International Electronic Journal of Elementary Education

January 2022, Volume 14, Issue 3, 269-281

Elementary Students’ Exploration

of the Structure of a Word Problem
Using Representations
Ji-Eun Leea, Sunghwan Hwangb,*
Received : 11 October 2021 Abstract
Revised : 13 December 2021
Accepted : 28 january 2022 Word problems are frequently used in school mathematics
DOI : 10.26822/iejee.2022.243 to offer students the opportunity to explore mathematical
relationships and structure. However, previous studies have
a
Ji-Eun Lee, Department of Teacher Development and
Educational Studies, Oakland University, Michigan, USA reported that word problems are misused or abused in ways
E-mail: lee2345@oakland.edu that overlook the original intent of exploring mathematical
ORCID: https://orcid.org/0000-0003-2246-6082 structure and relationship. This study aims to share a series
Corresponding author: Sunghwan Hwang, Department
b,* of a small group of third-grade students’ explorations
of Elementary Mathematics Education, Seoul National while debating the mathematical relationships in solving
University of Education, Seoul, South Korea a word problem with representations over several days.
E-mail: ihwang413@gmail.com
ORCID: https://orcid.org/0000-0001-8212-6368 Although the exploration took longer than planned, it was
worthwhile. It offered students a space to express confusion,
showcase their knowledge, test conjectures, and imagine
alternative contexts. Ultimately, these explorations helped
students recognize multiple relationships within the context
of specific problems while bringing their attention to real-
world related applications. The retrospective analysis of
class episodes offers insight into learning opportunities
to support students in exploring mathematical structure
and relationships while discussing and debating the word
problem context.

Keywords:
Additive and Multiplicative Relationships, Classroom Culture,
Elementary Education, Schematic Representation, Structure,
Word Problems

Introduction

U nderstanding and generalizing mathematical

relationships and structures in learning mathematics
are critical (Davydov, 1990; Mason, 2003; Sierpinska, 1994;
Thompson, 2011). The ability to “look closely to discern a
pattern or structure” is an essential skill that mathematics
learners should develop (National Governors Association
Center for Best Practices & Council of Chief State School
Officers, 2010, p. 8). Also, “the detection and exploitation
of structural relationships” is considered an essential
Copyright © mathematics component (Greer & Harel, 1998, p. 22).
www.iejee.com
ISSN: 1307-9298 Word problems are frequently used in school mathematics
to offer students the opportunity to explore mathematical
© 2022 Published by KURA Education & Publishing. relationships and structure. However, studies have reported
This is an open access article under the CC BY- that students face varying challenges and difficulties
NC- ND license. (https://creativecommons.org/
in handling world problems (Verschaffel et al., 2020).
licenses/by/4.0/)
Researchers discussed situations in which word problems are

269
 January 2022, Volume 14, Issue 3, 269-281

misused or abused, resulting in blocking the intended students to engage in mathematical discussions about
results of mathematical exploration (e.g., Mason, 2001; using and understanding schematic representation
Verschaffel et al., 2000). In particular, studies reported to improve students' problem-solving abilities in word
that many students tend to dive into calculations by problems.
grabbing given numbers and using known procedures
and operations or rely on keywords, rather than In short, despite the many possible supporting tools and
analyzing the structure of the problem as a means approaches, “what seems to matter most is not the
to solve the problem (Littlefield & Reiser, 1993; Savard apparatus itself, but how it is used” (Mason, 2018, p.332).
& Polotskaia, 2017; Stigler et al., 1990; Verschaffel et al., Good tools and approaches to real-world problems,
2000). Additionally, students generally produce one such as the aforementioned ones, can be (and often
answer in the form of a numerical symbol and seem are) incorrectly presented through a teacher-led, top-
unwilling to bring anything further into the problem- down presentation rather than as an apparatus for
solving process. Students also believe that there is only student-centered exploration. When students have
one correct answer or one correct process for finding more opportunities to play with, be curious about, and
a solution for a word problem. These two tendencies explore word problems by changing the context and
often keep students from paying attention to the numerical parameters, it can be more enjoyable for
context of the word problem, while these students them to explore structural relationships in the context
generally have difficulty with problem-solving of a word problem (Mason, 2018).
(Schwieger, 1999).
This study shares a series of explorations undertaken
Educators employ various approaches to help by a small group of third-grade students over several
students pay attention to and analyze the days. Using schematic representations and real-world
mathematical structure of a problem. For example, examples, this group discussed the mathematical
students are often encouraged to represent or relationships involved in the following story problem:
model the relationships in ways that allow them to “A father is 32 years old, and his son is 4 times younger
manipulate the quantities and reveal the structure, than him. How old will they be in 4 years?” The purpose
supporting their discovery of the required arithmetic of this study is to show the students’ exploration
operation. Some researchers supported schematic- process through three vignettes while providing
based instruction, claiming that schematic diagrams interpretational space for readers. In this retrospective
better serve students (e.g., Terwel et al., 2009). Several analysis of teaching episodes, this study focuses on
studies highlighted a conceptual correlation between the following questions: (a) What types of confusion
schemas and problem-solving (e.g., Jitendra & Star, and curiosity did the students exhibit while using
2011; Steele & Johanning, 2004; van Garderen et schematic representations to identify the additive and
al., 2013). Other researchers offered activities that multiplicative relationships? (b) How did the students
help students discern different word problem story make sense of the mathematical relationships
grammar (e.g., Xin, 2012). underlying schematic representations? (c) What kind
Previously, mathematics educators focused on of classroom culture should be established to support
teaching predefined schematic representations based student reasoning and justification?
on cognitive psychology (Fagnant & Vlassis, 2013).
They asked students to memorize several predefined Literature Review
representations to solve certain types of problems,
and students were expected to develop an ability to Problem Structure
categorize problems based on the representations
used (Schoenfeld, 1992). However, with increased Although the term “structure” has been widely
attention to sociocultural perspective (Cobb & Hodge, used in mathematics education without clear
2011) and mathematical process (National Council definitions, researchers consider knowledge about
of Teachers of Mathematics [NCTM], 2014), today’s structure as an awareness of a network of local and
mathematics educators are encouraged to provide general relationships (Venkat et al., 2019). Venkat et
students with opportunities to examine mathematical al. (2019) noted that emerging structures involving
structures of problems and represent them through analyzing, forming, and seeing local relationships
student-oriented investigations (Fagnant & Vlassis, can be observed when young students analyze and
2013). For example, NCTM (2014) suggests to “allow distinguish local relationships, ultimately allowing
students to select and discuss their choices to represent them to identify mathematical structures with more
the problem situations” (p. 28). When students gain general mathematical relationships and properties.
authority in their mathematics investigation, they can For young mathematics students, exploring the
make sense of connections between representations, different potential structures embedded in additive
understanding central mathematical ideas, and and multiplicative situations is a critical pathway for
experiencing authentic mathematical problem- developing students’ understanding and ability to
solving processes. Teachers should encourage operate within these structures flexibly (Mason, 2018).

270
 Elementary Students’ Exploration of the Structure of a Word Problem Using Representations / Lee & Hwang

Typically, curriculum using word problems includes Wright et al., 2006) frequently incorporate five or 10
multiple structures within additive and multiplicative frames to help students’ early numeracy knowledge
situations. It is common that students make additive by illustrating the structure of numbers and the place
errors in multiplicative missing-value word problems value concept. In the Singapore curriculum, various
and multiplicative errors in additive missing-value models, such as bar models, support students’ deeper
word problems. understanding in solving word problems (Kaur, 2019;
Ng & Lee, 2005, 2009).
Researchers highlighted that an important difference
between additive and multiplicative relationships is More explicit use of models in elementary
the nature of invariance (Behr & Harel, 1990; Degrande mathematics curriculum can be found in Davydov’s
et al., 2019). In other words, quantities are linked curriculum (Davydov et al., 1999), where the critical
additively in the additive structure, and the actual role of symbols and models is emphasized. In the
difference between quantities remains invariant. latter curriculum, students manipulate real objects
In contrast, the ratio (e.g., relative difference) of and graphic models such as line segments and
quantities linked multiplicatively (e.g., linked through schematics to represent implicit and explicit structural
multiplication and division), what is invariant is the relationships. As they progress, the use of concrete
ratio between quantities. objects and graphic models decreases, and the use
of symbolic formulas increases. For instance, physical
When considering the word problem at hand, we objects or graphic models of a part-whole relationship
can consider several structures.
help students initially see all involved quantities and
A father is 32 years old, and his son is 4 times younger their connection. Later, students can formulate
than him. How old will they be in 4 years? algebraic equations for this mathematical relationship
(Lee, 2002; Schmittau, 2005). Several studies reported
First, the given relation (“4 times younger”) supports the effectiveness of using various tools to represent
students in multiplicative reasoning. Thus, the son’s and visualize relationships between quantities when
current age is 8 because 32 divided by 4 is 8. Second, solving word problems (Kaur, 2019; Ng & Lee, 2005,
the question turns students to additive reasoning. 2009; Schmittau, 2005).
After finding the son’s age, students can find missing
values (“in 4 years”) through different strategies. As Word Problem and Schematic Representation
the examples below show, 4 years are added to the
current ages of father and son: Mathematics educators highlighted the importance
of word problems in learning mathematics (NCTM,
Father’s age in 4 years: 32 + 4 = 36 2000; van Garderen et al., 2013; Vula et al., 2017).
Son’s age in 4 years: 8 + 4 = 12 Word problems refer to problems that are “typically
composed of a mathematics structure embedded
Alternatively, noting the actual difference between in a more or less realistic context” (Depaepe et al.,
the current ages of father and son, 24, the final missing 2010, p. 154). Word problems help students construct
value is identified as follows: mathematical representations and understand
mathematical relationships and structures. They help
The difference between the current ages of father
them explore the relationship between reality and
and son: 32 – 8 = 24
abstract mathematical concepts and operations
Father’s age in 4 years: 32 + 4 = 36
(Jitendra, 2019). Studies showed that students usually
Son’s age in 4 years: 36 – 24 = 12 go through problem-representation and problem-
solution phases to solve word problems (Depaepe et
Thus, the situation can be explained differently al., 2010; Jitendra, 2019). In the problem-representation
depending on the relationships students recognize. phase, students comprehend the problem and
construct representations (or models) to illustrate
Representing the Problem Structure the problem situation clearly. However, students
work through the constructed representations in the
Researchers noted that students have difficulties in problem-solution phase and interpret and evaluate
understanding structures and analyzing quantitative the outcome.
relationships of word problems (Mason, 2018). Several
researchers highlighted the importance of visualizing In a well-known classification scheme for
and representing the problem contexts to support representation types, Lesh et al. (1987) emphasized
students’ attention and analyze the structure and flexibility and variability in meaningful use of
relationships underlying a problem. Therefore, in representations among contextual, visual, verbal,
mathematics curricula and programs, it is prominent physical, and schematic (or symbolic) representations.
to include various representations of a problem to elicit The visual representation retains most of the detailed
the structure and relationships within it. For example, information of the original contexts and clearly
materials used in the Math Recovery Program (e.g.,

271
 January 2022, Volume 14, Issue 3, 269-281

represents concrete visualization of objects to help culture should be established to best support students
students understand the problem contexts (Hegarty as they learn word problems. Therefore, further studies
& Kozhevnikov, 1999; Viseu et al., 2021). However, can be conducted to examine students’ exploration
schematic representations abstractly represent a of mathematical structures of word problems with
structural relationship of mathematical elements in a representations.
problem. As schematic representations are “meaning-
based representations” (Terwel et al., 2009, p. 27), Methods
they discard unimportant information and select
mathematically important relationships and structures Context and Participants
used in the problem-solving process. Therefore,
students are expected to convert verbal information The class episodes were taken from a three-year
into symbolic expressions, such as line, diagram, teaching experiment conducted in a private school
and shapes, and use them to construct arithmetic in the US (Lee, 2002). The first author taught a
operations during the problem representation phase. cohort of seven students using the first three years
Some studies reported that mathematics of elementary mathematics curriculum developed
educators often introduced predefined schematic by Davydov and his colleagues (Davydov et al.,
representations and asked students to memorize 1999). There were two male students and five female
those representations to solve word problems students. For five students, this private school was their
(Fagnant & Vlassis, 2013). However, findings of some first formal education setting, and two students had
studies revealed that solving word problems with some public school experience. There were three or
representations does not always increase students’ four mathematics classes per week, and each class
performance (Diezmann & English, 2001; Terwel et al., session lasted approximately 50–60 minutes. The
2009; Verschaffel et al., 2020). For example, Terwel et curriculum consists of a series of problems. Students
al. (2009) examined the effect of teacher-provided were accustomed to engaging in an in-depth
representations on solving word problems with fifth- discussion (or debate) on a small number of problems
grade students and reported minimal improvement each session.
in student problem-solving abilities. However, their
counterpart group, which was asked to construct The class was in the third year of the experiment
representations through collaboration, showed when discussing the word problem that this study
considerable improvement. As the reasons for these discusses. Prior to this discussion, the students were
different outcomes, the researchers explained that accustomed to using literal variables, while they had
the collaboration allowed students to improve their the freedom to refer to known or unknown quantities
understanding of problem structures and enhance using some tools such as question marks, blanks,
students’ capabilities to generate new problem-solving underlines, or verbal descriptors. These students were
strategies. Similarly, Lehrer et al. (2000) examined also accustomed to problems that were impossible to
elementary school students and found that student- solve due to insufficient or contradictory information.
generated representations were more beneficial Such problems aimed to facilitate the students’
for developing their conceptual competence than justification and reasoning process. The students
using teacher-sanctioned representations. However, called them trap problems (Lee, 2007). The students
these findings did not reveal that teachers should not were also familiar with using various representations
teach schematic representations to their students; such as line segments and schematic representations.
instead, it means that teachers should first give Students used self-invented schematic representations
students opportunities to learn and use predefined at times, but they usually used mutually agreed-upon
representations. Teachers should then allow students representations. Figure 1 shows some examples. As
to construct their schematic representation based on shown, students were encouraged to relate various
their understanding paired with thoughtful discussion relationships by analyzing the structure of the given
and analysis among classmates (Diezmann & English, schematic representations.
2001; Lehrer et al., 2000).
When the class episodes in the following section
Previous studies have largely adopted quantitative occurred, the students had already studied additive
research methods to examine the effect of employing and multiplicative relationships and analyzed various
schematic representations on students’ word contexts (word problems). In previous experiences,
problem-solving abilities. Thus, limited qualitative the problem contained only one relationship — either
information on what types of confusion and curiosity the additive or multiplicative relationship. Thus, the
is exhibited by students when using schematic invariance of difference or ratio was maintained.
representations. We also lack understanding of how The discussion presented in this study happened
students make sense of mathematical relationships when students needed to consider both additive and
underlying schematic representations, and there is multiplicative relationships in the same context.
little guidance on what types or aspects of classroom

272
 Elementary Students’ Exploration of the Structure of a Word Problem Using Representations / Lee & Hwang

Data Sources and Data Analysis important to examine their context. A descriptive case
study clearly describes a phenomenon and focuses
The primary data sources of the study were classroom on tracing “the sequence of interpersonal events
discourse and field notes that the first author over time (and describing) a subculture of it” (Yin,
documented after each class session, describing 2003, p. 4). For example, if researchers investigate the
interactions between the students, the teacher, and development of students’ interactions over time, they
among students. For this study, the authors focused can examine student participation and discourse, as
on the three days of class vignettes related to the well as their teachers’ roles and discourse, to get a
discussion on the given word problem. A descriptive complete picture of the classroom environments.
case study design (Yin, 2003) was used to examine
student challenges during word-problem solving, and The second author had an unbiased third-party
how students resolved those challenges through a role. As the first author was a teacher and thus
series of small group discussions. directly participated in the classroom interactions,
the second author also independently examined
A case study examined a few cases of a phenomenon the raw data. Then, the two authors collaborated
in a real context (Creswell & Poth, 2016). As individual during several online meetings to compare and
cases are strongly connected in space and time, it is discuss the interpretation of the raw data at hand.
Figure 1
Examples of Schematic Representations Students Used
Schematic Representation Example Related symbolic representations

x=b+c
Line segment for a part- b+c=x
whole relationship x-b=c
x-c=b

x=a+b
Schematic for a part-whole a+b=x
relationship x-a=b
x-b=a

Morgan is 10 years old. Jamie is 2 years older than Morgan.

How old is Jamie?
10 + 2 = ?
? - 2 = 10
? - 10 = 2

Schematic for the additive

relationship Moran is 10 years old. Jamie is 2 years
younger than Morgan. How old is Jamie?
10 - 2 = ?
10 - ? = 2
? + 2 = 10

Morgan is 10 years old. Jamie is 2 times older than Morgan.

How old is Jamie?
10 × 2 = ?
? ÷ 2 = 10
? ÷10 = 2

Schematics for the multi-

plicative relationship Morgan is 10 years old. Jamie is 2 times younger than
Morgan. How old is Jamie?
10 ÷ 2 = ?
? × 2 = 10
2 × ? = 10

273
 January 2022, Volume 14, Issue 3, 269-281

This process helped ensure the credibility of this Figure 2

study and provided the two authors the opportunity Initial Schematic Representation
to retrospectively analyze key learning instances
that occurred during the process (Yackel, 2001) from
insider and outsider perspectives. More specifically,
we took the following steps: (a) independently review
classroom discourse and descriptive field notes and
identify what important elements and moments,
(b) compare each other’s elements and moments,
(c) jointly identify the most salient themes, and (d)
juxtapose the data with interpretations based on
extant literature. In short, we allowed the analysis to
emerge from our understanding and interpretations Using the information provided, students were able to
of the events unfolding on the data, instead of find the values for all unknown quantities, as illustrated
approaching the data with a predetermined coding in Figure 2.
scheme. The reliability of the analysis is not obtained
by the coincidence of interpretations among us. • [Father now]: 32 years old
Instead, we directly presented classroom discourse to • [Son now × 4]: 32, [Son now]: 32 ÷ 4 = 8
increase the external validity and transparency of the • [Father will be]: 32 + 4 = 36
study (Creswell & Poth, 2016).
• [Son will be]: 8 + 4 = 12
Descriptions of Class Episodes
Meanwhile, the students found the values of all
The following three vignettes taken from the first unknown quantities and completed the problem.
author’s field notes show the students’ confusion However, one student, Jordan, started talking about
regarding the use of schematic representations, the relationship between Father will be (father’s age
differentiation between additive and multiplicative in 4 years) and Son will be (son’s age in 4 years), noting
relationships and the critical moments in which the that they did not show the relationship between these
students chronologically shifted their attention. The two quantities.
class episodes also show how students make sense of
the mathematical relationships underlying schematic Jordan: “Father (will be) is 4 times older than his
representations. In the teaching episodes, the teacher’s son (will be).”
role was minimal and focused only on facilitating the Chris: “The problem did not ask for that
discussion and recording (or helping record) students’ relationship.”
discussions in visual forms. All students’ names are
pseudonyms. Jordan: “However, we know that the
relationship between the father’s age and the
Vignette 1 son’s age stays the same.”

“A father is 32 years old, and his son is 4 times younger Although Chris resisted to do extra work (not due to
than the father. How old will they be in 4 years?” any mathematical reason), students agreed that they
could put “4 times” in the schematic representation
To solve this question, the student first determined the between the father’s age and son’s age in 4 years
known (father’s age now) and unknown quantities (Figure 3).
(son’s age now, father’s age in 4 years, and son’s
age in 4 years) and the relationships between given Figure 3
quantities. Based on that, the students drew the Examining the Unasked Relationship
following schematic representation to help class
discussion (see Figure 2). Students used descriptors
for the quantities instead of literal variables in the
schematic representations. In students’ terms, the
father’s age in 4 years was noted as “father will be”
and son’s age in 4 years as “son will be.” Unknown
quantities were noted using question marks. Students
noted the additive relationship between quantities
“by” and multiplicative relationship “times.”

274
 Elementary Students’ Exploration of the Structure of a Word Problem Using Representations / Lee & Hwang

When completing the problem, students realized that father will be and son will be cannot be connected in
something was wrong in the schematic representation this situation. This attempt was more likely to support
because the father will be in 4 years (36) was not 4 Chris’ argument in Vignette 1. While Chris argued
times greater than the son will be in 4 years (12). At that we did not need to find that relationship in the
this point, Chris again suggested deleting “4 times” problem, Morgan tried to show that it was not possible
between the father’s and son’s ages in 4 years. Chris to find the relationship by changing the shape of the
believed that we were not responsible for explaining schematic.
the relationship between father will be and son will be
by deleting the connection between them. Morgan: “We cannot connect father (will be) and son
(will be) now, so we cannot say that it is a trap.”

Other students disagreed with Chris, stated that the Chris: “But we can still figure out the relationship
relationship would still exist even after it was deleted. between father will be and son will be, and it should
be the same as the relationship 4 years ago.”
Then, they concluded that this was a trap problem due
to the contradictory information. Chris also agreed Morgan: “We don’t need it.”
that it was a trap problem because the relationship Chris: “We don’t need it, but we can do it. (Chris
between the father’s age and the son’s age should connected father will be and son will be and noted
stay the same in 4 years. However, he continued to the relationship as “4 times” as illustrated in Figure 5).
argue that there was no need to talk about this issue
Figure 5
as the question did not ask about this relationship.
Peers’ Reaction to Morgan’s First Attempt
Vignette 2

The discussion began with a review of previous day’s

conclusions as to why this problem was a trap:

The relationship between the father’s age and the

son’s age should remain the same even after 4 years.
The father is always 4 times older than the son.
However, here, 36 divided by 12 is 3, not 4. It does not
make sense. So, it is a trap.

While all students agreed on this conclusion made in

the session, one student, Morgan, changed her mind.
Morgan stated, “I changed my mind. I think it is not
a trap.” When asked to explain, Morgan redrew the
previous schematic representation as an attempt
to disconnect father (will be) from son (will be), as Morgan, then drew another schematic on the board,
depicted in Figure 4. as illustrated in Figure 6.

Figure 4 Figure 6
Morgan’s First Attempt to Change the Shape of the Morgan’s Second Attempt to Change the Shape of
Schematic Representation the Schematic Representation

Morgan: “Now, we cannot connect father (will be)

and son (will be).”
Jamie: “Still, we can connect them, and we can write
‘4 times’ there” (she added a line between father will
be and son will be, as shown in Figure 7)

Figure 7
Peers’ Reaction to Morgan’s Second Attempt

The students rejected Morgan’s conjecture that the

Morgan tried to justify her argument by changing the shape of schematic representation would make a
shape of the schematic representation, believing that difference. However, they could not find why the

275
 January 2022, Volume 14, Issue 3, 269-281

relationship between father will be and son will be Although Alex’s explanation was good and the
were not 4 times bigger or smaller. Thus, this continued students used the phrase “age difference,” they were
to be a trap problem. unable to connect the different explanations for this
problem and the original problem regarding the ages
Vignette 3 of father and son in 4 years. At this point, the teacher
encouraged the students to think about an additional
In this session, a student suggested taking a different problem in a similar context.
example. Alex tried another example using Morgan’s
age and the age of an adult in the classroom. We Teacher: “Alex’s explanation was very interesting.
drew a schematic on the board along with Alex’s Can we make another example? Can you use Fran’s
age and Mrs. L’s baby’s age this time?” (They all knew
explanation. that Fran was 10 years old, and Mrs. L’s daughter was
two years old.) What is the relationship between Fran
Alex: “Morgan, how old are you?” now and Mrs. L’s daughter now?”
Morgan: “11.”
Interestingly, this time, some students said, “by 8,” and
Alex: “Mrs. L., how old are you?”
some of them said “5 times.” The teacher wrote down
Mrs. L.: “32.” both relations in the schematic (Figure 10).

At this moment, the teacher asked the students how Figure 10

we could write this relationship in the schematic. The Teacher’s Variation Problem
students answered easily (by 21), and it was recorded
in the schematic (Figure 8).

Figure 8
Alex’s New Example

Teacher: “Let’s think about their ages in two years.”

Students: “Fran will be 12, and Baby will be 4.”
(The teacher put the numbers in the blanks as
illustrated in Figure 11).

Figure 11
Teacher’s Variation Problem: Expanded Version

Alex: “Morgan, how old will you be in 4 years?”

Morgan: “15.”
Alex: “How old will Mrs. L be in 4 years?”
Morgan: “36.”
Alex: “OK. What is the age difference between you
and Mrs. L in 4 years?”
Morgan: “21.” Teacher: “Can you tell me the relationship between
Fran will be, and Baby will be?
(Again, students wrote ‘by 21’ in the schematics.)
Students: “It will be the same.”
Alex: “See, the age difference did not change.”
(Figure 9) (The teacher put both – “by 8” and “5 times” – in the
schematic as illustrated in Figure 12).
Morgan: “Yes, but we don’t need to figure that out.”

Figure 12
Figure 9
Teacher’s Expanded Version with Additive and
Alex’s New Example: Expanded Version
Multiplicative Relationships Noted

276
 Elementary Students’ Exploration of the Structure of a Word Problem Using Representations / Lee & Hwang

Jamie: “Wait a minute. 12 minus 4 is 8, but 12 divided between the father’s and son’s ages would remain
by 4 is 3, not 5…” the same in 4 years. Two aspects are noteworthy.
Chris: “It is another trap!” First, the students were unable to identify the additive
(But this time, nobody agreed with Chris). relationship between the father’s age (now) and
son’s age (now) bound by the given multiplicative
Jordan: “No, it is not a trap.”
relationship between these two quantities.

Students talked about this situation for a while.

Second, the students seemed to overgeneralize the
Eventually, they concluded that the age difference
invariance of relationships. In previous experiences,
was the same, but it was not true for the times
the students only examined problems involving the
relationship. They returned to the original problem.
additive or multiplicative relationships; thus, the
Students said, “The age difference of 24 is the same
invariance of difference or ratio was preserved. In the
every year, but the “times” relationship is not like that.
problem context reported in this study, the students
had to think about both additive and multiplicative
Interpretations of Class Episodes
relationships. Perhaps, their past learning experiences
in techniques and language patterns triggered this
This section revisits each vignette and comments on
overgeneralized conclusion (Mason, 2003). What
several key points by juxtaposing the descriptive data
happened in this vignette is evidence of students’
from class episodes with our interpretations based on
unstable understanding of multiple relationships in
what we learned from other studies.
one context.
Comments on Vignette 1: Smooth Beginning, Doing
Comments on Vignette 2: False Attention to the Shape
Unasked Work, and a Trap!
of the Schematic Representations
Initially, the solution process was smooth. Examining
In this vignette, Morgan attempted to prove that
the given multiplicative relationship between the
the problem was not a trap problem because there
father’s age (now) and son’s age (now), students
should be no relationship between the father’s age
identified the son’s age (will be). Students then noted
and the son’s age in 4 years. In Vignette 1, students’
the additive relationships between current ages and
discussion was about “we don’t need to do it” vs. “we
ages in 4 years, which should have a 4-year difference.
still can figure it out.” In this vignette, Morgan tried to
Students’ recognition of these relationships resulted
explain that it was not possible to find the relationship
in incorrect answers. Jordan’s proposal to further
between ages in 4 years by changing the shape of
investigate the relationship between the father’s and
the schematic representations. Morgan attempted to
the son’s ages in 4 years completely changed the
separate these two quantities as much as possible,
way of the discussion. Two aspects were notable in
thinking that students could not “connect” them in
terms of students’ attitudes toward problem-solving.
the representations, and thus there should be no
First, students had discussions on doing unasked work.
relationship. Other students’ reactions to Morgan’s idea
Although there was some resistance to doing extra
(i.e., students were able to connect two quantities,
work, there was a consensus that they could do it.
regardless of the shapes Morgan created) eventually
Rather than using the “keyword method” or adopting
persuaded Morgan. Still, the idea that the problem
“number grabbing” approach (e.g., Littlefield & Reiser,
itself was a trap remained.
1993), being curious about unasked questions helped
focus more on the structure of the problem context.
Morgan’s inaccurate attention to the shape of the
Second, although there was no valid conclusion,
schematic representations revealed the students’
the students acknowledged the possibility that the
potential confusion, which had not been surfaced
problem may have incomplete or contradictory
before. At the same time, the overall discussion in
information. In the end, it could not be solved (i.e., a trap
Vignette 2 showed the flexibility of the students in
problem). These two aspects show different attitudes
using schematic representations. Ng and Lee (2009)
toward problem-solving from what Schwieger (1999)
reported in their study that some students seem to
points out as reasons for students’ difficulties with
treat the schematic models as an algorithm. However,
problem-solving: (a) unwillingness to bring anything
it was not an algorithm learned by rote; rather, it was
additional to the problem-solving process other than
a problem-solving heuristic that required students
one numerical, symbolic answer and (b) belief of a
to reflect on accurately depicting the information
singular solution and method for problem-solving.
presented in word problems. Similarly, in this vignette,
Morgan’s false attention provided an opportunity to
While students showed desirable attitudes toward
demonstrate students’ cognitive flexibility in using
problem-solving, this session ended with an incorrect
schematic representations.
mathematical conclusion. The entire group agreed
upon the erroneous conclusion that the multiplicative
relationship (“times” relationship in students’ words)

277
 January 2022, Volume 14, Issue 3, 269-281

Comments on Vignette 3: Reconstructing the Problem eventually recognize multiple relationships within the
particular problem context and general contexts.
Noting the importance of distinguishing between Additionally, the exploration revealed students’
additive and multiplicative situations, Mason (2018) unexamined assumptions about the use of schematic
suggested that encouraging students to make representations.
connections or develop problems with given
structural relations would be an important area for We particularly noted that there were several
further exploration. Students’ discussion in Vignette 3 instances where students themselves exhibited
appeared to align with this suggestion. Alex’s attempt intellectual perturbations (Harel et al., 2014) or
to restructure the problem using different quantities cognitive conflicts. Without such student-generated
helped students attend to the additive relationship. perturbations and conflicts, the proposed problem
Alex focused on the difference between the ages might have ended up as a computational problem.
because it was not immediately feasible to find the What if Jordan did not ask to find the unasked
multiplicative relationship between the selected question in Vignette 1? What if Morgan did not pay
two quantities (32 and 11). At this point, the teacher her false attention to the shape of the schematic
intentionally suggested two quantities (10 and 2) so representations in Vignette 2? What if Alex did not
that the students could identify both additive and suggest restructuring the problem using different
multiplicative relationships. Alex’s and the teacher’s examples? What if the teacher did not provide
attempts resonate with the notion of variation in strategic variation in the quantities to shift students’
structuring sense-making regarding tasks (Watson attention? Such unexpected questions helped the
& Mason, 2006). Both aimed to expose the target students focus on the structures and relationships
mathematical structure by strategically varying some rather than just performing calculations.
features of the problem while keeping other features. Consistent with previous studies (e.g., Lehrer et al., 2000;
Reconstructing the schematic representations using Terwel et al., 2009), the findings of this study revealed
these quantities (i.e., strategic variation) promoted that students could resolve their confusion through
students’ focus of attention and encouraged them collaboration. While the teacher did not ask students
to notice what was invariant in this context. The to use a particular schematic representation, they
students’ willingness to reconstruct the context with constructed and reconstructed the representations
the teacher’s purposeful support was helpful. through group discussion to reveal the difference
between the father’s and son’s ages concerning
Discussion and Implications additive and multiplicative relationships. Therefore,
students could identify important mathematical
The story problem used in this article can be quickly elements in additive and multiplicative word problems
solved using several steps of analysis and calculation. and explain structural relationship of these problems
However, exploring this word problem with schematic using schematic representations. These findings
representations took an unexpected path, resulting revealed that teachers might provide students with
in a much longer exploration than expected. Some mathematical tools to support their investigation,
may say that this is a failure of lesson planning and its reasoning, and justification.
enactment. Others may question whether it was worth
spending a long time discussing only one problem. These findings also highlighted the teachers’ roles in
While admitting that the presented class episodes solving complex word problems with representations.
in this study were atypical in terms of the duration of Using mathematical tools, such as representations,
the discussion, we saw the value of allowing such an alone could not guarantee students’ mathematical
atypical learning process to occur. learning (Lehrer et al., 2000). As Mason (2018) claimed
“what seems to matter most is not the apparatus
Regarding the mathematical content, the students’ itself,” (p. 332) but how teachers and students use
lengthy investigation was fueled by their initial them. If the apparatus is not used properly, its use
confusion and curiosity about the additive and might lead to rote learning. Therefore, teachers should
multiplicative relationships and the related invariant be cautious when using schematic representations in
and variant relationships. Considering the importance mathematics classrooms. For example, as shown in this
of constructing multiplicative reasoning for students’ study, teachers could first teach their student types of
learning of mathematics throughout the middle schematic representations that they could use and
grades and beyond (Zwanch & Wilkins, 2021), this explain the meanings of individual representations.
was a timely opportunity for students to think about Next, teachers could provide challenging problems
different relationships among quantities. Although it and ask students to justify their reasoning by presenting
took longer than planned, it was worthwhile because additional questions. These processes might arouse
it offered students a space to express their confusion, students’ curiosity and help them manipulate the
demonstrate their knowledge, test conjectures, quantities to reveal the mathematical structure of the
construct a similar but different problem context, and problem.

278
 Elementary Students’ Exploration of the Structure of a Word Problem Using Representations / Lee & Hwang

As such, teachers should create a mathematics Davydov, V. V. (1990). Types of generalization in

learning environment that allows students to instruction: Logical and psychological problems
investigate, reason, and justify (Depaepe et al., 2010; in the structuring of school curricula, Soviet
NCTM, 2000). As active investigators, teachers should studies in mathematics education. Volume 2.
believe in their students’ mathematical abilities and National Council of Teachers of Mathematics.
refrain from transmitting mathematical knowledge (Originally published in 1972)
and algorithms (NCTM, 2014). When teachers consider
their students as passive listeners, students are unlikely Davydov, V. V., Gorbov, S. F., Mikulina, G. G., & Savel’eva,
to present unasked questions and investigate them. O. V. (1999). Mathematics: Class 1. Unpublished
Moreover, teachers should respect students’ authority manuscript. State University of New York.
in learning mathematics and create a classroom
culture where all students’ answers are respected Degrande, T., Verschaffel, L., & Van Dooren, W. (2019).
(Cobb & Hodge, 2011). While some students gave To add or to multiply? An investigation of the
incorrect answers in this study, most students did not role of preference in children’s solutions of word
criticize their ideas, and the teacher did not directly problems. Learning and Instruction, 61, 60–71.
correct them. Instead, the students attempted https://doi.org/10.1016/j.learninstruc.2019.01.002
to justify their arguments using representations
and discussion, while the teacher supported their Depaepe, F., De Corte, E., & Verschaffel, L. (2010).
argumentation. Therefore, mathematics educators Teachers' approaches towards word problem
should be concerned with their classroom culture, solving: Elaborating or restricting the problem
particularly whether it facilitates or hinders students’ context. Teaching and Teacher Education, 26(2),
understanding of mathematical relationships and 152–160. https://doi.org/10.1016/j.tate.2009.03.016
structures in learning mathematics (Davydov, 1990;
Mason, 2003; Zwanch & Wilkins, 2021). Diezmann, C. M., & English, L. D. (2001). Promoting the use
of diagrams as tools for thinking. In A. A. Cuoco
The study has some limitations. Given that this study & F. R. Curcio (Eds.), The roles of representation in
examined a small group of students in a single school mathematics: 2001 yearbook (pp. 77–89).
classroom, the findings of the study could not be National Council of Teachers of Mathematics.
generalized to other contexts. Therefore, studies with
larger samples might yield more generalizable results. Fagnant, A., & Vlassis, J. (2013). Schematic representations
However, we hope that these classroom episodes in arithmetical problem solving: Analysis of
give teachers and teacher educators ideas to explore their impact on grade 4 students. Educational
more optimal learning environments for students to Studies in Mathematics, 84(1), 149–168. https://
raise awareness about the mathematical structure doi.org/10.1007/s10649-013-9476-4
and relationships in solving word problems.
Greer, B., & Harel, G. (1998). The role of isomorphisms
References in mathematical cognition. The Journal of
Mathematical Behavior, 17, 5–24. https://doi.
Behr, M., & Harel, G. (1990). Understanding the org/10.1016/s0732-3123(99)80058-3
multiplicative structure. In G. Booker, P. Cobb,
& T. N. de Merldicutti (Eds.), Proceedings of the Harel, G., Fuller, E., & Soto, O. D. (2014). DNR-based
14th Conference of the International Group instruction in mathematics: Determinants of a
for the Psychology of Mathematics Education DNR expert’s teaching. In Y. Li., E. Silver, & S. Li
Volume 3 (pp. 27–34). México: Consejo Nacional (Eds.), Transforming mathematics instruction:
de Ciencia y Tecnología, Gobierno del Estado Multiple approaches and practices (pp. 413–
de Morelos. 437). Springer. https://doi.org/10.1007/978-3-319-
04993-9_23
Cobb, P., & Hodge, L. L. (2011). Culture, identity, and
equity in the mathematics classroom. In E. Hegarty, M., & Kozhevnikov, M. (1999). Types of visual-
Yackel, K. Gravemeijer, & A. Sfard (Eds.), A journey spatial representations and mathematical
in mathematics education research (pp. 179– problem solving. Journal of Educational
195). Springer. https://doi.org/10.1007/978-90- Psychology, 91(4), 684–689. https://doi.
481-9729-3_11 org/10.1037/0022-0663.91.4.684

Creswell, J. W., & Poth, C. N. (2016). Qualitative inquiry Jitendra, A. K. (2019). Using schema-based instruction
and research design: Choosing among five to improve students’ mathematical word
approaches. Sage publications. problem solving performance. In A. Fritz, V.
G. Haase, & R. Pekka (Eds.), The international
handbook of mathematical learning difficulties
(pp. 595–609). Springer.

279
 January 2022, Volume 14, Issue 3, 269-281

Jitendra, A. K., & Star, J. R. (2011). Meeting the needs of Mason, J. (2018). Structuring structural awareness: A
students with learning disabilities in inclusive commentary on Chap. 13. In M. G. B. Bussi & X.
mathematics classrooms: The role of schema- H. Sun (Eds.), Building the foundation: Whole
based instruction on mathematical problem- numbers in the primary grades (pp. 325–340).
solving. Theory Into Practice, 50(1), 12–19. Springer Open. https://doi.org/10.1007/978-3-
319-63555-2_14
Kaur, B. (2019). The why, what and how of the ‘‘model’’
method: A tool for representing and visualising National Council of Teachers of Mathematics.
relationships when solving whole number (2000). Principles and standards for teaching
arithmetic word problems. ZDM: International mathematics.
Journal on Mathematics Education, 51, 151–168.
https://doi.org/10.1007/s11858-018-1000-y National Council of Teachers of Mathematics. (2014).
Principles to actions: Ensuring mathematical
Lee, J. (2002). An analysis of difficulties encountered success for all.
in teaching Davydov's mathematics curriculum
to students in a US setting and measures National Governors Association Center for Best
found to be effective in addressing them. Practices & Council of Chief State School
Publication No. 3051878 [Doctoral dissertation, Officers. (2010). Common core state standards
State University of New York at Binghamton]. for mathematics.
ProQuest Dissertations Publishing.
Ng, S. F., & Lee, K. (2005). How primary five pupils use
Lee, J. (2007). Context in mathematics learning: the model method to solve word problems. The
Problems and possibilities. Teaching Children Mathematics Educator, 9, 60–83.
Mathematics, 14(1), 40–44.
Ng, S. F., & Lee, K. (2009). The model method: Singapore
Lehrer, R., Schauble, L., Carpenter, S., & Penner, D. (2000). children’s tool for representing and solving
The interrelated development of inscriptions algebraic word problems. Journal for Research
and conceptual understanding. In P. Cobb, in Mathematics Education, 40, 282–313. https://
E. Yackel, & K. McClain (Eds.), Symbolizing and doi.org/10.5951/jresematheduc.40.3.0282
communicating in mathematics classrooms:
Perspectives on discourse, tools, and Savard, A., & Polotskaia, E. (2017). Who’s wrong? Tasks
instructional design (pp. 325–360). Lawrence fostering understanding of mathematical
Erlbaum. relationships in word problems in elementary
students. ZDM: International Journal on
Lesh, R., Post, T. R., & Behr, M. (1987). Representations Mathematics Education, 49, 823–833. https://
and translations among representations in doi.org/10.1007/s11858-017-0865-5
mathematics learning and problem solving. In
C. Janiver (Ed.), Problems of representations in Schmittau, J. (2005). The development of algebraic
the teaching and learning of mathematics (pp. thinking: A Vygotskian perspective. ZDM:
33–40). Lawrence Erlbaum. International Journal on Mathematics
Education, 37(1), 16–22. https://doi.org/10.1007/
Littlefield, J., & Reiser, J. J. (1993). Semantic features of bf02655893
similarity and children’s strategies for identifying
relevant information in mathematical story Schoenfeld, A. H. (1992). Learning to think
problems. Cognition and Instruction, 11, 133–188. mathematically: Problem solving,
https://doi.org/10.1207/s1532690xci1102_2 metacognition and sense-making in
mathematics. In D. A. Grouws (Ed.), Handbook
Mason, J. (2001). On the use and abuse of word of research on mathematics learning and
problems for moving from arithmetic to teaching (pp.334–370). Macmillan. https://doi.
algebra. In H. Chick, K. Stacey, J. Vincent & J. org/10.1177/002205741619600202
Vincent (Eds.), The future of the teaching and
learning of algebra (pp. 430–437). Proceedings Schwieger, R. (1999). Teaching elementary school
of the 12th ICMI study conference, University of mathematics: A problem solving approach.
Melbourne, Melbourne. Wadsworth Publishing.

Mason, J. (2003). Structure of attention in the Sierpinska, A. (1994). Understanding in mathematics.

learning of mathematics. In J. Novotná (Ed.), Falmer. https://doi.org/10.4324/9780203454183
Proceedings, international symposium on
elementary mathematics teaching (pp. 9–16), Steele, D. F., & Johanning, D. I. (2004). A schematic-
Charles University, Prague. theoretic view of problem solving and
development of algebraic thinking. Educational
Studies in Mathematics, 57(1), 65–90.
280
 Elementary Students’ Exploration of the Structure of a Word Problem Using Representations / Lee & Hwang

Stigler, J. W., Lee, S. Y., & Stevenson, H. W. (1990). Wright, R. J., Martland, J., Stafford, A. K., & Stanger, G.
Mathematical knowledge of Japanese, (2006). Teaching Number: Advancing children’s
Chinese, and American elementary school skills & strategies (2nd ed.). Sage Publications.
children. National Council of Teachers of
Mathematics. Xin, Y. P. (2012). Conceptual model-based problem
solving: Teach students with learning difficulties
Terwel, J., van Oers, B., van Dijk, I., & van den Eeden, to solve math problems. Sense Publications.
P. (2009). Are representations to be provided or https://doi.org/10.1007/978-94-6209-104-7
generated in primary mathematics education?
Effects on transfer. Educational Research Yackel, E. (2001). Explanation, justification and
and Evaluation, 15(1), 25–44. https://doi. argumentation in mathematics classrooms. In
org/10.1080/13803610802481265 van den Heuvel-Panhuizen (Ed.), Proceedings
of the conference of the international group
Thompson, P. W. (2011). Quantitative reasoning and for the psychology of mathematics education,
mathematical modeling. In L. L. Hatfield, 25th, Volume 1 (pp. 9 – 24). Freudenthal Institute.
S. Chamberlain, & S. Belbase (Eds.), New
perspectives and directions for collaborative Yin, R. K. (2003). Case study research: Design and
research in mathematics education, WISDOMe methods (3rd ed.). Sage Publication. https://doi.
Monographs Volume 1 (pp. 33–57). Laramie: org/10.3138/cjpe.30.1.108
University of Wyoming.
Zwanch, K., & Wilkins, J. L. M. (2021). Releasing the
van Garderen, D., Scheuermann, A., & Jackson, C. (2013). conceptual spring to construct multiplicative
Examining how students with diverse abilities reasoning. Educational Studies in Mathematics,
use diagrams to solve mathematics word 106, 151–170. https://doi.org/10.1007/s10649-020-
problems. Learning Disability Quarterly, 36(3), 09999-4
145–160. https://doi.org/10.1177/0731948712438558

Venkat, H., Askew, M., Watson, A., & Mason, J. (2019).

Architecture of mathematical structure. For the
Learning of Mathematics, 39(1), 13–17.

Verschaffel, L., Greer, B., & De Corte, E. (2000). Making

sense of word problems. Swets & Zeitlinger.

Verschaffel, L., Schukajlow, S., Star, J., & Van Dooren,

W. (2020). Word problems in mathematics
education: A survey. ZDM: International Journal
on Mathematics Education, 52, 1–16. https://doi.
org/10.1007/s11858-020-01130-4

Viseu, F., Pires, A. L., Menezes, L., & Costa, A. M. (2021).

Semiotic representations in the learning of
rational numbers by 2nd grade Portuguese
students. International Electronic Journal of
Elementary Education, 13(5), 611-624. https://doi.
org/10.26822/iejee.2021.216

Vula, E., Avdyli, R., Berisha, V., Saqipi, B., & Elezi, S. (2017).
The impact of metacognitive strategies and
self-regulating processes of solving math word
problems. International Electronic Journal of
Elementary Education, 10(1), 49-59. https://doi.
org/10.26822/iejee.2017131886

Watson, A., & Mason, J. (2006). Seeing an exercise as a

single mathematical object: Using variation to
structure sense-making. Mathematical Thinking
and Learning, 8(2), 91–111. https://doi.org/10.1207/
s15327833mtl0802_1

281