews were brief

and, depending on when the interview was conducted, was originally designed to

gauge the degree to which they were committed to trying, or continuing to use,

vertical non-permanent surfaces in their teaching and how they were using them.

However, participants wanted to talk about much more than just this. They wanted

to discuss innovations they had made, the ways in which this was changing their

Table 2 Distribution of participants in VNPS study

Elementary Middle Secondary Total

Learning team 21 43 41 105

Multi-session workshops 12 28 42 82

Single workshops 35 24 54 113

Total 68 95 137 300

P. Liljed a hl

373

teaching practice as a whole, the reactions of the students and their colleagues, as

well as a variety of other details pertaining to vertical non-permanent surfaces. With

time, these impromptu conversations changed the initial interview questions to

begin to also probe for these more nuanced details. For the purposes of this chapter,

however, only the data pertaining to the original intent will be presented.

In addition to the interview data, there were also ﬁ eld notes from 20 classroom

visits. These visits were implemented for the purposes of checking the ﬁ delity of the

interview data—to see if what teachers were saying is actually what they were

doing. In each case, this proved to be the case. It was clear from these data that

teachers were true to their words with respect to their use of vertical non-permanent

surfaces. However, these visits, like the interviews, offered much more than what

was expected. I saw innovations in implementation, observed the enthusiasm of the

students, and witnessed the transformational effect that this was having on the

teaching practices of the participants .

Results and Discussion

In general, almost all of the teachers who were introduced to the notion of vertical

non-permanent surfaces were determined to try it within their teaching and were

committed to keep doing it, even after 6 weeks (see Fig. 1 ). This is a signiﬁ cant

uptake rarely seen in the literature. This is likely due, in part, to the ease with which

it is modelled in the various professional development settings. During these ses-

sions, not only are the methods involved easily demonstrated but the teachers

immediately feel the impact on themselves as learners when they are put into a

group to work on a vertical non-permanent surface.

An interesting result from this aggregated view is that there were more teachers

using non-permanent vertical surfaces after 6 weeks than there was after 1 week.

This has to do with access to these vertical non-permanent surfaces. Many teachers

struggled to ﬁ nd such surfaces. There were some amazing improvisations in this

regard, from using windows to bringing in a number of novel surfaces, from shower

curtains to glossy wall boards. One teacher even stood her classroom tables on end

to achieve the effect. As time went on, teachers were able to convince their admin-

istrators to provide them with enough whiteboards that these improvisations no lon-

ger became necessary. For some teachers, this took more time than others and

speaks to the delayed uptake seen in Fig. 1 . However, it also speaks to the persis-

tence with which many teachers pursued this idea with.

A disaggregated look at the data shows that neither the grade levels being taught

(see Fig. 2 ) or the type of professional development setting in which the idea was

presented (see Fig. 3 ) had any signiﬁ cant impact on the uptake.

Literature on teacher change typically implies that sustained change can only be

achieved through professional development opportunities with multiple sessions

and extended contact. That is, single workshops are not effective mediums for pro-

moting change (Jasper & Taube,

2004 ; Little & Horn, 2007 ; Lord, 1994 ; McClain

Building Thinking Classrooms: Conditions for Problem-Solving

374

& Cobb, 2004 ; Middleton, Sawada, Judson, Bloom, & Turley, 2002 ; Stigler &

Hiebert, 1999 ; Wenger, 1998 ). The introduction of vertical non-permanent surfaces

as a workspace doesn’t adhere to these claims. There are many possible reasons for

this. The ﬁ rst is that the introduction of non-permanent vertical surfaces was

achieved in a single workshop could be, as mentioned, due to the simple fact that it

is a relatively easy idea for a workshop leader to model and for workshop partici-

pants to experience. Forty ﬁ ve minutes of solving problems in groups standing at a

whiteboard coupled with a whole group discussion on the affordances of recreating

this within their own classrooms is enough to convince teachers to try it. And trying

it leads to a successful implementation. Unlike many other changes that can be

made in a teacher’s practice, vertical non-permanent surfaces (as demonstrated in

100

96

100

100

100

95

99

99

100

95

96

96

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six

weeks

intends to keep doing it

after six weeks

PERCENT

elementary (n=68) middle (n=95) secondary (n=137)

Fig. 2 Uptake of VNPS by grade levels ( n = 300)

100

95

98

98

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six weeks

weeks

intends to keep doing it

after six weeks

PERCENT

Fig. 1 Uptake of VNPS ( n = 300)

P. Liljed a hl

375

the ﬁ rst study) was well received by students, was easy to manage at a whole class

level, and had an immediate positive effects on classroom thinking behaviour.

Together, the ease of modelling coupled with a successful implementation meant

that vertical non-permanent surfaces did not need more than a single workshop to

change teaching practice.

These possible reasons are supported by the comments of teachers from the

interviews after week 1 and week 6. The following comments were chosen from the

many collected for their conciseness.

“ I will never go back to just having students work in their desks .”

“ How do I get more whiteboards ?”

“ The principal came into my class … now I’m doing a session for the whole staff on Monday .”

“ My grade - partner is even starting to do it .”

“ The kids love it. Especially the windows .”

“ I had one girl come up and ask when it will be her turn on the windows .”

Not only is the implementation of vertical non-permanent surfaces immediately

effective for these teachers, it is also infectious with other teachers quickly latching

on to it and administrators quickly seeing the affordances it offers.

But if vertical non-permanent surfaces are the solution, what was the problem?

When I began the research on students’ workspace, the default was students sitting

in desks—sometimes individually in rows, other times clustered in groups. The

move from the desks to the vertical workspaces was made, not because I saw some-

thing speciﬁ cally wrong with students being in desks, but rather through adherence

to the contrarian approach that was adopted early on in the more general research

project. Looking back now at students working in desks, from the perspective of the

affordances that having them stand at a non-permanent vertical surface offers, I see

100

96

100

100

100

93

95

95

100

96

98

98

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six

weeks

intends to keep doing it

after six weeks

PERCENT

learning team (n=105) multi-session workshops (n=82) single workshop (n=113)

Fig. 3 Uptake of VNPS by professional development setting ( n = 300)

Building Thinking Classrooms: Conditions for Problem-Solving

376

more clearly the problems that desks introduced into my efforts to build and main-

tain thinking classrooms. Primarily, this has to do with anonymity and how desks

allow for and even promote this. When students stand at a whiteboard or a window,

they are all visible. There is nowhere to hide. When students are in their desks, it is

easy for them to become anonymous, hidden and safe—from participating and

from contributing. It is not that all students want to be hidden, to not participate, but

when the problems gets difﬁ cult, when the discussions require more thinking, it is

easy for a student to pull back in their participation when they are sitting. Standing

in a group makes this more difﬁ cult. Not only is it immediately visible to the teacher

but it is also clear to the student who is pulling back. To pull back means to step

towards the centre of the room, towards the teacher, towards nothing. There is no

anonymity in this.

Forming Groups

The research into how best to form groups began, like it did with student work sur-

faces, by looking at how groups are typically formed in a classroom. In most cases,

this is either a strategically planned arrangement decided by the teacher or self-

selected groupings of friends as decided by the students. Teachers tend to make

groupings in order to meet their educational goals. These may include goals around

pedagogy, student productivity, or simply the construction of a peaceful work envi-

ronment. Meanwhile, students, when given the opportunity, tend to group them-

selves according to their social goals. This mismatch between educational and

social goals in classrooms creates conditions where, no matter how strategic a

teacher is in her groupings, some students are unhappy in the failure of that group-

ing to meet their social goals (Kotsopoulos, 2007 ; Slavin, 1996 ).

This disparity results in a decrease in the effectiveness of group work. This led to

the exploration of alternative grouping methods. The fact that strategic grouping

strategies were often not working, coupled with the contrarian approach of action in

such instances, meant that random grouping methods needed to be explored.

Working with the same type of population of teachers described above, a variety of

random grouping methods were implemented and studied. This preliminary research

showed, very quickly, that there was little difference in the effectiveness of strategic

groupings and randomized groupings when the randomization was done out of sight

of the students. The students assumed that all groupings had a hidden agenda, and

merely saying that they were randomly generated was not enough to change class-

room behaviour.

However, when the randomization was done in full view of the students, changes

were immediately noticed. When randomization was done frequently—twice a day

in elementary classrooms and every class in middle and secondary classrooms—the

changes in classroom behaviour was profound. Within 2–3 weeks:

• Students became agreeable to work in any group they were placed in.

• There was an elimination of social barriers within the classroom.

P. Liljed a hl

377

• Mobility of knowledge between students increased.

• Reliance on the teacher for answers decreased.

• Reliance on co-constructed intra- and inter-group answers increased.

• Engagement in classroom tasks increased.

• Students became more enthusiastic about mathematics class.

To conﬁ rm these observations, one grade 10 (age 15–16) was studied. The details

and results of this research have already been published in a chapter entitled The

Affordances of Using Visibly Random Groups in a Mathematics Classroom

(Liljedahl, 2014 ). What follows is a summary of this research.

The class in which the study was done belonged to Ms. Carley, a teacher with

eight years experience who was a participant in one of the learning teams I was

leading. Ms. Carley had joined the team because she was dissatisﬁ ed with the results

of group work in her teaching. She knew that group work was important to learning,

but, until now, had felt that her efforts in this regard had been unsuccessful. She was

looking for a better way. So, when I suggested to the group that they try using vis-

ibly random groups she made an immediate commitment to start using this method

in one of her classrooms.

Data consisted of interview transcripts and ﬁ eld notes collected over a 3-month

period immediately prior to and during an implementation of visibly random groups

in Ms. Carley’s class. These data were analysed using analytic induction (Patton,

2002 ) anchored in the a priori and grounded observations from my initial experi-

mentation with random groupings.

These results both conﬁ rmed and nuanced the initial observations. Students very

quickly shed their anxieties about what groups they were in. They began to collabo-

rate in earnest. After three weeks, a porosity developed between group boundaries

as both intra- and inter-group collaboration ﬂ ourished. With this heightened mobi-

lization of knowledge came a decrease in the reliance on the teacher as the knower

in the room. In the end, there was a marked heightening of enthusiasm and engage-

ment for problem-solving in particular, and in mathematics class in general. In

short, Ms. Carley’s class became a thinking classroom .

Visibly Random Groupings: Teacher Uptake

Similar to the research on the vertical non-permanent surfaces a pseudo-quantitative

study was done on the uptake by teachers on the idea of visibly random groupings

(VRG). Tapping into the similar populations of teachers engaged in learning teams,

multi-session workshops, and single workshops between 2009 and 2011, a popula-

tion of 200 teachers were selected to participate (see Table 3 ).

These teachers were introduced to the idea of visibly random groupings in a

similar fashion as above—through modelling and immersion. They were likewise

interviewed immediately after their professional development experience, 1 week

after their experience, and 6 weeks after. The results of this analysis can be seen in

Fig.

4 .

Building Thinking Classrooms: Conditions for Problem-Solving

378

The dip in the uptake between week 1 and week 6 was minor. What was interest-

ing was the uptick in intension after week 6. From the interviews, it became clear

that the teachers who had come away from using visibly random groups did so

because, after 3–4 weeks, things were working so well that they thought they could

now allow the students to work with who they wanted. Once they saw that this was

not as effective, they recommitted to going back to random groupings.

Like with vertical non-permanent surfaces, there was no discernible difference in

uptake between elementary, middle or secondary teachers. However, unlike the pre-

vious study, there was a slight difference depending on the nature of the profes-

sional development environment they were participating in (see Fig. 5 ).

From the interviews, it seemed that although the immediate delivery of the idea

was accomplished within a single session, the support of the learning team helped

teachers to get on board late if they hesitated in implementing in the 1st week. This

explains the uptick in the number of learning team members who started using ran-

domized groups in between the ﬁ rst and the sixth week. This also explains why

there was no such uptick amongst the single workshop participants who had no

follow-up session, or amongst the multi-session participants who did not have a

second session until 8 weeks after the initial idea was presented.

Regardless, there was still a signiﬁ cant uptake by those teachers who only expe-

rienced one 90 min session on the use of visibly random groupings. This can be

explained in the same way as it was for the vertical non-permanent surfaces—it was

easily modelled and the affordances became immediately apparent. As well, the

students took to it quickly with little resistance once the participants implemented it

within their own classrooms.

Table 3 Distribution of participants in VRG study

Elementary Middle Secondary Total

Learning team 15 22 31 68

Multi-session workshops 25 19 14 58

Single workshops 10 25 39 74

Total 50 66 84 200

93

91

88

93

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six weeks

weeks

intends to keep doing it

after six weeks

PERCENT

Fig. 4 Uptake of VRG ( n = 200)

P. Liljed a hl

379

As with the research on the vertical non-permanent surfaces, the research on vis-

ibly random groupings included 14 classroom visits. Unlike the research on VNPS,

however, the purpose of these visits was not to check the ﬁ delity of the interview

data. Rather, it was to see if teachers were continuing to use VRG’s even 6–9

months after their last work with me. In each of the 14 visits, I saw a continued use

of VRG strategies. And like with my visits in the VNPS research, these visits

offered much more than what was expected. I saw innovations in implementation,

observed the enthusiasm of the students, and witnessed the transformational effect

that this was having on teaching practices.

VNPS and VRG Taken Together: Teacher Uptake

Once it was established that both vertical non-permanent surfaces and visibly ran-

dom groupings were effective practices for building aspects of a thinking classroom

and that these methods had good uptake by teachers, it was easy to bring them

together. From a professional development perspective, this is no more difﬁ cult

than presenting each one separately. VNPS and VRG are easily modelled together,

with the participants being put into visibly random groupings to work on vertical

non-permanent surfaces. So, this is what was done with a population of teachers

similar to the ones described above. From this, 124 participants were followed to

gauge the uptake of being exposed to both of these methods simultaneously. The

results can be seen in Fig.

6 .

93

93

99

99

91

88

83

91

95

92

82

89

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six

weeks

intends to keep doing it

after six weeks

PERCENT

learning team (n=105) multi-session workshops (n=82) single workshop (n=113)

Fig. 5 Uptake of VRG by professional development setting ( n = 200)

Building Thinking Classrooms: Conditions for Problem-Solving

380

Like with visibly random groupings, there was no signiﬁ cant difference in uptake

by grade level and a slight difference in uptake as disaggregated by the professional

development setting in which the combined methods were presented. Like with vis-

ibly random groupings, the teachers in the learning team setting were more consis-

tently implementing the methods presented, whereas those teachers in the single

workshop sessions were less likely to get on board late and more likely to drop off

early (see Fig. 7 ). Despite these differences, however, the uptake across for each

group was impressive with much enthusiasm for it.

With respect to the effect on students, my observations during ten classroom

visits showed the combined beneﬁ ts of the two interventions. The fact that the stu-

dents were so comfortable working with each other, coupled with the high visibility

94

90

90

92

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six weeks

weeks

intends to keep doing it

after six weeks

PERCENT

Fig. 6 Uptake of both VNPS and VRG ( n = 124)

100

95

100

100

100

94

95

95

100

94

95

95

0

10

20

30

40

50

60

70

80

90

100

immediate intension tried it in the first week still doing it after six

weeks

intends to keep doing it

after six weeks

PERCENT

learning team (n=53) multi-session workshops (n=26) single workshop (n=45)

Fig. 7 Uptake of VNPS and VRG by professional development setting ( n = 124)

P. Liljed a hl

381

of the work afforded by the vertical surfaces, allowed for enhanced intra-group

knowledge mobilization. The teachers often commented that they saw huge

improvements in the classroom community.

“I used to think I had a community in my classroom. Now I see what a community can look

like .”

My observation of the student actions during these ten classroom visits conﬁ rmed this .

General Findings: All Nine Elements

The results from research on students’ workspace and grouping methods are indicative

of the ﬁ ndings of research into each of the nine aforementioned elements. From the

design-based research on each of these—independently or in conjunction with

others—emerged a set of teaching practices that are conducive to either the building,

or maintenance, of a thinking classroom. In what follows brieﬂ y, these are:

1 . The type of tasks used and when and how they are used

Lessons need to begin with good problem-solving tasks. At the early stages of

building a thinking classroom, these tasks need to be highly engaging,

collaborative tasks that drive students to want to talk with each other as they try

to solve them (Liljedahl, 2008 ). Once a thinking classroom is established, the

problems need to permeate the entirety of the lesson and emerge rich mathe-

matics (Schoenfeld, 1985 ) that can be linked to the curriculum content to be

‘taught’ that day.

2 . The way in which tasks are given to students

Tasks need to be given orally. If there are data or diagrams needed, these can be

provided on paper, but the instructions pertaining to the activity of the task need

to be given orally. This very quickly drives the groups to discuss what is being

asked rather than trying to decode instructions on a page.

3 . H ow groups are formed, both in general and when students work on tasks

As presented above, groupings need to be frequent and visibly random. Ideally,

at the beginning of every class, a visibly random method is used to assign stu-

dents to a group of 2–4 for the duration of that class. These groups will work

together on any assigned problem-solving tasks, sit together or stand together

during any group or whole class discussions.

4 . Student workspace while they work on tasks

As discussed, groups of students need to work on vertical non-permanent sur-

faces such as whiteboards, blackboards, or windows. This will make visible all

work being done, not just to the teacher but to the groups doing the work. To

facilitate discussion, there should be only one felt pen or piece of chalk per

group.

5 . Room organization, both in general and when students work on tasks

The classroom needs to be de-fronted. The teacher must let go of one wall of the

classroom as being the designated teaching space that all desks are oriented

Building Thinking Classrooms: Conditions for Problem-Solving

382

towards. The teacher needs to address the class from a variety of locations within

the room and, as much as possible, use all four walls of the classroom. It is best

if desks are placed in a random conﬁ guration around the room.

6 . How questions are answered when students are working on tasks

Students only ask three types of qu estions: (1) proximity questions—asked when

the teacher is close; (2) stop-thinking questions—most often of the form ‘is this

right’; and (3) keep-thinking questions—questions that students ask so they can

get back to work. Only the third of these types should be answered. The ﬁ rst two

types need to be acknowledged but not answered.

7 . The ways in which hints and extensions are used while students work on tasks

Once a thinking classroom is established, it needs to be nurtured. This is done

primarily through how hints and extensions are given to groups as they work on

tasks. Flow (Csíkszentmihályi 1990 , 1996 ) is a good framework for thinking

about this. Hints and extensions need to be given so as to keep students in a per-

fect balance between the challenge of the current task and their abilities in work-

ing on it. If their ability is too high, the risk is they get bored. If the challenge is

too great, the risk is they become frustrated.

8 . When and how a t eacher levels their classroom during or after tasks

Levelling needs be done at the bottom. When every group has passed a minimum

threshold, the teacher needs to engage in discussion about the experience and

understanding the whole class now shares. This should involve a reiﬁ cation and

formalization of the work done by the groups and often constitutes the ‘lesson’

for that particular class.

9 . Assessment, both in general and when students work on tasks

Assessment in a thinking classroom needs to be mostly about the involvement of

students in the learning process through efforts to communicate with them where

they are and where they are going in their learning. It needs to honour the

activities of a thinking classroom through a focus on the processes of learning

more so than the products and it needs to include both group wo rk and individual

work.

Discussion

However, this research also showed that these are not all equally impactful or pur-

poseful in the building and maintenance of a thinking classroom. Some of these are

blunt instruments capable of leveraging signiﬁ cant changes while others are more

reﬁ ned, used for the ﬁ ne-tuning and maintenance of a thinking classroom. Some

are necessary precursors to others. Some are easier to implement by teachers than

others, while others are more nuanced, requiring great attention and more practice

as a teacher. And some are better received by students than others. From the whole

of these results emerged a three-tier hierarchy that represent not only the bluntness

and ease of implementation but also an ideal chronology of implementation (see

Table 4 ).

P. Liljed a hl

383

In the aforementioned research, I presented the results of research into teachers

implementing teaching practices from stage one, either separately or together.

However, the effect on these teachers is more profound than the numbers and graphs

indicated above. This experience with elements in stage one propels them to thirst

for more, both in particular and in general. They want more tasks, more examples

of how to make random groupings, how to ﬁ nd vertical surfaces. But they also want

to know more about assessment, how to ask and answer questions, how to organize

their rooms, how to give instructions and how to sustain the engagement they have

experienced while at the same time feeling like they are getting through the curricu-

lum. In short, their experience with the teaching methods associated with stage one

elements is quite naturally propelling them into wanting to engage in the elements

in stages two and three.

These results are not deﬁ nitive, exhaustive or unique. The teaching methods that

emerged as effective for each of these elements emerged as a result of an a priori

commitment to make change in a contrarian fashion. This continued until positive

effects began to emerge, at which point reﬁ nements were recursively explored. It is

possible that a different approach to the research would have yielded different

methods. Different methods could, likewise, emerge a different set of stages opti-

mal for the development of thinking classrooms.

Conclusions

The main goal of this research is about ﬁ nding ways to build thinking classrooms.

One of the sub-goals of this work on building thinking classrooms was to develop

methods that not only fostered thinking and collaboration but also bypassed any

classroom norms that would potentially inhibit this from happening. Using the

methods in stage one while solving problems, either together or separately, was

almost universally successful. They worked for any grade, in any class and for any

teacher. As such, it can be said that these methods succeeded in bypassing whatever

norms existed in the over 600 classrooms in which these methods were tried.

Further, they not only bypassed the norms for the students but also the norms of the

Table 4 Nine elements as chronologically implemented

Stage one Stage two Stage three

• Begin lessons with problem-solving

tasks

• Oral instructions • Levelling

• Vertical non-permanent surfaces • De-fronting the room • Assessment

• Visibly random groups • Ans wering questions • Managing ﬂ ow

BLUNTNESS

DIFFICULTY OF IMPLEMENTATION

Building Thinking Classrooms: Conditions for Problem-Solving

384

teachers implementing them. So different were these methods from the existing

practices of the teachers participating in the research that they were left with what I

have come to call ﬁ rst-person vicarious experiences . They are ﬁ rst person because

they are living the lesson and observing the results created by their own hands. But

the methods are not their own. There has been no time to assimilate them into their

own repertoire of practice or into the schema of how they construct meaningful

practice. They simply experienced the methods as learners and then were asked to

immediately implement them as teachers. As such, they experienced a different way

in which their classroom could look and how their students could behave. They

experienced, through these other ly methods, an other ly classroom—a thinking

classroom.

The results of this research sound extraordinary. In many ways, they are. It

would be tempting to try to attribute these to some special quality of the profes-

sional development setting or skill of the facilitator. But these are not the source of

these remarkable results. The results, I believe, lie not in what is new but what is not

old. The classroom norms that permeate classrooms in North America, and around

the world, are so robust, so entrenched, that they transcend the particular classrooms

and have become institutional norms (Liu & Liljedahl, 2012 ). What the methods

presented here offer is a violent break from these institutional norms, and in so

doing, offer students a chance to be learners much more so than students (Liljedahl

& Allan, 2013a , 2013b ).

By constructing a thinking classroom, problem-solving becomes not only a

means but also an end. A thinking classroom is shot through with rich problems.

Implementation of each of the aforementioned methods associated with the nine

elements and three stages relies on the ubiquitous use of problem-solving. But at the

same time, it also creates a classroom conducive to the collaborative solving of

problems.

Afterword

Since this research was completed, I have gone back to visit several of the class-

rooms of teachers who ﬁ rst took part in the research. These teachers are still using

VNPS and VRG as well as having reﬁ ned their practice around many of the other

nine aforementioned elements. Unlike many other professional development initia-

tives and interventions I have seen implemented over the years, these really seemed

to have had a lasting impact on teacher practice. The reason for this seems to come

from two sources. First, teachers talk about how much their students like the ‘new’

way of doing mathematics. So much so, in fact, that when they go back to using

direct instruction, even for brief periods of time, the students object. The second and

more intrinsic reason is that they feel more effective as teachers. Their students are

exhibiting the traits that they had been striving for but were unable to achieve

through nuanced changes to their initial teaching practice.

P. Liljed a hl

385

References

Anderson, T., & Shattuck, J. (2012). Design-based research: A decade of progress in education

research? Educational Researcher, 41 (1), 16–25.

Brousseau, G. (1984). The crucial role of the didactical contract in the analysis and construction of

situations in teaching and learning mathematics. In H.-G. Steiner (Ed.), Theory of Mathematics

Education: ICME 5—Topic Area and Miniconference: Adelaide, Australia . Bielefeld, Germany:

Institut fuer Didaktik der Mathematik der Universitaet Bielefeld.

Cobb, P., Confrey, J., diSessa, A., Lehrer, R., & Schauble, L. (2003). Design experiments in edu-

cational research. Educational Researcher, 32 (1), 9–13.

Csíkszentmihályi, M. (1990). Flow: The psychology of optimal experience . New York: Harper &

Row.

Csíkszentmihályi, M. (1996). Creativity: Flow and the psychology of discovery and invention .

New York: HarperCollins.

Design-Based Research Collective. (2003). Design-based research: An emerging paradigm for

educational inquiry. Educational Researcher, 32 (1). 5–8, 35–37.

Engeström, Y., Miettinen, R., & Punamäki, R. (1999). Perspectives on activity theory . Cambridge,

England: Cambridge University Press.

Fenstermacher, G. (1986). Philosophy of research on teaching: Three aspects. In M. C. Whittrock

(Ed.), Handbook of research on teaching (3rd ed., pp. 37–49). New York: Macmillan.

Fenstermacher, G. (1994, revised 1997). On the distinction between being a student and being a

learner . Paper presented at the annual meeting of the American Educational Research

Association, New Orleans, LA.

Fernandez, C., Llinares, S., & Valls, J. (2012). Learning to notice students’ mathematical thinking

through on-line discussions. ZDM—The International Journal on Mathematics Education,

44 (6), 747–759.

Hill, H., Ball, D., & Schilling, S. (2008). Unpacking pedagogical content knowledge:

Conceptualizing and measuring teachers’ topic-speciﬁ c knowledge of students. Journal for

Research in Mathematics Education, 39 (4), 372–400.

Jacobs, V. R., Lamb, L. L., & Philipp, R. A. (2010). Professional noticing of children’s mathemati-

cal thinking. Journal for Research in Mathematics Education, 41 (2), 169–202.

Jasper, B., & Taube, S. (2004). Action research of elementary teachers’ problem-solving skills

before and after focused professional development. Teacher Education and Practice, 17 (3),

299–310.

Kotsopoulos, D. (2007). Investigating peer as “expert other” during small group collaborations in

mathematics. In Proceedings of the 29th annual meeting of the North American Chapter of the

International Group for the Psychology of Mathematics Education . Lake Tahoe, NV: University

of Nevada, Reno.

Liljedahl, P. (2005). Mathematical discovery and affect: The effect of AHA! experiences on under-

graduate mathematics students. International Journal of Mathematical Education in Science

and Technology, 36 (2–3), 219–236.

Liljedahl, P. (2008). The AHA! experience: Mathematical contexts, pedagogical implications .

Saarbrücken, Germany: VDM Verlag.

Liljedahl, P. (2014). The affordances of using visually random groups in a mathematics classroom.

In Y. Li, E. Silver, & S. Li (eds.) Transforming Mathematics Instruction: Multiple Approaches

and Practices . New York, NY: Springer.

Liljedahl, P., & Allan, D. (2013a). Studenting: The case of “now you try one”. In A. M. Lindmeier

& A. Heinze (Eds.), Proceedings of the 37th Conference of the International Group for the

Psychology of Mathematics Education (Vol. 3, pp. 257–264). Kiel, Germany: PME.

Liljedahl, P. & Allan, D. (2013b). Studenting: The case of homework. In Proceedings of the 35th

Conference for Psychology of Mathematics Education—North American Chapter . Chicago, IL,

USA.

Building Thinking Classrooms: Conditions for Problem-Solving

386

Little, J. W., & Horn, I. S. (2007). ‘Normalizing’ problems of practice: Converging routine conversation

into a resource for learning in professional communities. In L. Stoll & K. S. Louis (Eds.),

Professional learning communities: Divergence, depth, and dilemmas (pp. 79–92). Berkshire,

England: Open University Press.

Liu, M., & Liljedahl, P. (2012). ‘Not normal’ classroom norms. In T. Y. Tso (Ed.), Proceedings of

the 36th Conference of the International Group for the Psychology of Mathematics Education .

Taipei, Taiwan.

Lord, B. (1994). Teachers' professional development: Critical colleagueship and the roles of

professional communities. In N. Cobb (ed.), The Future of Education: Perspectives on National

Standards in America (pp. 175–204). New York, NY: The College Board.

Mason, J. (2002). Researching your own practice: The discipline of noticing . New York: Routledge.

Mason, J. (2011). Noticing: Roots and branches. In M. G. Sherin, V. Jacobs, & R. Philipp (Eds.),

Mathematics teacher noticing (pp. 35–50). New York: Routledge.

Mason, J., Burton, L., & Stacey, K. (1982). Thinking mathematically . London: Addison-Wesley.

McClain, K., & Cobb, P. (2004). The critical role of institutional context in teacher development.

In Proceedings of 28th Annual Conference for the Psychology of Mathematics Education

(Vol. 3, pp. 281–288).

Middleton, J. A., Sawada, D., Judson, E., Bloom, I., & Turley, J. (2002). Relationships build

reform: Treating classroom research as emergent systems. In L. D. English (Ed.), Handbook of

international research in mathematics education (pp. 409–431). Mahwah, NJ: Lawrence

Erlbaum Associates.

Norton, A. H., & McCloskey, A. (2008). Teaching experiments and professional development.

Journal of Mathematics Teacher Education, 11 (4), 285–305.

Patton, M. Q. (2002). Qualitative research and evaluation methods . Thousand Oaks, CA: Sage.

Pólya, G. (1957). How to solve it (2nd ed.). Princeton, NJ: Princeton University Press.

Schoenfeld, A. (1985). Mathematical problem solving . Orlando, FL: Academic Press.

Sherin, M. G., Jacobs, V., & Philipp, R. (2011). Situating the study of teacher noticing. In M. G.

Sherin, V. Jacobs, & R. Philipp (Eds.), Mathematics teacher noticing (pp. 3–14). New York:

Routledge.

Shulman, L. S. (1986). Those who understand: Knowledge growth in teaching. Educational

Researcher, 15 (2), 4–31.

Slavin, R. E. (1996). Research on cooperative learning and achievement: What we know, what we

need to know. Contemporary Educational Psychology, 21 , 43–69.

Stigler, J., & Hiebert, J. (1999). The Teaching Gap. Best Ideas from the World’s Teachers for

Improving Education in the Classroom . New York, NY: The Free Press.

van Es, E. (2011). A framework for learning to notice student thinking. In M. G. Sherin, V. Jacobs,

& R. Philipp (Eds.), Mathematics teacher noticing (pp. 134–151). New York: Routledge.

Wenger, E. (1998). Communities of practice . New York: Cambridge University Press.

Yackel, E., & Rasmussen, C. (2002). Beliefs and norms in the mathematics classroom. In G. Leder,

E. Pehkonen, & G. Törner (Eds.), Beliefs: A hidden variable in mathematics education?

(pp. 313–330). London: Kluwer Academic.

P. Liljed a hl