The ASTR

Vol. Research Journal

4 · November 2020
PRINT ISSN: 2619-8428 • ONLINE ISSN: 2619-8436
International Peer Reviewed Journal
This journal is produced by the Asian Society of Teachers for Research, Inc.

Instrumental and Relational

Understanding in Math Problem Solving
Among Grade V Pupils
BELEN T. PUGOY
[Link]
belentejeropugoy@[Link]
DepEd, District of Compostela
Compostela, Cebu, Philippines

RENATO C. SAGAYNO
[Link]
renatosagayno@[Link]
University of Cebu
Cebu City, Philippines

CECILIA O. ARES
[Link]
fractalsares09@[Link]
University of Cebu
Cebu City, Philippines

JIGGER B. ABABON
[Link]
jiggerababon5@[Link]
University of Cebu
Cebu City, Philippines

STEVEN Y. RAZONABLE
[Link]
[Link]@[Link]
University of Cebu
Cebu City, Philippines

24
 Volume 4 · November 2020

Grammar Test: 98/100

Originality: 99/100
Gunning Fog Index: 10.97
Flesch Reading Ease: 52.66

ABSTRACT

In teaching mathematics, the essential task of the teachers is to motivate

their students to understand and appreciate the concepts and usefulness of
mathematics. This study determined the mathematics performances of the Grade
V pupils in the context of Skemp’s (1976) relational and instrumental concepts
of understanding. The study utilized the descriptive-comparative method of
research using a researcher-made test to gather data on the performances of the
Grade V pupils in instrumental and relational understanding types of tests. Most
of the pupils involved in the study failed in the instrumental type of test and
more so in the relational type of test. The majority of schools manifest significant
differences in the pupils’ performance in the instrumental and relational types
of tests. Based on the mean scores, the pupils’ performances in the relational
type of test is more deficient than the instrumental type of test. The analysis
of variance and the post hoc test supports the idea that there is a variation in
pupils’ performances in the instrumental type of test. Likewise, a variation in the
relational type of test exists.

Keywords: Mathematics, problem-solving, instrumental understanding,

relational understanding, teaching and learning, descriptive-comparative research,
Philippines

INTRODUCTION

Mathematics teaching today rests on the assumption that teachers are

important figures in improving the teaching and learning of mathematics (Ferri,
2018). Consequently, the teacher must have long-term support and adequate
resources to engage the students effectively in the teaching-learning process
(Liu, Chen, Lin, & Huang, 2017). Teachers’ primary concern in the classroom is
to deliver the lesson. Motivating the students to understand and appreciate
various mathematical concepts is considered an essential task of the teachers
(Simamora & Saragih, 2019). Hence, teachers need to innovate (Wang, Utemov,
Krivonozhkina, Liu, & Galushkin, 2018) and make their examples as realistic

25
 The ASTR Research Journal

as possible. Relatable and straightforward math problems would make the

students enjoy the topic and learn more (Clarke & Roche, 2018; Menanti, Sinaga,
& Hasratuddin, 2018). Although some students understand the lessons and do
well, they hate the subject because they cannot see real-life applications. For
them to love math is by making it practical and relatable (Batool, 2019).
Exposing young learners to math word problems would enhance their
potentials in choosing and applying varied strategies in handling real-life
problems (Dröse & Prediger, 2019). The working memory capacity of children
with difficulties in learning mathematics constitutes a moderating effect of the
selected teaching strategy. The advantages of children with high functional
memory capacity on performing better than their counterparts depend on the
type of teaching strategy that motivates them and catches their attention. Some
teaching strategies do not yield excellent outcomes. However, those strategies
which focus on capturing the attention of the pupils to engage in the lesson
benefitted both learners with high and low working memory capacity (Swanson,
2016).
Ideally, mathematics classrooms should offer students an environment with
equal opportunity to learn, a balanced focus on conceptual understanding, and
procedural fluency (Smith & Freels, 2017). It encourages the students to engage
actively in problem-solving, reasoning, communicating, making connections
(Santos-Trigo & Reyes-Martínez, 2019), and using multiple representations
(Kang & Liu, 2018). Mathematics classrooms should also be technologically
well-equipped learning centers. Technology is readily accessible to enhance
understanding, aside from incorporating various assessments compliant with
the instructional goals and practices. As a result, the students will manifest a
good performance (Young, 2017).
A study conducted in New Zealand advocated using challenging tasks
as an innovative approach to teach mathematics. The researchers believe
that students learn mathematics best when they build connections between
mathematical ideas for themselves. In challenging mathematics tasks, students
are encouraged to think for themselves, discuss mathematics, determine
appropriate strategies, and establish connections. The goal of the approach is
to bring out the students from the habit of applying mathematical concepts and
rules without deeply understanding the problem (instrumental approach). As a
result, the students would engage in a situation where they have to seek a more
profound understanding and connections of the various mathematical concepts
to a specific problem (relational approach) (Ingram et al., 2019).
In educational psychology, personalizing problems is an effective teaching
strategy. Personalized questions can elicit the students’ interest to engage and re-
engage with specific ideas (Bernacki & Walkington, 2018). In the cognitive view,

26
 Volume 4 · November 2020

learning is extending and transforms existing knowledge. On the older beliefs

of cognition, the emphasis is on knowledge acquisition. However, the newer
approaches to understanding emphasized knowledge construction. Among
constructivists, learning is on knowledge in use rather than the storing of inert
facts, concepts, and skills. Learning objectives include developing abilities to find
and solve ill-structured problems, critical thinking, inquiry, self-determination,
and openness to multiple perspectives (Wolfolk, 2016). However, several math
teachers indicated that their students struggled with representation and
understanding math problems. The causes of students’ difficulties include text
difficulties, unfamiliar contexts, and using inappropriate strategies. Embedding
math word problems in a more familiar context would turn it more comfortable
for students (Macdonald & Banes, 2017).
The result of the examination conducted by Trends in International
Mathematics and Science Study (TIMSS) in 1999 showed that the Philippines
was ranked 36th in mathematics and science among 38 participating countries.
In TIMMS 2003, the Philippines was listed in 23rd place among 25 participating
countries in Grade 4 science and mathematics. Furthermore, the country was in
42nd place in science and 41st in mathematics among 45 countries for second-
year high school (Gonzales et al., 2004). In March 2019, the Philippines participated
again in the said survey. Based on the results, the overall performance of Filipino
Grade 4 pupils is ranked the lowest among fifty-eight countries that participated
in the said international assessment (Baclig, 2021).
Literature search reveals that there are studies that focus on specific math
concept (Anwar, Yuwono, & As’ari, 2016; Utomo, 2020), the tendency of students
to adopt instrumental approach (Anderson, 1996), and exposing students
to relational instruction compared to students who received instrumental
instruction followed by relational instruction (Pesek & Kirshner, 2000). However,
the researchers cannot find a study conducted regarding instrumental and
relational understanding in math problem solving among grade school pupils
from public schools in the Philippines. Furthermore, the researchers cannot
find a study of the variation of the pupils’ performances in the instrumental and
relational types of tests analyzed in the context of geographical setting.
In the School District of Compostela, located 31 kilometers north of Cebu
City, Philippines, the average performance of the pupils in the mathematics
component of the Philippines’ National Achievement Test (NAT) is consistently
below the national passing mark of 75% for the last five years. Such observation
became the driving force in conducting this study by looking into the mathematics
performances of the Grade V pupils in the context of Skemp’s (1976) relational
and instrumental concepts of understanding. Furthermore, this study holds on
to the theories of cognitive dissonance (Festinger, 1957) and cognitive flexibility

27
 The ASTR Research Journal

(Spiro, Coulson, Feltovich, & Anderson, 1988). Hopefully this study would provide
additional information on the existing literature on students’ mathematics
performances and encourage other researchers to conduct further studies on
this topic.

FRAMEWORK

Skemp (1976) stated that there are two different approaches to understanding:
instrumental and relational understandings. Both types of understanding give
the correct answers, but relational is much more extensive. Moreover, relational
understanding is the one considered the better option over the other. However,
there are advantages and disadvantages to both. The favorable outcomes for
the first type of understanding are the disadvantages of the other. Instrumental
understanding would push the learners to have a mathematical rule and able
to use or manipulate it. In short, the learning process involves knowing and
applying the rule. As a result, the learner has to remember many separate rules
that seem unconnected from each other. Relational understanding would not
only push the students to learn the mathematical rule and its uses; but also
enable them to know why a rule works and connects with another rule. Many
teachers teach instrumental mathematics because it is usually easier for students
to understand; the rewards are more immediate and apparent. One can often
get the correct answer more quickly and reliably. However, teaching through
relational understanding will make students more adaptable to new tasks; they
can remember the concept easily; can rationalize knowledge effectively as a goal
in itself, and relate to schemas in organic quality. However, a teacher might make
a reasoned choice to use instrumental understanding in teaching mathematics
for self-convenience.
Meanwhile, Festinger’s (1957) theory of cognitive dissonance asserts that
in any given situation where two cognitions are inconsistent, one person tends
to seek consistency in his/her beliefs and attitudes. As a state of unpleasant
psychological tension, cognitive dissonance motivates a person to reduce his/
her mental inconsistencies by making his/her views more consistent with his/
her behavior. As explained by Baumeister and Bushman (2017), the theory of
cognitive dissonance centers on one’s effort to reach internal consistency.
People’s inner needs guarantee the character of their beliefs and views.
Disharmony is the result of inconsistent or conflicting beliefs in which people
strive to avoid. Cooper (2007) stressed that too personal cognition might lead
to more significant dissonance. Belief in highly valued things would lead to
more substantial disharmony. The proportion of dissonant thoughts against
consonant thoughts could play a role in how strong the feelings of dissonance

28
 Volume 4 · November 2020

are. The higher the strength of the dissonance, the more pressure there is to
relieve the feelings of discomfort. The three key strategies to reduce or minimize
cognitive dissonance include the following: 1) focusing one-self on more
reassuring thoughts that outweigh the dissonant belief or behavior, 2) reducing
the importance of contradicting belief, and 3) amending the different trust to
be consistent with other beliefs or behavior. A person experiencing dissonance
has three optional courses of action to minimize the conflict: alter the behavior,
amend the belief, or give reasons for the behavior. However, people tend to
either change their beliefs or rationalize. The motivation to reduce dissonance
may cause irrational or even dangerous behavior.
Additionally, cognitive flexibility theory concerns transferring desired
knowledge and skills to any learner beyond the initial learning situation. Its
emphasis is on presenting information from multiple perspectives and the use
of many case studies that present diverse examples. It also asserts that active
learning is context-dependent, so instruction needs to be very specific. On
the importance of constructed knowledge, the development of the learners’
representation is the prime intention. Complex and ill-structured domains are
the foci of the cognitive flexibility theory. The flexibility of cognition refers to the
natural ability to restructure knowledge in various ways and adapt to varying
situations (Spiro et al., 1988).
Further, cognitive flexibility theory supports the use of interactive
technology. In principle, this theory says that learning activities must provide
multiple representations of content, avoid oversimplifying the content domain,
and support context-dependent knowledge of the instructional materials.
Instruction should be case-based and emphasize knowledge construction,
non-transmission of information, and knowledge sources should be highly
interconnected rather than compartmentalized. Manipulating the way of
presenting information and its corresponding processes leads to the cognitive
flexibility of the learner (Spiro & Jehng, 2012).

OBJECTIVES OF THE STUDY

The study determined the 1) level of performance in Mathematics of the

Grade V pupils in instrumental and relational understanding type of test; 2) the
significance of the differences between the pupils’ performances in instrumental
and relational understanding types of tests per school; 3) the significance of the
difference between the pupils’ performances in the instrumental and relational
type of tests per school; and 4) the significance of the difference among the
pupils’ performances in the instrumental and relational understanding type of
tests when grouped in terms of school.

29
 The ASTR Research Journal

METHODOLOGY

The study utilized the descriptive-comparative method of research using a

researcher-made test to gather data on the performances of the Grade V pupils
in instrumental and relational understanding types of tests. The locale of the
study is the School District of Compostela - Compostela, Cebu, Philippines. The
School District of Compostela is 31 kilometers away from Cebu City, going to the
northern part of Cebu Province. Compostela District is composed of 15 complete
elementary schools. The research subjects were the Grade V pupils enrolled in the
public elementary schools in the District of Compostela, Cebu. The total number
of Grade V pupils in the district is 1,127. From this number, 498 pupils composed
the research subjects of this study. The sampling technique used is random
cluster sampling. One class of Grade V pupils from each school was represented
in the survey. The sample size of 498 Grade V pupils represented 51% of the total
population of Grade V pupils. The central limit theorem advances that a sample
size of 30 or more is considered a large sample. Hence, the sample size of this
study is already more than sufficient.
A researcher-made test, prepared based on the math competencies for
Grade V pupils, was used to measure the respondents’ level of performance in
Mathematics using the instrumental and relational understanding approach.
Experts in test construction and other math teachers validated the researcher-
made tests. After incorporating the suggestions of the validators, the primary
researcher subjected the teacher-made tool to pilot testing in one of the
Grade V classes in the locality. The pilot testing indicates that both types of
researcher-made tests passed the reliability test using Cronbach’s alpha. The
instrumental type of researcher-made test got a reliability coefficient of 0.81,
while the relational kind of test got a reliability coefficient of 0.78. In this study,
the researchers utilized the following score ranges and categories of DepEd to
summarize the scores of the research subjects; as follows: 0 – 74% described as
beginning, 75 – 79% described as developing, 80 – 84% described as approaching
proficiency, 85 – 89% described as proficient, and 90 – 100% described as
advanced. The researchers utilized the frequency count, proportion, t-test for
paired two samples for means, and the one-way analysis of variance (ANOVA) in
summarizing, analyzing, and interpreting the data.

30
 Volume 4 · November 2020

RESULTS AND DISCUSSION

Pupils’ Level of Performances in Mathematics

Table 1. Pupils’ performances in the instrumental understanding type of test

Approaching
Advanced Proficient Developing Beginning
Proficiency Total Mean Rank
(90-100) (85-89) (75-79) (74 & below)
(80-84)

School 1* 0 23 13 8 7 51 80.88 2
School 2** 0 0 0 0 35 35 45.71 8
School 3** 0 2 1 0 22 25 61.00 4
School 4* 7 13 6 11 5 42 82.87 1
School 5* 0 2 6 3 27 38 58.19 6
School 6* 0 0 0 1 34 35 52.06 7
School 7** 0 0 0 0 41 41 44.11 11
School 8* 0 0 0 1 39 40 45.00 10
School 9** 0 0 0 0 18 18 42.75 12
School 10** 1 8 6 7 27 49 71.03 3
School 11** 0 0 0 1 23 24 42.59 13
School 12** 0 0 1 5 19 25 60.89 5
School 13** 0 0 0 0 17 17 39.71 15
School 14* 0 0 0 0 32 32 45.31 9
School 15** 0 0 0 0 26 26 39.96 14
Grand
8 48 33 37 372 498 56.89
Total
% 1.61 9.64 6.63 7.43 74.70 100.00
*lowland school; **upland school

Table 1 shows that the pupils’ general performance from the fifteen
public schools in the instrumental type of test is categorized as beginning
proficiency (74.70%) with an overall average score of 56.89. Moreover, schools
#4 (rank 1) and #1 (rank 2) are the only two schools whose pupils’ performances
in the instrumental type of test are categorized as proficient with average scores
of 82.87 and 80.88. These school (#4 and #1) are located in the town proper.
Although most pupils’ performances of the remaining schools are categorized
as beginning proficiency, school #10 (rank 3) has some pupils belonging to the
higher performance categories with an average score of 71.03. This school
is located along the national highway and two kilometers going north from
the town proper. Further, the remaining schools are categorized as beginning

31
 The ASTR Research Journal

proficiency with average scores of less than 70.00. As indicated in Table 1, the
top three schools (#4, #1, and #10) are situated in the lowland areas. Meanwhile,
the five least performing schools are #7 (rank 11), #9 (rank 12) , #11 (rank 13),
#15 (rank 14), and #13 (rank 15). These schools are all situated in remote upland
areas. The findings indicate that pupils of schools situated in the town proper
and those in school along the national highway have better performances than
their counterparts in remote upland areas in the instrumental type of test. It
could also mean that the quality of mathematics teaching in the schools situated
in the town proper and schools along the national highway is better than the
schools situated in the remote upland areas.

Table 2. Pupils’ performances in the relational understanding type of test

Advanced Proficient Approaching Developing Beginning Total Mean Rank
(90-100) (85-89) Proficiency (75-79) (74 &
(80-84) below)

School 1* 0 0 5 14 32 51 70.86 2
School 2** 0 0 0 0 35 35 36.92 12
School 3** 0 0 0 0 25 25 45.89 7
School 4* 0 3 11 17 11 42 76.72 1
School 5* 0 1 0 2 35 38 49.56 4
School 6* 0 0 1 2 32 35 48.03 6
School 7** 0 0 0 0 41 41 36.65 13
School 8* 0 0 0 0 40 40 34.53 15
School 9** 0 0 0 0 18 18 37.50 11
School 10* 0 0 2 4 43 49 63.32 3
School 11** 0 0 0 1 23 24 42.36 8
School 12** 0 0 0 1 24 25 48.22 5
School 13** 0 0 0 0 17 17 39.56 10
School 14* 0 0 0 0 32 32 35.78 14
School 15** 0 0 0 1 25 26 40.92 9
Grand Total 0 4 19 42 433 498 49.42
% 0.00 0.80 3.82 8.43 86.95 100.00
*lowland school; **upland school

Table 2 shows that the pupils’ general performance from the fifteen public
schools in the relational type of test is categorized as beginning proficiency (86.95%)
with an overall average score of 49.42. Moreover, school #4 (rank 1) is the only
school with pupils› performances categorized under developing proficiency and

32
 Volume 4 · November 2020

approaching proficiency with an average score of 76.72 in the relational type of

test. As mentioned earlier, school #4 is located in the town proper. Although
most pupils’ performances are categorized as beginning proficiency, schools #1
(rank 2) and #10 (rank 3) have some pupils belonging to the higher performance
categories with average scores of 70.86 and 63.32. School #1 is situated in the
town proper, and school #10 is located two kilometers from the town proper
but along the national highway. Further, the remaining schools are categorized
as beginning proficiency with average scores of less than 60.00. Meanwhile, the
five least performing schools are #9 (rank 11), #2 (rank 12), #7 (rank 13), #14 (rank
14), and #8 (rank 15). Schools #9, #2, and #7 are upland schools, while schools
#14 and #8 are lowland schools. As indicated, only one school from the town
proper has performed better in the relational type of test, and the rest have mean
scores of less than 75 passing marks. As observed, schools #14 and #8 (lowland
schools) are the least performing schools based on the ranking. This observation
is fascinating considering that these schools are in the lowland area but far from
the town proper and national highway. Other possible factors may have caused
this observation but were not captured in this study. Hence, a good material for
future researches. Nevertheless, the overall findings indicate that the pupils have
more difficulty in the relational type of test. In other words, the majority of the
pupils in the participating schools have difficulty answering the mathematical
questions written in relational form. Hence, this is a manifestation of their low
ability to master the lessons in mathematics.

Differences Between the Pupils’ Performances in the Instrumental and

Relational Type of Tests per School

Table 3. Significance of the difference between the pupils’ performances in the

instrumental and relational type of tests per school
Instrumental Relational Decision
School df t-stat t-crit Significance
(Mean) (Mean) on Ho
School 1* 80.88 70.86 50 8.475 2.009 Reject Ho Significant
School 2** 45.71 36.92 34 5.282 2.032 Reject Ho Significant
School 3** 61.00 45.89 24 6.159 2.064 Reject Ho Significant
School 4* 82.87 76.72 41 4.670 2.020 Reject Ho Significant
School 5* 58.19 49.56 37 4.415 2.026 Reject Ho Significant
School 6* 52.06 48.03 34 2.342 2.032 Reject Ho Significant
School 7** 44.11 36.65 40 3.783 2.021 Reject Ho Significant
School 8* 45.00 34.53 39 5.917 2.023 Reject Ho Significant

33
 The ASTR Research Journal

Instrumental Relational Decision

School df t-stat t-crit Significance
(Mean) (Mean) on Ho
Failed to
School 9** 42.75 37.50 17 1.598 2.110 Not Significant
Reject Ho
School 10* 71.03 63.32 48 5.335 2.011 Reject Ho Significant

Failed to
School 11** 42.59 42.36 23 0.107 2.069 Not Significant
Reject Ho
School 12** 60.89 48.22 24 6.262 2.064 Reject Ho Interpretation

Failed to
School 13** 39.71 39.56 16 0.133 2.120 Not Significant
Reject Ho
School 14* 45.31 35.78 31 5.660 2.040 Reject Ho Significant

Failed to
School 15** 39.96 40.92 25 -0.400 2.060 Not Significant
Reject Ho

*lowland school; **upland school; α = 0.05

Table 3 shows that eleven schools have test statistics greater than the critical
values at a 0.05 level of significance. Based on their means, the mean scores in
the relational type of test are lesser than the mean scores in the instrumental
type of test. The results imply that the pupils’ performances in the relational
type of test are significantly lesser than their instrumental type of test. In other
words, the pupils of schools #1, #2, #3, #4, #5, #6, #7, #8, #10, #12, and #14 have
more difficulty in answering the relational type of test. Schools #2, #3, and #7 are
located in the upland areas, while the rest are located in the lowland areas. These
findings indicate that regardless of the geographical location of the schools,
students who failed to master the various mathematical concepts and operations
would have difficulty answering the instrumental type of test and more difficulty
in the relational type of test. This scenario is supported by (Macdonald & Banes,
2017) when they said that many math teachers indicated that their students
struggled with representation and understanding the math problems. They
stressed that pupils’ difficulties in mathematics center on understanding the text
used in the word problems and unfamiliarity of the contexts used.
Meanwhile, out of fifteen schools, only four schools have test statistics less
than the critical values. It means that there is no sufficient evidence to say that
the mean scores in the instrumental type of test are different from the mean
scores in the relational type of test. Hence, the pupils’ performances from
schools #9, #11, #13, and #15 in the instrumental and relational types of tests
are relatively similar. As indicated in Table 4, these schools are situated in the
upland areas, and the mean scores are far below the 75 passing marks. These
findings indicate that the students in these four upland schools may have just
guessed in answering the two types of tests. As a result, their scores in the

34
 Volume 4 · November 2020

instrumental and relational types of tests are deficient but do not reflect that the
relational type of test is more complex than the instrumental type of test. When
this group of pupils was answering the tests, their thought is to pass the tests.
However, their poor ability and non-mastery of the mathematical concepts and
operations give them difficulty answering the tests. As a result, they resort to
guessing the answers to complete the tests. The theory of cognitive dissonance
(Festinger, 1957) can explain this situation. In the theory of cognitive dissonance,
people tend to seek consistency of their beliefs and attitudes when caught in a
situation of two conflicting cognitions. Cognitive dissonance gives unpleasant
psychological tension, motivating them to reduce their mental inconsistencies
by doing things more consistent with their behavior.
Based on the findings, one can say that the pupils involved in this study
are more exposed to the instrumental understanding teaching approach than
the relational understanding approach. The decreasing trend of the mean
scores between the instrumental and relational types of tests is a manifestation
of pupils’ less exposure to the relational understanding teaching approach.
The finding validates Skemp’s (1976) explanation that many teachers teach
mathematics in the instrumental understanding way because it is easier to
manage, and the rewards are more immediate and apparent. Bernacki and
Walkington (2018) suggested that personalizing math word problems is an
effective teaching strategy to elicit students’ interest to engage and re-engage
with specific ideas. Activating interest can lead to outcomes such as increased
attention, persistence, confidence, and ultimately learning. Bridging what the
learners already know and the formal mathematics concepts is the ultimate goal
for personalizing problems.
Additionally, Spiro et al. (1988) emphasized that information, like
mathematics, should be presented from multiple perspectives and diverse
examples. The development of a learner’s representation is significant in
knowledge construction. Hence, active learning should be context-dependent.
Spiro and Jehng (2012) reiterated that learning activities must provide multiple
representations of content. Moreover, oversimplifying the content domain must
be avoided. Further, they encourage support on context-dependent knowledge
of the instructional materials.

35
 The ASTR Research Journal

Differences of the Pupils’ Performances Among Schools

Table 4. Significance of the differences in pupils’ performances in instrumental

type of test
Source of
Variables Variation
SS df MS F P-value F crit Significance

Pupils’ Per- Between

111,204.830 14 7,943.202 52.760 0.0000 1.712 Significant*
formances in Groups
Instrumental
Within
Type of Test 72,717.232 483 150.553
Groups
Among
Schools
Total 183,922.060 497

* α = 0.05

One-way ANOVA was used to test the null hypothesis that there are no
significant differences in the mean scores of the pupils’ performances in the
instrumental type of test among the schools involved in the study. Table 4 shows
that the F statistic (52.760) is greater than the F critical (1.712). Likewise, the
p-value (0.0000) is extremely smaller than the 0.05 and 0.01 levels of significance.
These results rejected the null hypothesis that there is no significant difference
in the mean scores among schools in the instrumental type of test. Hence, there
is statistical evidence that at least one school has a different mean score than
the other schools. Using Tukey’s post hoc test, five groups are generated. Table 5
shows the summary of the groupings.

Table 5. Groupings using Tukey’s post hoc method for instrumental type of test
N Mean Groupings
School #4* 42 82.87 A
School #1* 51 80.88 A
School #10* 49 71.03 B
School #3** 25 61.00 B C
School #12** 25 60.89 B C
School #5* 38 58.19 C
School #6* 35 52.06 C D
School #2** 35 45.71 D E
School #14* 32 45.31 D E
School #8* 40 45.00 D E
School #7** 41 44.11 D E

36
 Volume 4 · November 2020

N Mean Groupings
School #9** 18 42.75 D E
School #11** 24 42.59 D E
School #15** 26 39.96 E
School #13** 17 39.71 E
*lowland school; **upland school; means that do not share a letter are significantly different

As indicated in Table 5, schools #4 and #1 belong to group A and do not

share a letter. Hence, the mean scores of schools #1 and #4 are significantly
different from the others. Group B contains schools #10, #3 and #12. However,
school #10 does not share a letter which means that it differs from the others.
Group C contains schools #3, #12, #5 and #6. However, school #5 does not share
a letter which means that it differs from the others. Group D comprised seven
schools, but all of them shared a letter which means that it does not differ from
the others. Group E comprises eight schools, but schools #15 and #13 do not
share a letter. Hence, schools #15 and #13 significantly differ from the others.
The post hoc test shows that six schools have significantly different mean scores
from the others. Therefore, there is sufficient evidence to support that the pupils’
performances in the instrumental type of test from the fifteen schools involved
in the study vary significantly.

Table 6. Significance of the differences in pupils’ performances in relational type

of test
Source of P-
Variables SS df MS F F crit Significance
Variation value

Between
98,944.247 14 7,067.446 57.568 0.0000 1.712 Significant*
Pupils’ Per- Groups
formances in
Relational Type Within
of Test Among 59,296.571 483 122.767
Groups
Schools
Total 158,240.820 497
* α = 0.05

One-way ANOVA was also used to compare the pupils’ performances in

the relational type of examination to test the null hypothesis that there are no
significant differences in the mean scores among the schools involved in the
study. Table 6 shows that the F statistic (52.760) is more significant than the F
critical (1.712). Likewise, the p-value (0.0000) is extremely smaller than the 0.05
and 0.01 levels of significance. Hence, the null hypothesis is rejected in favor of

37
 The ASTR Research Journal

the alternative hypothesis. It means that there are significant differences in the
mean scores among schools in the instrumental type of examination. Hence,
there is statistical evidence that at least one school has a different mean score
than the other schools. Using Tukey’s post hoc test, six groups are generated.
Table 7 summarized the results.

Table 7. Groupings using Tukey’s post hoc method for relational type of test
N Mean Groupings
School #4* 42 76.72 A
School #1* 51 70.86 A
School #10* 49 63.32 B
School #5* 38 49.56 C
School #12** 25 48.22 C D
School #6* 35 48.02 C D
School #3** 25 45.89 C D E
School #11** 24 42.36 C D E F
School #15** 26 40.92 C D E F
School #13** 17 39.54 C D E F
School #9** 18 37.5 D E F
School #2** 35 36.9 E F
School #7** 41 36.65 E F
School #14* 32 35.76 F
School #8* 40 34.51 F
*lowland school; **upland school; means that do not share a letter are significantly
different

As shown in Table 7, group A consists of schools #4 and #1, which do not

share a letter. Hence, the mean scores for schools #4 and #1 are significantly
different from the others. Group B has school #10 only and does not share a letter.
Hence, the mean score of school #10 is significantly different from the others.
Group C consists of seven schools, but school #5 does not share a letter. Hence,
the mean score of school #5 is significantly different from the others. Group
D consists of seven schools, but all of them shared letters. Hence, all schools
belonging to group D are not significantly different from the others. Group E
has seven schools, but all of them also shared letters. Hence, the mean scores of

38
 Volume 4 · November 2020

all the schools in group E are not significantly different. Lastly, group F has eight
schools, but schools #14 and #8 do not share a letter. Hence, the mean scores of
schools #14 and #8 are significantly different from the others. In this post hoc
test, six schools have significantly different mean scores from the others. Just like
in the instrumental type of test, it can be said that there is sufficient evidence to
support that the performances of the pupils in relational type of test from the
fifteen schools involved in the study also varies significantly.

CONCLUSIONS

Most of the pupils involved in the study failed in the instrumental type of
test and more so in the relational type of test. It indicates that pupils are more
exposed to the instrumental approach of teaching than the relational teaching
approach. Furthermore, the low scores of the pupils in the instrumental type
of test indicate their low or non-mastery of the mathematical concepts and
operations introduced by their teachers. Hence, they have more difficulty in the
relational type of test. The study also revealed that the pupils’ performances in the
upland schools are generally deficient compared to some schools in the lowland
areas. To note, some schools in the lowland area manifested better performances
than the rest of the schools. It implies that the quality of math instruction in some
lowland schools is better than the rest of the schools involved in the study. The
majority of schools manifest significant differences in the pupils’ performance in
the instrumental and relational types of tests.
Based on the mean scores, the pupils’ performances in the relational type
of test is more deficient than the instrumental type of test. Hence, the pupils
have more difficulty in the relational type of test. On the differences in the pupils’
performances in the instrumental type of test, it is statistically evident that
there is a significant variation. Post hoc test reveals six schools with mean scores
statistically different from the others in the instrumental type of test. Likewise,
it is statistically evident that there is a significant variation in the mean scores
of pupils in the relational type of test. Post hoc tests also reveal six schools with
mean scores statistically different from the others in the relational type of test.
Hence, the analysis of variance and the post hoc test supports the idea that there
is indeed a variation in pupils’ performances in relational and instrumental types
of test.
As mentioned earlier, this variation is a manifestation of the differences in
the quality of instruction. However, other factors may have an indirect effect on
the pupils’ performances, which are not covered in this study. Future studies may
identify those factors to have broader background information about the low

39
 The ASTR Research Journal

performance in mathematics. School administrators may assess the teaching

practices of the mathematics teachers and provide appropriate training to
enhance their teaching competence and desire to teach mathematics effectively.

LITERATURE CITED

Andaya, O. J. F. (2014). Factors that affect mathematics achievements of students

of Philippine Normal University-Isabela Campus. Researchers World, 5(4), 83.

Anderson, J. A. (1996). Experiments to investigate instrumental and relational

understanding among undergraduates. International Journal of
Mathematical Education in Science and Technology, 27(6), 813–819. https://
[Link]/10.1080/0020739960270605

Anwar, R. B., Yuwono, I., As’ari, A. R., Sisworo, & Rahmawati, D. (2016). Mathematical
representation by students in building relational understanding on concepts
of area and perimeter of rectangle. Educational Research and Reviews, 11(21),
2002–2008. [Link]

Baclig, C. E. (2021, July 23). PH’s grade 4 students lowest in math, science around
the world – int’l study. [Link]. Retrieved from [Link]

Batool, T. (2019). Who love mathematics? Relating attitudinal factors with

mathematics in public sector secondary schools. Pakistan Journal of Social
Sciences (PJSS), 39(4), 1465–1474.

Baumeister, R. F., & Bushman, B. J. (2017). Social psychology and human nature
(4th ed.). Boston, MA: Cengage Learning.

Bernacki, M. L., & Walkington, C. (2018). The role of situational interest in

personalized learning. Journal of Educational Psychology. [Link]
org/10.1037/edu0000250

Clarke, D., & Roche, A. (2018). Using contextualized tasks to engage students in
meaningful and worthwhile mathematics learning. Journal of Mathematical
Behavior, 51, 95–108. [Link]

Cooper, J. (2007). Cognitive dissonance: 50 years of a classic theory. Thousand

Oaks, CA: Sage Publications, Inc.

40
 Volume 4 · November 2020

Dröse, J., & Prediger, S. (2019). Enhancing Fifth Graders’ Awareness of Syntactic
Features in Mathematical Word Problems: A Design Research Study on the
Variation Principle. Journal für Mathematik-Didaktik, 1-32.

Ferri, R. B. (2018). Learning how to teach mathematical modeling in school and

teacher education (1st ed.). Springer International Publishing. [Link]
org/10.1007/978-3-319-68072-9

Festinger, L. (1957). A theory of cognitive dissonance. Standford, CA: Stanford

University Press.

Gonzales, P., Partelow, L., Pahlke, E., Lerner, R., Jocelyn, L., Kastberg, D., & Williams,
T. (2004). Highlights from the Trends in International Mathematics and Science
Study (TIMSS) 2003. National Center for Education Statistics. Retrieved from
[Link]

Ingram, N., Holmes, M., Linsell, C., Livy, S., McCormick, M., & Sullivan, P. (2019).
Exploring an innovative approach to teaching mathematics through the
use of challenging tasks: a New Zealand perspective. Mathematics Education
Research Journal, 1-26. [Link]

Kang, R., & Liu, D. (2018). The Importance of Multiple Representations of

Mathematical Problems: Evidence from Chinese Preservice Elementary
Teachers’ Analysis of a Learning Goal. International Journal of Science and
Mathematics Education, 16(1), 125–143. [Link]
016-9760-8

Liu, C. C., Chen, W. C., Lin, H. M., & Huang, Y. Y. (2017). A remix-oriented
approach to promoting student engagement in a long-term participatory
learning program. Computers and Education. [Link]
compedu.2017.03.002

Macdonald, L. R., & Banes, L. C. (2017). More than words: Struggling readers’
comprehension of word problems. Journal of Teacher Action Research (Vol. 3).

Menanti, H., Sinaga, B., & Hasratuddin, D. (2018). Improve Mathematical

Connections Skills with Realistic Mathematics Education Based Learning
(pp. 29–35). Atlantis Press. [Link]

41
 The ASTR Research Journal

Pesek, D. D., & David, K. (2000). Interference of instrumental instruction in

subsequent relational learning. Journal for Research in Mathematics
Education, 31(5), 524–540. [Link]

Santos-Trigo, M., & Reyes-Martínez, I. (2019). High school prospective

teachers’ problem-solving reasoning that involves the coordinated use of
digital technologies. International Journal of Mathematical Education in
Science and Technology, 50(2), 182–201. [Link]
9X.2018.1489075

Simamora, R. E., & Saragih, S. (2019). Improving Students’ Mathematical Problem

Solving Ability and Self-Efficacy through Guided Discovery Learning in Local
Culture Context. International Electronic Journal of Mathematics Education,
14(1), 61–72. [Link]

Skemp, R. R. (1976). Relational Understanding and Instrumental Understanding.

Mathematics Teaching.

Smith, T. M., & Freels, A. (2017). Promoting Student Success in Algebra I Instructional
Practices That Promote Conceptual Understanding and Procedural Fluency
Profile of Practice Brief.

Spiro, R. J., Coulson, R. L., Feltovich, P. J., & Anderson, D. K. (1988). Cognitive
flexibility theory: Advanced knowledge acquisition in ill-structured domains.
Champaign, IL.

Spiro, R. J., & Jehng, J. C. (2012). Cognitive flexibility and hypertext: Theory and
technology for the nonlinear and multidimensional traversal of complex
subject matter. In D. Nix & R. Spiro (Eds.), Cognition, education, and multimedia:
Expolring ideas in high technology. New York, NY: Routledge.

Swanson, H. L. (2016). Word Problem Solving, Working Memory and Serious

Math Difficulties: Do Cognitive Strategies Really Make a Difference? Journal
of Applied Research in Memory and Cognition. [Link]
jarmac.2016.04.012

Utomo, D. P. (2020). The pattern of a relational understanding of fifth-grade

students on integer operations. Journal of Research and Advances in
Mathematics Education, 5(2), 119–129. [Link]
v5i2.9545

42
 Volume 4 · November 2020

Wang, Z., Utemov, V. V., Krivonozhkina, E. G., Liu, G., & Galushkin, A. A. (2018).
Pedagogical readiness of mathematics teachers to implement innovative
forms of educational activities. Eurasia Journal of Mathematics, Science
and Technology Education, 14(1), 543–552. [Link]
ejmste/80613

Wolfolk, A. (2016). Educational psychology (3rd ed.). Singapore: Pearson Education

South Asia PTE. Ltd.

Young, J. R. (2017). Technology Integration in Mathematics Education: Examining

the Quality of Meta-Analytic Research. International Journal on Emerging
Mathematics Education (IJEME), 1(1), 71–86. [Link]
v1i1.5713

43