International Journal of Evaluation and Research in Education (IJERE)

Vol. 13, No. 6, December 2024, pp. 4115~4124

ISSN: 2252-8822, DOI: 10.11591/ijere.v13i6.28395  4115

Modelling computational thinking with game-based learning

among primary school students’

Zulkifley Mohamed1, Nor Hasbiah Ubaidullah1, Noor Wahida Md Junus1, Kasthuri Devi Angamuthu2,
Ahmad3
1
Department of Mathematics, Faculty of Science and Mathematics, Sultan Idris Education University, Tanjong Malim, Malaysia
2
Ministry of Education, Putrajaya, Malaysia
3
Department of Mathematic Education, Faculty of Teacher Training and Education, Universitas Muhammadiyah Purwokerto,
Purwokerto, Indonesia

Article Info ABSTRACT

Article history: The computational thinking (CT) skills of students will be revised, increasing
their future viewpoint in the sphere of scientific activities, notably in
Received Aug 21, 2023 education interest. Game-based learning (GBL) appears to have the potential
Revised Mar 29, 2024 to improve students’ motivation to learn. Students’ GBL is associated with
Accepted May 7, 2024 higher mathematics performance, and GBL’s strong relationship with CT may
have an even larger effect. The entirety of this CT education research is
focused on undergraduate classrooms; little is revealed about how GBL
Keywords: support CT in K-12, particularly in primary schools. This study utilized a
Structural Equation Model (SEM) in modelling the relationship between CT
Computational thinking and GBL among primary school students. A sample of 90 primary school
Game-based learning students from Malaysia was chosen. In this study, the Partial Least Squares-
Hierarchical component model Structural Equation Model (PLS-SEM) was employed to develop the model.
Partial least squares The results demonstrate that empirical evidence, coupled with prior
Structural equation model observations verified the model developed. The developed model successfully
confirmed all the indicator variables stated in the constructs as all of the
associations within the model were significant. In conclusion, the lower order
components (LOC) along with the hierarchical component model (HCM) in
PLS-SEM depicted the relationship between CT and GBL, substantiated
empirically.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Zulkifley Mohamed
Department of Mathematics, Faculty of Science and Mathematics, Sultan Idris Education University
Tanjong Malim, Perak-35900, Malaysia
Email: [email protected]

1. INTRODUCTION
Computational thinking (CT) is now a key component of education in the cutting-edge educational
system. It is especially important to obtain a thorough grasp of student’ CT skills from many viewpoints as the
endeavor to incorporate CT in primary education grows [1]. Many scholars perceive CT as a collection of skills
separate from those required for computer interaction or programming tasks. This requires students to acquire
both domain-driven understanding as well as the ability to solve problems [2]. Chen et al. [3] identified the
prominence of CT growth in K-12 education as an issue that was conversed the most through a clustering
analysis of collected terms. Additionally, Durak and Saritepeci study [4] found that thinking styles, academic
performance in mathematics, and attitudes toward mathematics were highly predictive of CT proficiency.
Computational thinking encompasses problem-solving strategies and a spectrum of cognitive abilities
that surpass mere programming, applicable across various fields and domains [5]. The core components of CT

Journal homepage: http://ijere.iaescore.com

4116  ISSN: 2252-8822

including problem formulation, decomposition, pattern discernment, algorithmic design and generalization.
In today’s data-driven and technologically advanced world, CT is a crucial skill set for everyone who needs to
solve complicated problems or make judgments. It is not just for computer scientists or programmers. It can be
developed by practice and application in diverse problem-solving scenarios and is useful in a wide range of
domains, including science, engineering, business, social sciences, and everyday life [6]. The abilities
cultivated by CT are highly transferable and valuable across diverse settings, encompassing logical reasoning,
critical thinking, creativity and systematic problem solving.
A teaching strategy known as game-based learning (GBL) makes use of games as a pedagogical tool
to encourage learning and engagement among students. As stated by Khaldi et al. [7], a range of game formats
utilized on computers or mobile devices can serve as educational tools. These games can be made to teach
specific subjects, like arithmetic, science, history, or a foreign language, or they can be made to improve more
general abilities, like problem-solving, critical thinking, teamwork, and decision-making. GBL motivates
students to actively engage, explore, and persist in problem solving by harnessing their enthusiasm [8]. GBL
promotes active learning, in which students actively contribute to their education rather than merely absorbing
it. Games offer quick, relevant feedback that can assist students in analyzing their performance, correcting their
errors, and developing new abilities. Numerous advantages of GBL have been demonstrated, including
enhanced learning outcomes, greater motivation and engagement, improved problem solving and critical
thinking abilities, and growth in social skills like cooperation and communication [9]. It can be implemented
across a range of educational environments, spanning formal classrooms to informal learning setups, and can
be tailored to suit diverse age groups, subjects, and learner demographics.
The principles of CT and GBL may be blended to produce interesting educational experiences. In
order to address difficult issues, CT involves breaking things down into smaller, more manageable stages and
applying logic and algorithmic thinking. On the other hand, GBL entails using games as a teaching and learning
method in which students participate in immersive and interactive experiences to learn new information and
skills. CT can be incorporated into GBL in several ways. In problem-solving context, games provide students
a rich framework in which to employ CT techniques while addressing problems. It frequently involve
challenges, puzzles, and missions that demand players to exercise critical thinking, situational analysis, and
strategy development. Playing games is a fun and engaging technique for students to improve their problem-
solving abilities. Creating rules, mechanisms, and interactions that control how the game world behaves is a
key component of algorithmic thinking in game design. This demands algorithmic thinking since game
designers must establish the order and logic of the activity.
By planning, creating, and testing various game elements such as characters, objects, and actions,
learners can practice algorithmic thinking [10]. According to Greipl et al. [11], play is the cornerstone of any
game. Obviously, playing is a common occurrence in a child's development and often has a number of positive
developmental consequences. One may contend that playing a game with others (e.g., in competition or
cooperation) creates a unique setting that stimulates and motivates not just for winning the game or challenge
but frequently for improving one's competence or skill. Moreover, there is evidence indicating that GBL
confers benefits over conventional teaching methods in fostering deep learning [12]. When compared to the
conventional approach, neither of the GBL strategies involving parents or not has much enhanced the children'
CT in educational circumstances [13]. The outcomes also demonstrated that, in comparison to the other two
strategies, the incorporation of parental involvement in the GBL approach significantly enhanced students'
receptiveness towards acquiring CT. CT and GBL are closely related since both concepts require problem-
solving, critical thinking, and logical reasoning abilities [14]. GBL denotes the utilization of educational games
as a pedagogical strategy to enhance learning and stimulate engagement among learners.
On the contrary, CT entails a problem-solving methodology where intricate problems are
deconstructed into more manageable steps, and logical, algorithmic thinking is employed to resolve them. GBL
and CT place a strong emphasis on the improvement of problem-solving abilities [15]. In GBL, students
encounter obstacles or issues that force them to use their critical thinking skills, situational analysis, and
problem-solving abilities. Similar to analytical thinking, CT concentrates on formulating issues, designing
algorithms, and using logic to solve complicated problems. Both concepts inspire students to tackle challenges
methodically and strategically and provide with creative solutions. Similar to this, in GBL, players frequently
need to adhere to certain guidelines, protocols, or algorithms in order to accomplish their objectives. This might
entail developing strategies, making choices depending on the facts at hand, and carrying out tasks in a
sequential order. Through engaging and interactive games, GBL gives students the chance to practice and
improve their algorithmic thinking abilities. Both GBL and CT foster critical thinking abilities. Students must
analyze events, evaluate actions, and accomplish well-informed conclusions in order to advance in GBL.
Applying critical thinking abilities like analyzing data, making logical arguments, and solving problems is
required for this. Similar to this, CT requires using critical thinking abilities while formulating problems,
designing algorithms, and assessing the efficacy of solutions. Yadav and Oyelere [16] revealed that playing

Int J Eval & Res Educ, Vol. 13, No. 6, December 2024: 4115-4124
Int J Eval & Res Educ ISSN: 2252-8822  4117

games while studying could provide students a fun setting in which to practice and refine their critical thinking
abilities. Collaborative and teamwork activities are prominent in GBL, as students get together to accomplish a
task or resolve an issue. This promotes teamwork, communication, and collaboration among students as they work
towards a common goal. Similarly to this, CT can entail group problem-solving, where students emerge with
algorithms, analyze issues, and assess solutions. In real-world situations where teamwork and collaboration are
valued, CT and GBL both foster collaborative abilities. The ability of GBL to inspire and engage students is well
established. The immersive and engaging learning experiences that may be offered by well-designed educational
games has the potential to boost students’ desire to engage fully in the learning process. Similarly, CT also
incorporates practical problem-solving and inspires learners to think critically and creatively.
The majority of the CT activities included in the study were GBL, cooperative learning, problem-
based learning, and project-based learning [17]. It was found in a study by Ubaidullah et al. [18] that lacking
highly developed CT practices could result in individuals encountering major issues that might undermine their
educational or career aspirations. As the result of the educational benefits that CT offer to both instructors and
students, there has been extensive implementation of teaching strategies like project-based learning and CT
corresponding problem-solving procedures [19]. Numerous research probing at the relationship between CT
and GBL have been performed. Digital games used in CT education have favorable impacts, but such results
strongly rely on the environment in which learning is being put into practice and the users who are utilizing
the games [20]. It should be emphasized that the favorable impact of GBL on students' CT implies, in a practical
sense, that educators think about ways to support instructors in incorporating GBL into their classes so that
students may learn more CT. It is essential to create a model that takes into consideration the relationship
between GBL and CT based on the information found in this literature.
CT can be employed in a variety of disciplines and grade levels, which presents both potential and
challenges. Engagement between educators and researchers from many fields and educational backgrounds is
encouraged so that CT may be assessed and promoted. Hence, this study seeks to model the relationship
between CT and GBL among primary school students’ by utilizing SEM. The study explores theoretical
underpinnings on students' CT and GBL in attempt to estimate the relationship amidst CT and GBL through
SEM. The notion of learning through gaming elucidates how students acquire computational thinking concepts
while actively participating in a game and accomplishing its designated objectives [21]. Students may explore
CT concepts including problem-solving, decomposition, abstraction, and pattern identification when learning
through games and to enhance CT skills, establishing entertaining gaming is essential [14]. In this study, a
SEM for the relationship between CT and GBL among primary school students’ is established and examined.

2. METHOD
The study population consisted of eleven-year-old Malaysian primary school students. Following this,
the sample was chosen using the multistage cluster sampling procedure. In the first stage, the researchers
randomly select a state in Malaysia from 13 states. A state selected consisted of 13 districts. In a second stage,
a district was randomly selected which consisted of 11 localities/cities. In a third stage, a locality/city was then
randomly selected which consisted of 12 areas. In a fourth stage, an area was randomly selected which
consisted of five primary schools. At the final stage, a primary school was then randomly selected and the
entire standard 5 (11 years of age) students were selected as a sample for the study. The number of students
selected was 90 which comprises of three classes with the estimate of the population proportion of 0.177 and
bound of error of 0.012. The general guideline suggests, as proposed by Hair et al. [22], which the smallest
sample size allowed for a PLS-SEM model ought to be ten times the number of independent variables, taking
into account both measurement and structural models, or ten times the highest count of inner model paths
leading to a specific construct within the inner model. Less than nine variables are involved in the most intricate
regression in this study. Moreover, as noted by Hair et al. [22], the entire intricacy of a structural model has
minimal impact on PLS-SEM sample size requirement. This is justified by the fact that not all relationships
within the structural model are computed concurrently by the PLS-SEM procedure.
A GBL module incorporated with the CT was developed and implemented in the activities of teaching
and learning. The module developed consisted of the topics of fraction for primary school in Malaysia. In
assessing the students’ acceptance on the module development, the study utilized technology acceptance model
(TAM). This study assessed students’ CT by utilizing the Korkmaz CT scales survey instrument [23]. The CT
scales survey instrument consisted of eight items to assess creativity, six items to assess algorithmic thinking,
five items to assess critical thinking and six items to assess students’ problem-solving. For this study, the
adaptation of the CT scale survey instrument underwent further validation by seven panels of experts
specializing in computational thinking, mathematics, and education. The research utilized the content validity
index to validate CT scale survey instrument in which I-CVI for item validity and S-CVI for each and every
item on the scale. The computed I-CVI ranged from 0.7143 to 1.000, and S-CVI was recorded as 0.8971, shows
the allowed level of validity [24]. Moreover, Krippendorff’s Alpha (KA) reliability measure was utilized to
Modelling computational thinking with game-based learning among primary … (Zulkifley Mohamed)
4118  ISSN: 2252-8822

evaluate the inter-rater reliability among the panel of experts. The obtained KA value was 0.8136, indicating
that the CT instrument scales yielded precise measurements [25].
Meanwhile, TAM instrument was used to assess GBL module developed. TAM instrument consisted
of 20 items where five items were regards to perceived usefulness, six items were about perceived ease of use,
six items were to assess students’ attitude and four items were about behavioral intention to use. TAM
instrument used was validated by seven panels of experts in the field of mathematics and mathematics
education. The established I-CVI were between 0.7143 and 1.000, and the S-CVI had a value of 0.8857. This
shows that the TAM instrument used has the sufficient content validity [24]. Additionally, the KA inter-rater
reliability among the panels of experts for the TAM instrument was recorded as 0.7988. This indicates that the
TAM instrument produced measurements that were accurate [25]. The latent variables and indicator variables
derived from the survey instruments with its description were depicted in Table 1.
Following the culmination of data gathering, the response data was evaluated using the SmartPLS.
The research utilized a two-stage disjoint approach employing PLS-SEM reflective-reflective hierarchical
component model (HCM) to evaluate the relationship between CT and GBL. The initial PLS-SEM portrayal
as a research framework is depicted in Figure 1. As represented in Figure 1, the reflective-reflective HCM are
CT and GBL. The lower order components (LOC) for HCM CT are algorithmic thinking, critical thinking,
creativity, problem-solving and cooperativity. While the LOC for HCM GBL are attitude, behavioral intention,
perceived usefulness, and perceived ease of use. The study’s hypothesis examines the significant relationship
among CT and GBL.

Table 1. Description of latent and indicator variables

Latent variable
Indicator variable Description
(construct)
Creativity (COY) COY1, COY2, COY3, Creativity: The act of portraying and utilizing thoughts and ideas.
COY4, COY5, COY6,
COY7, COY8
Algorithmic AO1, AO2, AO3, AO4, Algorithmic thinking: Encompasses the capacity to understand, employ, assess, and
thinking (AO) AO5, AO6 devise algorithms.
Critical thinking CG1, CG2, CG3, CG4, Critical thinking: The utilization of cognitive abilities or methods to increase the
(CG) CG5 probability of expected outcomes.
Problem-solving PM1, PM2, PM3, PM4, Problem-solving: In the realm of education, this is acknowledged as the process of
(PM) PM5, PM6 identifying a numerical problem based on specific values and determining its solution.
Cooperativity COO1, COO2, COO3, Cooperative learning: A teaching method known as cooperative learning endeavors to
(COO) COO4 improve the learning results for both individual students and groups in small-group
environments.
Perceived UF1, UF2, UF3, UF4, Perceived usefulness: Perception of oneself on performance and efficiency of a
usefulness (UF) UF5 technology.
Perceived ease of EF1, EF2, EF3, EF4, Perceived ease of use: Perception of oneself on the easiness of features of a
use (EF) EF5, EF6 technology.
Attitude (AD) AD1, AD2, AD3, AD4, Attitude: Students' attitudes are their assessments of whether using the technology will
AD5 be advantageous to them.
Behavioral BE1, BE2, BE3, BE4 Behavioral intention: The term behavioral intention describes how someone intends to
intention (BE) use technology for learning both now and in the future.

Creativity Critical
thinking

Algorithmic
thinking CT Cooperativity

Behavioral intention Problem-solving

TAM
(GBL)

Attitude

Perceived Perceived
of ease of use
usefulness

Figure 1. Framework of the study

Int J Eval & Res Educ, Vol. 13, No. 6, December 2024: 4115-4124
Int J Eval & Res Educ ISSN: 2252-8822  4119

3. RESULTS AND DISCUSSION

3.1. Structural and measurement model
The study's model is composed of two primary parts: the outer model, also referred to as the structural
model, and the measurement model, which also incorporates the reflective-reflective HCM. The
comprehensive model encompasses two reflective-reflective HCMs and nine measurement models, each
represented by its latent and indicator variables. For CT reflective-reflective HCM, comprises of LOC, namely
algorithmic thinking, critical thinking, creativity, problem-solving and cooperativity. The measurement model
algorithmic thinking consisted of the indicator variables AO1, AO2, AO3, AO4, AO5 and AO6. The indicator
variables CG1, CG2, CG3, CG4 and CG5 were composed in critical thinking. COY1, COY2, COY3, COY4,
COY5, COY6, COY7 and COY8 were the indicator variables for creativity. The indicator variables PM1, PM2,
PM3, PM4, PM5 and PM6 were composed in problem-solving. While the indicator variables COO1, COO2,
COO3 and COO4 were composed in cooperativity.
The second reflective-reflective HCM was GBL consisted of four lower component models, namely
attitude, behavioral intention, perceived usefulness and perceived ease of use. The indicator variables for
attitude were AD1, AD2, AD3, AD4 and AD5. The indicator variables for behavioral intention were BE1,
BE2, BE3 and BE4. The indicator variables for perceived usefulness were UF1, UF2, UF3, UF4 and UF5.
While the indicator variables for perceived ease of use were EF1, EF2, EF3, EF4, EF5 and EF6. The path
diagram connecting CT and GBL constitutes the study structural model. The reflective-reflective HCM,
measurement models and structural model with its components loading and coefficient of determination were
depicted in Figure 2. All indicator variables' component loadings for their respective latent variables are greater
than 0.700, indicating that each indicator variable sufficiently reflects its latent variable [26].

Figure 2. The regression coefficient and coefficient of determination for each construct

3.2. Measurement model validity and reliability

This section examined the constructs’ validity and reliability in addition to evaluating each
measurement model. When assessing the measurement model, various evaluations should be considered,

Modelling computational thinking with game-based learning among primary … (Zulkifley Mohamed)
4120  ISSN: 2252-8822

including the AVE for convergent validity, the HTMT for discriminant validity, Cronbach’s alpha (α) and
composite reliability for internal consistency, and outer loading for indicator reliability. A HTMT value below
0.900 indicates the establishment of discriminant validity between two reflective constructs. While the AVE
higher than 0.500, Cronbach’s alpha (α) and composite reliability value between 0.70 and 0.95, the convergent
validity and reliability respectively are established [22]. Table 2 displays the validity, internal consistency and
reliability statistics from the PLS-SEM. The values of AVE for algorithmic thinking, critical thinking,
creativity, problem-solving, cooperativity, attitude, behavioral intention, perceived usefulness and perceived
ease of use are greater than 0.500. This demonstrates the establishment of convergent validity, wherein all
indicator variables in the model converge to represent the underlying constructs developed, as recommended
by Hair et al. [22]. In addition, the fact that all of the constructs in Table 3 had HTMT values less than 0.850,
which imply discriminant validity, have their own distinct identities and are unrelated to any other constructs
in the study. As for internal consistency and reliability, all the Cronbach’s alpha (α) and composite reliability
value, namely AO, CG, COY, PM, COO, AD, BE, UF, and EF are higher than 0.700. This indicates that each
model's indicator variable represents the related constructs.

Table 2. The reliability value of the constructs’ indicator variables

Construct Indicator variable Indicator reliability Composite reliability Cronbach’s alpha AVE
Creativity COY1 0.762 0.888 0.899 0.553
COY2 0.750
COY3 0.804
COY4 0.763
COY5 0.706
COY6 0.688
COY7 0.829
COY8 0.627
Algorithmic thinking AO1 0.690 0.851 0.862 0.569
AO2 0.773
AO3 0.701
AO4 0.825
AO5 0.795
AO6 0.732
Critical thinking CG1 0.863 0.841 0.856 0.611
CG2 0.787
CG3 0.772
CG4 0.739
CG5 0.742
Problem-solving PM1 0.784 0.892 0.911 0.646
PM2 0.857
PM3 0.768
PM4 0.768
PM5 0.826
PM6 0.780
Cooperativity COO1 0.793 0.849 0.854 0.689
COO2 0.875
COO3 0.791
COO4 0.859
Perceived usefulness UF1 0.778 0.832 0.838 0.599
UF2 0.754
UF3 0.709
UF4 0.797
UF5 0.826
Perceived ease of use EF1 0.690 0.853 0.869 0.579
EF2 0.804
EF3 0.826
EF4 0.688
EF5 0.832
EF6 0.707
Attitude AD1 0.782 0.805 0.885 0.560
AD2 0.842
AD3 0.584
AD4 0.767
AD5 0.739
Behavioral intention BE1 0.818 0.793 0.812 0.615
BE2 0.727
BE3 0.745
BE4 0.842

Int J Eval & Res Educ, Vol. 13, No. 6, December 2024: 4115-4124
Int J Eval & Res Educ ISSN: 2252-8822  4121

Table 3. The HTMT value of the model constructs

Construct AO AD BE CT COO COY CG UF EF PM
AD 0.480
BE 0.426 0.633
CT 0.533 0.371 0.276
COO 0.220 0.214 0.268 0.594
COY 0.445 0.345 0.382 0.482 0.199
CG 0.464 0.474 0.440 0.769 0.202 0.426
UF 0.359 0.438 0.589 0.196 0.392 0.235 0.282
EF 0.173 0.265 0.290 0.538 0.759 0.210 0.216 0.548
PM 0.288 0.310 0.245 0.722 0.351 0.171 0.575 0.136 0.328
TAM (GBL) 0.226 0.440 0.610 0.345 0.672 0.232 0.136 0.698 0.829 0.246
Note: CT=Computational thinking, TAM (GBL)=TAM (Game-based learning)

3.3. Structural model evaluation

Structural models are evaluated by assessing the values of the formulated hypothesis path coefficients
as there is a relationship between CT and GBL. A PLS-SEM bootstrap procedure was conducted, utilizing the
t-score to evaluate the significance level of relationships. The findings indicated a significant relationship
between CT and GBL (r=0.607, t=7.165, p<0.001). The resulting R2 value, which is 0.370, indicates that 37
percent of the variance in CT can be impacted by GBL, or in other word, GBL contributed almost 37 percent
to CT. The study does not distinguish from the work of Ma et al. [27], which shown that GB significantly
improved students' CT on the whole. Furthermore, a substantial relationship among all LOC, namely
algorithmic thinking, critical thinking, creativity, problem-solving and cooperativity with the HCM CT as
indicated by their component loadings depicted in Figure 2. The research organized by Ponce et al. [26] indicate
that algorithmic thinking, as a CT component, demonstrates a similar correlation pattern to our study (r=0.225,
t=2.166, p<0.05). Study by Lemay et al. [28] found that critical thinking, as a CT constituent, does not
significantly differ from our findings, which exhibit a significant relationship between CT and critical thinking
(r=0.503, t=5.459, p<0.001). The study unveiled that creativity, as an integral part of CT, exhibits a significant
relationship (r=0.317, t=3.135, p<0.005), consistent with Durak and Saritepeci [4]. Additionally, our study
highlights the relationship between CT and problem-solving (r=0.475, t=5.064, p<0.001), as initially observed
by Palts and Pedaste [29]. Moreover, the study underscores a significant relationship between cooperativity
and CT (r=0.569, t=6.491, p<0.001), in-line with Durak and Saritepeci [4] research.
In the context of TAM for GBL, the study unveiled relationships throughout attitude, behavioral
intention, perceived usefulness, and perceived ease of use with HCM GBL. Yeo et al. [30] found a positive
and significant relationship between attitude towards game use and intention to use digital games, which aligns
with our study where GBL statistically relates to attitude (r=0.323, t=3.202, p<0.005). The study also identified
that behavioral intention is influenced by GBL, establishing a significant relationship between these constructs
(r=0.433, t=4.367, p<0.001), which corresponds with Razami and Ibrahim’s investigation [31] on the impact
of gamification on behavioral intention. Moreover, the study found that GBL and perceived usefulness have a
significant relationship (r=0.625, t=7.511, p<0.001), consistent with Krath et al. [32] emphasis on GBL's role
in enhancing knowledge acquisition through perceived usefulness. Finally, the study demonstrated that GBL
significantly contributes to students' perception of ease of use (r=0.516, t=5.651, p<0.001), mirroring research
by Musyaffi et al. [33] on gamification quality and perceived ease of use relationship.
Research by Theodoropoulos [34] revealed that students can improve problem-solving and critical
thinking through debugging games, where they identify and fix issues like bugs and errors. Debugging and
troubleshooting are essential CT skills. Computational creativity, using computational tools for original
solutions, can be fostered through GBL [35]. Students can create games, simulations, or interactive stories
using game engines or coding platforms to express creativity and apply CT. Game data collection and analysis
typically involve player behavior, game performance, or user feedback [36]. As they base decisions on gathered
data, this allows students to practice data analysis and decision-making. Learners strengthen analytical and
critical thinking through gaming data analysis, and learn to make data-driven decisions. Both CT and GBL can
boost learners’ motivation and involvement, positively impacting results. With a shared focus on cooperation,
problem-solving, critical thinking, and algorithmic reasoning, CT and GBL are closely linked. CT enhances
problem-solving and critical thinking, beneficial in GBL scenarios, while GBL provides an environment to
develop these skills.
In the perspective of SEM, the study demonstrated practical contribution in understanding the
relationship between GBL and CT by employing HCM PLS-SEM. Current research practices primarily
emphasize a model that examines only the LOC of CT. Several studies [4], [28], [37] have concentrated on
these LOC to analyze the relationship between CT and its constituent elements. Similarly, research by Alt [8]
has examined the constituent elements of GBL using similar LOC. This study signifies a pioneering initiative

Modelling computational thinking with game-based learning among primary … (Zulkifley Mohamed)
4122  ISSN: 2252-8822

in employing HCM PLS-SEM modeling, particularly in the spheres of GBL and CT, which are recognized as
vital elements in educational settings.

4. CONCLUSION
The model was adeptly developed and tested. PLS-SEM was utilized in the study to effectively
validate all indicator variables, and the findings demonstrate that the developed model was strengthened by
empirical evidence, aligning with previous results and the theoretical framework. The findings reveal that
students' CT and GBL relationship are statistically significant. The developed HCM has made a significant
contribution to the research methodology. The HCM methodological contribution is rooted in the framework
it provides for comprehending and examining processes of CT and GBL. The HCM enables researchers to
decomposed complex CT processes into smaller, more digestible components. This decomposition provides a
pulverized analysis of CT and GBL operations, which makes it possible to examine and comprehend how
various operations interact and contribute to overall performance. In conclusion, the practical contributions of
CT and GBL in education can be amalgamated to create impactful educational encounters nurturing creativity,
critical thinking and problem-solving skills. Beyond gaining expertise in a certain field pertaining to game
content, learners can cultivate CT abilities through enjoyable and interactive gameplay, leveraging the
captivating and immersive aspects of games. To the best of our understanding, this is the first study to employ
the HCM with a PLS-SEM in examining the relationship between CT and GBL. The study is exploratory in
design due to small sample size, which favors PLS-SEM over SEM-AMOS. The model's development can be
accelerated by conducting further analysis on larger data sets.

ACKNOWLEDGEMENTS
The authors would like to express gratitude towards Sultan Idris Education University, Malaysia, and
the Malaysian Ministry of Higher Education for supporting this work financially, which was made possible by
grant number 2020-0045-107-01.

REFERENCES
[1] Y.-C. Liu, T.-H. Huang, and C.-L. Sung, “The determinants of impact of personal traits on computational thinking with
programming instruction,” Interactive Learning Environments, vol. 31, no. 8, pp. 4835–4849, Nov. 2023, doi:
10.1080/10494820.2021.1983610.
[2] X. Tang, Y. Yin, Q. Lin, R. Hadad, and X. Zhai, “Assessing computational thinking: A systematic review of empirical studies,”
Computers & Education, vol. 148, Apr. 2020, doi: 10.1016/j.compedu.2019.103798.
[3] H. E. Chen, D. Sun, T.-C. Hsu, Y. Yang, and J. Sun, “Visualising trends in computational thinking research from 2012 to 2021: A
bibliometric analysis,” Thinking Skills and Creativity, vol. 47, Mar. 2023, doi: 10.1016/j.tsc.2022.101224.
[4] H. Y. Durak and M. Saritepeci, “Analysis of the relation between computational thinking skills and various variables with the
structural equation model,” Computers & Education, vol. 116, pp. 191–202, Jan. 2018, doi: 10.1016/j.compedu.2017.09.004.
[5] K.-D. Boom, M. Bower, J. Siemon, and A. Arguel, “Relationships between computational thinking and the quality of computer
programs,” Education and Information Technologies, vol. 27, no. 6, pp. 8289–8310, Jul. 2022, doi: 10.1007/s10639-022-10921-z.
[6] C. Lu, R. Macdonald, B. Odell, V. Kokhan, C. Demmans Epp, and M. Cutumisu, “A scoping review of computational thinking
assessments in higher education,” Journal of Computing in Higher Education, vol. 34, no. 2, pp. 416–461, Aug. 2022, doi:
10.1007/s12528-021-09305-y.
[7] A. Khaldi, R. Bouzidi, and F. Nader, “Gamification of e-learning in higher education: a systematic literature review,” Smart
Learning Environments, vol. 10, no. 1, Jan. 2023, doi: 10.1186/s40561-023-00227-z.
[8] D. Alt, “Assessing the benefits of gamification in mathematics for student gameful experience and gaming motivation,” Computers
& Education, vol. 200, Jul. 2023, doi: 10.1016/j.compedu.2023.104806.
[9] K. Park, B. Mott, W. Min, E. Wiebe, K. E. Boyer, and J. Lester, “Generating game levels to develop computer science competencies
in game-based learning environments,” in Artificial Intelligence in Education, 2020, pp. 240–245. doi: 10.1007/978-3-030-52240-
7_44.
[10] W. Zhao and V. J. Shute, “Can playing a video game foster computational thinking skills?” Computers & Education, vol. 141, Nov.
2019, doi: 10.1016/j.compedu.2019.103633.
[11] S. Greipl, K. Moeller, and M. Ninaus, “Potential and limits of game-based learning,” International Journal of Technology Enhanced
Learning, vol. 12, no. 4, pp. 363–389, 2020, doi: 10.1504/IJTEL.2020.110047.
[12] F. Crocco, K. Offenholley, and C. Hernandez, “A Proof-of-Concept Study of Game-Based Learning in Higher Education,”
Simulation & Gaming, vol. 47, no. 4, pp. 403–422, Aug. 2016, doi: 10.1177/1046878116632484.
[13] X. Zhang et al., “Developing rural Chinese children’s computational thinking through ­game-based learning and parental
involvement,” The Journal of Educational Research, vol. 116, no. 1, pp. 17–32, Jan. 2023, doi: 10.1080/00220671.2023.2167798.
[14] T. Turchi, D. Fogli, and A. Malizia, “Fostering computational thinking through collaborative game-based learning,” Multimedia
Tools and Applications, vol. 78, no. 10, pp. 13649–13673, May 2019, doi: 10.1007/s11042-019-7229-9.
[15] F. J. Agbo, S. S. Oyelere, J. Suhonen, and M. Tukiainen, “Design, development, and evaluation of a virtual reality game-based
application to support computational thinking,” Educational technology research and development, vol. 71, no. 2, pp. 505–537,
Apr. 2023, doi: 10.1007/s11423-022-10161-5.
[16] A. K. Yadav and S. S. Oyelere, “Contextualized mobile game-based learning application for computing education,” Education and
Information Technologies, vol. 26, no. 3, pp. 2539–2562, May 2021, doi: 10.1007/s10639-020-10373-3.

Int J Eval & Res Educ, Vol. 13, No. 6, December 2024: 4115-4124
Int J Eval & Res Educ ISSN: 2252-8822  4123

[17] T.-C. Hsu, S.-C. Chang, and Y.-T. Hung, “How to learn and how to teach computational thinking: Suggestions based on a review
of the literature,” Computers & Education, vol. 126, pp. 296–310, Nov. 2018, doi: 10.1016/j.compedu.2018.07.004.
[18] N. H. Ubaidullah, Z. Mohamed, J. Hamid, S. Sulaiman, and R. L. Yussof, “Improving novice students’ computational thinking
skills by problem-solving and metacognitive techniques,” International Journal of Learning, Teaching and Educational Research,
vol. 20, no. 6, pp. 88–108, Jun. 2021, doi: 10.26803/ijlter.20.6.5.
[19] A. Saad and S. Zainudin, “A review of project-based learning (PBL) and computational thinking (CT) in teaching and learning,”
Learning and Motivation, vol. 78, May 2022, doi: 10.1016/j.lmot.2022.101802.
[20] A. Theodoropoulos and G. Lepouras, “Digital game-based learning and computational thinking in P-12 education: A systematic
literature review on playing games for learning programming,” in Handbook of Research on Tools for Teaching Computational
Thinking in P-12 Education, 2020, pp. 159–183. doi: 10.4018/978-1-7998-4576-8.ch007.
[21] S. Sukirman, L. F. M. Ibharim, C. S. Said, and B. Murtiyasa, “A strategy of learning computational thinking through game based
in virtual reality: Systematic review and conceptual framework,” Informatics in Education, vol. 21, no. 1, Jun. 2021, doi:
10.15388/infedu.2022.07.
[22] J. F. Hair, G. T. M. Hult, C. M. Ringle, and M. Sarstedt, A primer on partial least squares structural equation modeling (PLS-
SEM). SAGE Publications, 2022.
[23] Ö. Korkmaz, R. Çakir, and M. Y. Özden, “A validity and reliability study of the computational thinking scales (CTS),” Computers
in Human Behavior, vol. 72, pp. 558–569, Jul. 2017, doi: 10.1016/j.chb.2017.01.005.
[24] M. S. B. Yusoff, “ABC of content validation and content validity index calculation,” Education in Medicine Journal, vol. 11,
no. 2, pp. 49–54, Jun. 2019, doi: 10.21315/eimj2019.11.2.6.
[25] K. Krippendorff, Content analysis: An introduction to its methodology. Sage Publications, 2018.
[26] B. Alfaro-Ponce, A. Patiño, and J. Sanabria-Z, “Components of computational thinking in citizen science games and its contribution
to reasoning for complexity through digital game-based learning: A framework proposal,” Cogent Education, vol. 10, no. 1, Dec.
2023, doi: 10.1080/2331186X.2023.2191751.
[27] J. Ma, Y. Zhang, Z. Zhu, S. Zhao, and Q. Wang, “Game-based learning for students’ computational thinking: A meta-analysis,”
Journal of Educational Computing Research, vol. 61, no. 7, pp. 1430–1463, Dec. 2023, doi: 10.1177/07356331231178948.
[28] D. John Lemay, R. B. Basnet, T. Doleck, P. Bazelais, and A. Saxena, “Instructional interventions for computational thinking:
Examining the link between computational thinking and academic performance,” Computers and Education Open, vol. 2, Dec.
2021, doi: 10.1016/j.caeo.2021.100056.
[29] T. Palts and M. Pedaste, “A model for developing computational thinking skills,” Informatics in Education, vol. 19, no. 1, pp. 113–
128, Mar. 2020, doi: 10.15388/infedu.2020.06.
[30] S. Yeo, T. Rutherford, and T. Campbell, “Understanding elementary mathematics teachers’ intention to use a digital game through
the technology acceptance model,” Education and Information Technologies, vol. 27, no. 8, pp. 11515–11536, Sep. 2022, doi:
10.1007/s10639-022-11073-w.
[31] H. Hafiza Razami and R. Ibrahim, “Models and constructs to predict students’ digital educational games acceptance: A systematic
literature review,” Telematics and Informatics, vol. 73, Sep. 2022, doi: 10.1016/j.tele.2022.101874.
[32] J. Krath, L. Schürmann, and H. F. O. von Korflesch, “Revealing the theoretical basis of gamification: A systematic review and
analysis of theory in research on gamification, serious games and game-based learning,” Computers in Human Behavior, vol. 125,
Dec. 2021, doi: 10.1016/j.chb.2021.106963.
[33] A. M. Musyaffi, W. A. Sulistyowati, C. W. Wolor, and A. A. Sasmi, “Game-based learning sustainability during social distance:
The role of gamification quality,” European Journal of Educational Research, vol. 11, no. 3, pp. 1289–1302, Jul. 2022, doi:
10.12973/eu-jer.11.3.1289.
[34] A. Theodoropoulos, “Participatory design and participatory debugging: Listening to students to improve computational thinking by
creating games,” International Journal of Child-Computer Interaction, vol. 34, Dec. 2022, doi: 10.1016/j.ijcci.2022.100525.
[35] D. Hooshyar, L. Malva, Y. Yang, M. Pedaste, M. Wang, and H. Lim, “An adaptive educational computer game: Effects on students’
knowledge and learning attitude in computational thinking,” Computers in Human Behavior, vol. 114, Jan. 2021, doi:
10.1016/j.chb.2020.106575.
[36] J. Tay, Y. M. Goh, S. Safiena, and H. Bound, “Designing digital game-based learning for professional upskilling: A systematic
literature review,” Computers & Education, vol. 184, Jul. 2022, doi: 10.1016/j.compedu.2022.104518.
[37] M.-J. Tsai, J.-C. Liang, S. W.-Y. Lee, and C.-Y. Hsu, “Structural validation for the developmental model of computational
thinking,” Journal of Educational Computing Research, vol. 60, no. 1, pp. 56–73, Mar. 2022, doi: 10.1177/07356331211017794.

BIOGRAPHIES OF AUTHORS

Zulkifley Mohamed has served as a lecturer in Sultan Idris Education

University, Malaysia since 2001. He was appointed as a professor in 2021. He obtained his
bachelor degree from MARA University of Technology Malaysia and Master of Science in
Applied Statistics and Operational Research from University of Salford, UK. In 2007 he
was awarded PhD in Statistics from National University of Malaysia. His research interest
is in statistical modelling, particularly in education. He can be contacted through email:
[email protected].

Modelling computational thinking with game-based learning among primary … (Zulkifley Mohamed)
4124  ISSN: 2252-8822

Nor Hasbiah Ubaidullah is currently being an associate professor at the Sultan

Idris Education University, Malaysia. She started her carrier as a lecturer in 1992 at several
institutes of higher learning in Malaysia. She received her PhD in Information Technology
from National University Malaysia in 2007. She is so passionate in research, especially in
information technology and IT education. She has been awarded several research grants
from the Malaysia Ministry of Higher Education. Nor Hasbiah can be contacted at email:
[email protected].

Noor Wahida Md Junus is the head of department at the Department of

Mathematics, Faculty of Science and Mathematics, Sultan Idris Education University,
Malaysia. She is a lecture by profession and serves Sultan Idris Education, University
Malaysia since 2012. She was awarded a PhD in Statistics from Malaysia Science University
in 2018. Her research area is in statistics and time series analysis. She can be contacted at
email: [email protected].

Kasthuri Devi Angamuthu is a Ph.D. Candidate at the Department of

Mathematics, Faculty of Science and Mathematics, Sultan Idris Education University,
Malaysia. She is a teacher by profession. She has more than 20 years of working experience
as a teacher particularly in teaching mathematics in primary schools in Malaysia. Her current
research project is on the development of the teaching and learning module for topics on
fraction in primary school. She can be contacted through: [email protected].

Ahmad has served as a lecturer in Universitas Muhammadiyah Purwokerto,

Indonesia since 1995. He was appointed as a professor in 2022. He obtained his bachelor
degree from Universitas Ahmad Dahlan, Indonesia and Master of Mathematics Education
from Universitas Pendidikan Indonesia, Bandung, Indonesia. In 2018 he was awarded PhD
in Mathematics Education from Sultan Idris Education University, Malaysia. His research
area is in mathematics education. He can be contacted through email: [email protected].

Int J Eval & Res Educ, Vol. 13, No. 6, December 2024: 4115-4124