Chapter 4

Assessing Conceptual Understanding in Mathematics

with Concept Mapping
JIN Haiyue WONG Khoon Yoong
Mathematics educators and mathematics curriculum worldwide have
emphasised the importance of students ability to construct connections among
mathematics concepts (conceptual understanding) instead of just the
competence to carry out standard procedures in isolated ways. Education
researchers have used different techniques to assess this conceptual
interconnectedness in students minds. In this chapter, we discuss the use of
concept mapping as an assessment tool in mathematics instruction, including
different types of concept mapping tasks, training in concept mapping,
applications in classroom settings, and evaluation of student-constructed
concept maps. Concept mapping can be a worthwhile tool in teachers repertoire
of assessment for learning.
1 Introduction: What and Why of Concept Mapping
Cognitive psychologists have proposed that knowledge should be
interconnected, and acquiring knowledge with understanding is to make
meaningful connections between facts, concepts, and procedures. In
mathematics, the importance of interconnectedness among mathematical
concepts has been emphasized under the label conceptual understanding
(Kilpatrick, Swafford and Findell, 2001; National Council of Teachers of
Mathematics, 2000). For example, Van de Walle, Karp, and Bay-Willams (2010)
define conceptual understanding as the knowledge about relationships or
foundational ideas of a topic (p. 24), and these relationships are built from
underlying concepts that are meaningful to the students. The Singapore
mathematics syllabus (Ministry of Education, Singapore, 2006) highlights that
students should develop a deep understanding of mathematical concepts and
make sense of various mathematical ideas, including their connections and
applications; that is, students should see mathematics as an integrated whole
instead of isolated pieces of knowledge. Two principal issues to actualize this
curriculum goal are finding ways to help students make connections among what
they have learned and to assess their conceptual interconnectedness so that the
information can be used by teachers to plan lessons and provide remediation.
Education researchers have experimented with different techniques to assess
conceptual interconnectedness (White and Gunstone, 1992), and this chapter
addresses this assessment issue by focusing on concept mapping.
Concept mapping has gained popular use in science education over the
past three decades, and is now being studied by mathematics educators
(Afamasaga-FuataI, 2009). Figure 1 shows a concept map describing the
relations among seven concepts related to triangles. As this figure shows, a
concept map consists of three elements: (1) nodes representing concepts,
usually enclosed in ovals or rectangles, (2) links showing connections between
concepts, and (3) linking phrases specifying relationships between pairs of
concepts. The nodes can be mathematical concepts, examples and nonexamples of the concepts, diagrams, symbols, and formulas. The links are
usually directional to show subject-object, pre-post, cause-effect, top-down
hierarchy, or other relationships between the concepts. The linking phrases can
be verbs or adjective phrases. When two or more nodes are linked, statements
are formed, and these statements are called propositions. For example, in Figure
1, the connection between the concepts triangle and acute-angled triangle forms
the proposition triangle, when it has an acute angle, is an acute-angled
1

triangle (note that this proposition is only partially correct because all the
angles of an acute-angled triangle must be acute). The propositions form the
basic units of meaning in concept maps (RuizPrimo, 2004), although simpler
concept maps may not have linking
ADA TABEL

phrases (e.g., Orton, 2004), resulting in loss of information about the nature of
the links.
Why is concept map useful in assessing conceptual understanding?
Research suggests that the degree of a students understanding is determined
by the number, accuracy, and strength of connections (Hiebert and Carpenter,
1992; Resnick and Ford, 1981). Thus, a concept is well understood if it has
sufficient number of accurate and strong links with other related concepts. From
this perspective, a concept map can provide a visual representation of the
interconnected properties of the concepts held by the student.
Concept map was first developed by Joseph Novak and his team in the
1970s as a tool to document changes in understanding of a wide range of
scientific concepts held by students as they moved from first grade to twelfth
grade (Novak and Musonda, 1991). It is based on Ausubels (1963) assimilation
theory that states that learning takes place by assimilating new concepts and
propositions into existing knowledge framework or cognitive schema of the
learner (Novak and Caas, 2006). This psychological foundation justifies the use
of concept map as a tool to trace students conceptual changes over time. Over
the past three
decades, its use as an assessment technique has been
extensively investigated, especially in science education (Caas, et al., 2003). In
mathematics education, researchers and educators have also reported positive
findings concerning the use of concept map as an assessment technique at
different educational levels (see chapters in AfamasagaFuataI, 2009). The
following sections will discuss four aspects of this use.
First, several types of concept mapping tasks are explained to show that
different tasks may address different aspects of students conceptual
understanding. Second, training techniques for the mapping tasks are illustrated
with examples to help teachers plan such training when concept mapping is new
to their students. Third, four different classroom applications of concept maps
are discussed with examples, viz. to detect students prior knowledge, to
measure learning outcomes, to track learning progress, and to serve as a
learning strategy. Finally, several methods to evaluate student-constructed
concept maps are given so that the teachers can use the assessment information
to plan meaningful learning; this will align with the perspective of assessment for
learning.
2 Types of Concept Mapping Tasks
Concept mapping tasks can be categorized along a continuum of low to high
degree of directedness according to whether the four components, namely,
concepts, links, linking phrases, and map structure, are fully, partially, or not
provided (Ruiz-Primo, Schultz, Li and Shavelson, 2001). In high-directed concept
mapping tasks, most of these components are provided; thus, the tasks are
relatively easy for students to complete, but they are limited in measuring the
interconnected properties of students understanding. In contrast, in low-directed
concept mapping tasks, students have greater freedom to express their
understating of a topic using components that they construct on their own. In
this case, the openness of the tasks is more challenging to the students.
Some examples of commonly used concept mapping tasks from highdirected to low-directed ones are provided below.
2.1 High-directed concept mapping tasks: Fill-in-the-map
Fill-in-the-map tasks provide students with several concepts and require them to
fill in a skeleton map with these concepts. Figure 2 shows two different examples
of fill-in-the-map tasks: Fill-in-the-nodes and Fill-inthe-lines. Distracters may be
3

included to encourage students to think carefully about which items are relevant
to the map. The Fill-in-thenodes task in Figure 2 is an incomplete concept map
with two blank nodes. Four concepts are provided, two of which are distracters.
On the other hand, the Fill-in-the-lines task has two unlabelled links. Two linking
phrases are provided, with no distracter. The teacher has to decide whether or
not to include distracters, depending on the stage of learning and the students
ability. In either case, students fill in the blanks with what they think are the
correct items based on their understanding.
To design this type of mapping task, teachers either construct a concept
map themselves or use an expert-constructed map (for example, through
working with other teachers or mathematicians). Then remove some of the
concepts or linking phrases from the map and add distracters, if desired. This
type of concept mapping task is easy to administer and to grade, for example, by
counting the number of correctly filled items.
ADA TABEL
2.2 Semi-directed concept mapping tasks
When one or two of the above mentioned four components of a complete
concept map is missing and the other remaining components are fully or partially
provided, the concept mapping task is considered to be semidirected. Compared
with the high-directed concept mapping tasks, the semi-directed mapping tasks
require more efforts to complete.
In the semi-directed concept mapping task shown in Figure 3, only
concepts and linking phrases are provided. Students need to construct a concept
map including all the given concepts but only the relevant linking phrases. An
example of a possible concept map is also shown.
ADA TABEL
A variation of a semi-directed concept mapping task, illustrated in
Mansfield and Happs (1991), is to provide students with a partial list of concepts
of a particular topic, say, quadrilaterals. The most inclusive concept, in this case,
quadrilateral, is placed at the top of the map, with the other less inclusive
concepts, such as rectangle and rhombus, at lower levels, thereby requiring
students to consider hierarchy among the
2.3 Low-directed concept mapping tasks: Free-style mapping
Low-directed concept mapping tasks, also called free-style mapping, require
students to fully construct the maps based on the mapping topic or a list of given
concepts. They are free to express ideas in their own ways covering the four
components of a concept map. When only a mapping topic is given, students
need to first identify some concepts relevant to the topic and then construct the
map accordingly. For most school students, a concept list is usually given
because they may have difficulty in selecting the appropriate concepts. Some of
them may provide concepts that are somewhat related but not relevant or
essential to the topic (Jin, 2007). For example, they may include the concept
mathematics within the topic of functions: functions are very important for the
learning of mathematics. This kind of propositions does not directly address
students understanding about functions, and irrelevant concepts may even
distract students from constructing meaningful maps. A given concept list will
help them focus on a certain knowledge domain; at the same time, the task can
4

allow students to include additional concepts that they think are relevant to the
given ones.
The concept map in Figure 4 by a Singapore Secondary 3 student is an
example of a low-directed concept mapping task with a given list of concepts
about quadrilaterals. The student had used all the concepts provided without
adding new ones. The map was well constructed, with the most general concept
polygon located at the top, followed by less inclusive concepts quadrilateral,
parallelogram, and similar shapes at the middle levels, and the least inclusive
concept diagonals at the bottom.
ADA TABEL

Teachers may begin with high-directed mapping tasks and then move to
low-directed ones. This is because the high-directed mapping tasks are relatively
easy for students to complete. Furthermore, starting with easier tasks allows
time for both teachers and students to become familiar with the purposes and
construction of concept maps before they tackle the more challenging lowdirected tasks.
3 Training on Concept Mapping
Where concept map has not been extensively used in mathematics lessons, it is
necessary to train students on the techniques of constructing informative
concept maps. High-directed and semi-directed mapping tasks are, however,
quite straightforward and do not require extensive
Concept training. Thus, this section will focus on training in free-style mapping
tasks where a list of concepts is given. This is also most commonly used by
researchers and widely reported in the literature.
The following procedures have been developed based on the literature and
pilot studies conducted in Singapore and China (Jin and Wong, 2010).
1. Introduction: Teachers first provide students with a preliminary idea of
what a concept map is, what it is used for, and what its attributes are, i.e.,
nodes, arrowed links, and linking phrases.
2. Demonstrate with examples: Teachers begin with an example with four
or five concepts that students have already learned. First, read aloud the
concepts and help students to recall their meanings. Second, write the
concepts onto separate cards so that they can be easily moved around to
explore various connections that make sense. Concepts that are most
closely related are arranged near to one another. Third, once the intended
connections have been decided upon, identify the relations between each
pair of concepts and draw directed lines between them. Fourth, write on
each line the relationships identified so that propositions can be formed.
Finally, go back and check to see if any concept or relationship has been
left out; redraw the map if necessary.
3. Student practice: Provide students with a different set of concepts for
practice and remind them to pay attention to the following: a. All the
given concepts should be included in the map. b. In arranging the
concepts, make sure enough space is left for adding linking phrases. c.
The lines should be directed (with arrow) so that the relationships are
clear. d. All the lines should be labeled with linking phrases. e. The entire
map should be clear and legible.
4. Consolidation: After some practice, students should have mastered the
basic skills of concept mapping. Teachers should further encourage
students to include additional relevant concepts into their concept maps,
to construct as many relationships as they can between the concepts, and
to describe the relationships using informative, detailed linking phrases.
As the concept map is used as a graphical representation of students
conceptual understanding, they should show as much of their knowledge
as possible in their map so that it is rich enough to capture the essential
attributes of their conceptual understanding. With this the teachers can
obtain a better idea about what the students have already grasped and
which contents they are still weak in.

Some researchers emphasize the hierarchical nature of concept maps

because of Ausubels learning theory that more general, superordinate concepts
should subsume more specific, detailed concepts. In mathematics, in light of
Skemps (1986) Schema Theory, a concept map that shows the hierarchy of the
concepts is a more comprehensive representation of the interrelatedness among
mathematical concepts. However, strict requirement on hierarchy may distract
students from constructing meaningful connections, which is the main concern of
most mapping tasks at the school level. Besides, some school students might
have difficulty distinguishing or expressing the hierarchy of abstract
mathematical concepts (Schau and Mattern, 1997). Thus, for primary or
secondary school students, it is appropriate to encourage rather than require
them to construct concept maps with strong hierarchy.
4 Classroom Applications of Concept Map
This section covers four different but related applications of concept map in
classroom teaching and assessment.
4.1 Using concept map to detect students prior knowledge
The prior knowledge that students bring to their learning experience affects how
they encode and later retrieve newly learned information (Dochy, 1994). Concept
map has been used to find out about this prior knowledge (DiCerbo, 2007; Gurlitt
and Renkl, 2008) so that more effective lessons and materials can be prepared to
link prior knowledge to new learning.
For example, before teaching addition of two unlike fractions, teachers
need to know what their students have mastered about prior concepts such as
like fractions, unlike fractions, equivalent fractions, and so on. They may design
one type of mapping task to do so. Take the case of a low-directed concept
mapping task. The teacher can ask students to construct concept maps using a
list of concepts as given or selected by the students. Figure 5 shows two studentconstructed concept maps with the above three concepts about fractions.
Student A has displayed clearly the correct relationships among the three
concepts and included relevant numerical examples. By contrast, Student B has
not mentioned substantial relationships among the concepts; the only
relationship unlike fractions are different from like fractions is very brief and
general. Furthermore, the example for like fractions is wrong, equating the
numerators rather than the denominators. Thus, Student B will have difficulty
learning the new topic on unlike fractions, and some remediation is necessary.
ADA TABEL
4.2 Using concept map to evaluate learning outcomes
Concept mapping tasks can be assigned to students to assess their
understanding of newly-taught concepts. To avoid compounding conceptual
mapping with the new learning, it is better to begin with semi-directed concept
mapping tasks. Figure 6 is an example measuring students understanding after
learning numbers, modelled after Mansfield and Happs (1991). In this task, some
related concepts are given in boxes while others are omitted, with the boxes left
blank for students to fill in. The omitted concepts are provided on the right-hand
side, together with some distracters. The hierarchical positions of the concepts
are fixed as given. To some extent, these positions give hints to the appropriate
concepts for the blank boxes. For example, those who know the relationship
between composite number and prime number can deduce that the blank box
next to composite numbers should be prime numbers. In addition to the blank
7

boxes, spaces are provided for students to add linking lines and labels in their
own words.
ADA TABEL

With the given concepts, either in the boxes or in the list, and the fixed
hierarchical positions of the concepts, this task focuses students attention on a
particular domain. Thus, teachers have more control over what they are testing.
This task can be converted to the low-directed type, for example, by providing
only the mapping topic numbers or offering a list of concepts related to numbers.
As mentioned in the earlier section, this new task is more challenging for
students to complete and teachers to grade; yet, its openness allows students to
have greater freedom to express their understanding, thus, giving more valuable
information to the teachers.
4.3 Using concept map to track students progress in learning
Concept map has been used to track the changes in a students cognitive
structure and its increasing complexity as he or she integrates new knowledge
into existing cognitive structure. For example, Mansfield and Happs (1991)
reported their use of concept map as an expedient evaluation tool in a study of
parallel lines among a group of 12-year old students. In their study, the same
concept list consisting of eleven concepts was provided in the pre- and postconcept mapping tests before and after instruction of parallel lines. They cited
the concept maps drawn by the student, Bruce: he used only five concepts in the
pre-concept map but seven, adding one new concept, in the post-concept map.
Furthermore, the propositions in the post-concept map were more informative,
for example, revealing a misconception not found in the preconcept map.
Although the two maps were well-constructed, the omission of some given
concepts suggests that further teaching of these concepts is required for that
student. Thus, comparing concept maps constructed before and after instruction
can help teachers determine how much progress their students have made and
how effective the instruction has been. With the information drawn from the
comparison, teachers can then adjust their plans for future lessons.
4.4 Constructing concept maps as a learning strategy
Concept mapping has been widely promoted as a learning strategy to help
students elaborate on their learning and thereby to develop deep conceptual
understanding (Afamasaga-FuataI, 2006; Jegede, Alaiyemola and Okebukola,
1990). This popular application extends concept mapping beyond its use as a
formative or summative assessment tool.
Concept map can serve as a scaffold for students to organize their own
knowledge. At the end of a unit of instruction, a high-directed concept mapping
task can help students clarify the nature and number of connections among
newly learned concepts. A low-directed concept mapping task will encourage
them to reflect on the possible relationships among the concepts and represent
these relationships in a pictorial way. They may even see links that they are not
initially aware of (De Simone, 2007), thereby developing deeper understanding
about the concepts. They can modify these maps as learning progresses. This
constructive activity can be more effective when students work together in
groups to discuss their ideas and combine their knowledge in order to learn and
construct new knowledge (Gao, Shen, Losh and Turner, 2007). This group activity
will also provide opportunity for groups to compare their conceptual structure
with other groups and this will inspire further learning. In recent years, with the
support of concept mapping software such as CmapTools (Novak, 1998) and
SmartDraw (http://www.smartdraw.com), students can now build and discuss
their concept maps at distant locations (Novak and Caas, 2006) and flexible
times.
9

Several studies (e.g., Kankkunen, 2001; Mohamed, 1993) have reported

students positive attitudes toward using concept map in science. In
mathematics, we conducted a study with a class of Grade 8 Chinese students (n
= 48) in 2009. The students attitudes toward concept map were collected
through a questionnaire and interviews after their one months experience with
concept mapping. Most of them agreed that concept mapping was useful in
learning mathematics. They expressed moderate to high levels of enjoyment of
concept mapping even though, at the same time, some of them admitted that
concept mapping was challenging and required hard thinking. These findings are
encouraging for teachers who wish to explore this technique in their
mathematics lessons.
5 Evaluation of Student-Constructed Concept Maps
Concept maps can be assessed in a holistic, qualitative way based on expert or
teacher impressions or scored using specific criteria. These methods should
result in meaningful grades or scores so that judgement about the quality of
students conceptual understanding can be made from concept maps.
The following sections describe several quantitative methods to score
concept maps by examining the links between individual concepts and the
quality of the whole map. These scores can be used to assess students
performance on the given concept mapping task. It is not necessary to use all
the methods below for classroom assessment; however, some of these methods
may be used in action research.
5.1 Links between concepts
As defined in the introduction section, a concept map is a directed network. The
number of links, including incoming and outgoing ones, connected to an
individual concept reflects the extent to which that concept connects to all the
other concepts in the network, with higher number of links showing that it has
stronger connections with other concepts in the domain. This will reflect the
students conceptual understanding of that concept. In the extreme case, an
isolated concept with no incoming and outgoing link suggests that the person is
not familiar with the concept, cannot recall the link, or has simply forgotten to
construct connections with it (which can happen under timed test conditions).
For missing links, the teachers may need to interview the students to find out the
reasons behind their lack of conceptual connections about the concepts.

10

Tables 1 and 2 show the number of links to the concepts (excluding

examples) found in the respective maps in Figure 5. Both maps have included all
the three given concepts. Each concept in Student As map has more links
compared to the same concept in Student Bs map. If the links are also
mathematically correct, then Student A has a better understanding of the
fraction concepts than Student B.
Table 1
Number of links to concepts in Student As map in Figure 5
Concepts
Incoming
links
Outgoing
links
Total
links
Like fractions 2 0 2
Unlike fractions 1 1 2
Equivalent fractions 0 2 2
Table 2
Number of links to concepts in Student Bs map in Figure 5
Concepts
Incoming
links
Outgoing
links
Total
links
Like fractions 1 0 1
Unlike fractions 0 1 1
Equivalent fractions 0 0 0

11

In addition to the number of links for each concept, it is also informative to

examine the connections between pairs of concepts, in order to find out, for
examples, how close or far apart are the two concepts (called the distance), how
strong is the connection in terms of the number of direct and indirect paths
between them (called the connectedness), and the quality of the propositions.
The distance and connectedness measures indicate how easy or difficult it is for
the students to access the respective pair of concepts in their cognitive
structure. Knowing which pairs of concepts are cognitively far apart or weak
in their students conceptual understanding will alert the teachers to plan more
focussed activities that can strengthen these particular links. These quantitative
measures are related to various indicators used in social network analysis
(Degenne and Fors, 1999).
The distance between two concepts i and j is defined based on the length
of the shortest path that connects them in the direction from i to j. When there is
a direct link from i to j without any in-between concept, their distance d(i, j) = 1;
when i and j are not connected, either directly or indirectly, their distance is
defined as zero. Hence, it is possible to have d(i, j) = 1 and d(j, i) = 0. When
there is an indirect link from i to j, their distance equals to the number of inbetween concepts in the shortest path plus one. For any two connected
concepts, the larger their distance, the further apart they are, and the harder for
the students to connect them in their thinking. For the two maps in Figure 5, the
distance matrices are given in Tables 3 and 4 respectively. For these simple
maps, all the connections between pairs of concepts are direct ones (d = 1). It is
also evident that Student A has constructed more links than Student B.
ADA TABEL

12

The connectedness between any two concepts counts the number of

different paths between them. Referring to Student As map in Figure 5, the
connectedness from equivalent fractions to like fractions is 2 since there are two
paths, one direct (equivalent fractions like fractions) and one indirect (equivalent
fractions unlike fractions like fractions); the connectedness of the same pair of
concepts in Student Bs map is 0, since there is no path between them. Pairs of
concepts with large connectedness are quite robust; when one connection is
broken, these concepts still have high chances of being linked together. Tables 5
and Table 6 show the corresponding connectedness matrices for the maps in
Figure 5. The pairs of concepts in Student As map have stronger or more robust
relationships than those in Student Bs map.
ADA TABEL

13

The third measure of the relationship between each pair of concepts is the
quality of the proposition indicated by the linking phrase between them. A
simplified scoring scheme is as follows (see McClure, Sonak and Suen, 1999;
Novak and Gowin, 1984):
(1) when a proposition has no linking phrase or indicates misconception,
score 0;
(2) when a proposition indicates a relationship between the
connected concepts but with partially correct or incomplete linking
phrases, score 1;
(3) when a proposition indicates a correct and meaningful relationship
between the
connected concepts, score 2.
This scheme is illustrated in Figure 7 below. Proposition (a) indicates a
misconception of the relationship between square and parallelogram; thus, it is
scored 0. Proposition (b) correctly indicates the relationship that square is a
parallelogram; however, the linking phrase is very brief and gives no further
information about how or why a square is a parallelogram. Compared to
proposition (b), the proposition (c) is more detailed as it gives a complete
description of a square in relation to parallelogram. To apply the scoring scheme,
teachers should decide how much their students are expected to master the
relationship between two concepts at their stage of learning. If the students
propositions have met the expectation, then score 2; if the expectation is only
partially met, score 1. With this in mind, a teacher may score proposition (b) 1 or
2 according to his/her expectation. For formative assessment, pay attention to
the actual propositions, in addition to the scores, in order to identify students
good understanding as well as misconceptions.
ADA TABEL
The usefulness of the distance, connectedness, and quality of proposition
scores becomes apparent for more complex concept maps. These scores allow
more objective comparison between studentconstructed concept maps.

14

5.2 Nature of the whole map

Student-constructed concept maps can be compared at the whole map level. The
notion of density and the sum of all separate proposition scores can be used to
indicate the holistic properties of a concept map.
The density refers to the ratio of the total number of links to the total
number of concepts in a concept map. This provides information about how
compact the concepts are tied together within a particular group in the map.
There is no expected value of what the density of a concept map could
reasonably be, but compact maps are likely to have strongly intertwined
associations in the persons cognitive structure. For example, the densities of the
concept maps in Figure 5 are 3 3 = 1 and 1 3 respectively, excluding examples
and the links to examples. This suggests that Student As map is more compact
than Student Bs. Nevertheless, a higher density does not necessarily indicate
better quality of a map since students may simply draw links without considering
whether the links are substantial (meaningful) or trivial. The scoring of
propositions helps to show this differentiation. Hence, the sum of all separate
proposition scores is the second measure of the quality of the whole map. For
example, in Figure 5, Student As map obtained 2 + 2 + 2 = 6 points as an
overall proposition score since all the three propositions in the map are
substantial ones; while Student Bs map obtained 1 point since there was only
one partially correct proposition between the three given concepts. In general,
high proposition sums are associated with competent students who can provide
many valid content-based propositions, whereas low sums are associated with
weak students who do not provide many meaningful propositions.
Meaningful comparisons between concept maps can be made only if they
cover the same given concepts. This is because some concepts are less compact
than other concepts. As a consequence, a concept map constructed with such
concepts will have fewer expected connections and therefore lower density and
proposition scores. A different approach is to compare student-constructed
concept maps against a criterion (or expert) map. The criterion map can be
constructed by one or more teachers by taking into consideration the learning
objectives. Any gaps between student maps and the criterion map (for examples,
isolated concepts in student maps) and student misconceptions will highlight
where further teaching is to be focussed on. Some students may achieve higher
scores than the criterion map if they have constructed insightful connections
that the teachers have not thought about; indeed, this shows that teachers may
learn from their students.
6 Conclusions
This chapter has described three different types of concept mapping tasks that
can be used as alternative assessment to supplement traditional paper-andpencil tests, with concept maps highlighting the degree
of conceptual
understanding while traditional tests covering standard skills and problem
solving. Of these three types, student-constructed concept map is particularly
suited to measure individual conceptual understanding of a set of concepts. By
asking students to explicitly consider how and why concepts are linked, teachers
can detect students progress and gaps in understanding and then adjust their
instruction accordingly. At the same time, the concept mapping tasks provide
students important learning opportunity to reflect on what they have learned and
help them see links that they may have missed.
Training students to construct their own concept maps and interpreting
these maps are likely to be time-consuming. Some efforts may be saved by using
simple scoring as explained above. Even so, it is not economical in terms of
curriculum time to use concept maps for assessment purposes only. The studies
15

on concept maps in mathematics (Afamasaga-FuataI, 2009; Schau and Mattern,

1997) have provided strong evidence of the advantages of using concept
mapping as a teaching and learning tool, including letting students construct
group concept maps. Once concept map has been used for instruction or
learning, the burden of training for assessment will be reduced as students will
have become familiar with features of concept maps that will be assessed.
Hopefully, the students will also be receptive to the idea of using concept
mapping as part of standard assessment. This will change concept mapping from
assessment of learning to assessment as learning.
In summary, concept mapping is an assessment technique that can be
applied at various stages of learning. The increasing number of studies in recent
years on the uses of concept map in mathematics and other subjects suggests
that students can benefit from concept mapping. Its effects on learning are also
well-documented. Thus, it is worthwhile for teachers to develop their skills in
using this assessment technique and to explore its use in their mathematics
lessons to meet the curricular goals of promoting conceptual understanding in
mathematics.

16