The effects of gamification method and cognitive style on children’s early reading ability

Abstract Fun and pervasive education into students’ daily lives is an important part of children’s development. Learning methods and cognitive styles are one of the important aspects of the dynamics of learning. This research was a quasi-experimental study using the pretest-posttest of a nonequivalent control group design. Therefore, research based on these is expected to impact students’ early reading ability. There are 20 students in the experimental and 20 students in the control classes. The age range is 5–6 years, with heterogeneous gender. The experimental group used the gamification method, while the control group used the Beyond Center and Circle Times (BCCT) method. The gamification method and students with field independence cognitive style are proven to have superiority in improving early reading ability.

also revealed by Synodi (2010) saying that playing is a part of a child's life. According to him, playing contributes children to know their environment, interact socially, master language, mature the physical-motoric children and teach children about social norms (Synodi, 2010).
The term play is very close to gamification. The first hypothesis was to test the effectiveness of the gamification method conducted by Nazaret (Del Río et al., 2019). He has done a research theme on gamification methods for children who are obese. This research involved 46 children aged 6-12 years with obesity. Nazareth used a quasi-experimental design involving two groups (experiment and control). This research lasted for 3 years (longitudinal research). The results of this research indicated that the gamification method could provide a significant difference. The experimental group greatly influenced the diet program using the gamification method compared to children in the control group.
Another research on the effectiveness of the gamification method was also carried out by Dodero (Dodero et al., 2014). His research involved 56 primary school children aged 8-9 years. The research about gamification he did was longitudinal. According to him, the gamification method applied in the class is a real game about making cycle prototypes, such as photosynthesis. The elements used in gamification in this research include rapid feedback, rules and challenges, cooperation, and rewards. As a result, students who were subjected to the gamification method influenced their students' interactions and learning outcomes.
In language teaching, gamification is also used in learning Chinese mastery. The research was conducted by Dicky (Rawendy et al., 2017). This research was conducted in primary schools with an age range of 6-12 years. From the research that has been done, it was shown that the pretest and posttest data had a significant difference where the posttest value was higher. This supports the idea that the gamification method can be effective in teaching language.
In this research there are two terms used in the gamification method namely Learning Mechanic (LM) and Games Mechanic (GM). The authors formulated this research as LM-GM. We will explain in detail in Figure 1. We formulated LM with an image on the left, while GM is formulated in the image on the right.

Figure 1. Learning and games mechanics as the basis for LM-GM implementation in learning
The selection of terms in Learning Mechanic (LM) above was done on the basis of a literature review that the authors conducted by taking a number of opinions (Keller, 2009), (Gagne et al., 1992) and (Papert & Harel, 1991). In searching for the application keyword of Game Mechanic (GM), the authors also conducted a literature review by taking the terms of several experts namely (Brainerd, 1978), (Järvinen, 2008), (Sicart, 2008). The authors conducted a review using the meta-analysis method by identifying keywords and research results from several research results.
Based on the above meta-analysis, the authors ventured to implement the implementation of gamification by applying LM-GM in learning. To apply a gamification method, the following steps were made based on GM-LM. The design of the gamification method was based on a rational basis, which was adjusting the basic needs of students and being able to meet the learning objectives. The design is based on the following six steps: (1) Determine the learning goal (learning outcomes), such as the theme of transportation) (2) Identify students' characteristics, such as age and gender.
(3) Determine the determinants of the psychological success of students, such as cognitive style and self-efficacy.
(4) Encourage students to perceive their knowledge into a game, such as students are asked to conquer a challenge, and students get points and bonuses) (5) Encourage students to imagine, such as students are asked to act as superheroes) (6) In gamification, players must always be happy, the formula (behaviour change = knowledge x motivation). The goal is to show that a fun game can effectively produce the expected behavior changes, which are following learning outcomes.
The author tried to provide an illustration of how to implement the gamification method based on Learning Mechanic and Game Mechanic in the following picture: Implementation in learning starts from the stages of starting games. Firstly, the authors chose a term in Game Mechanic (GM) on the right (marked with numbers). In the example above, the authors chose the level and the term in Learning Mechanic (LM) (marked with the alphabet). Instructional naming and selection based on the principle of similarity and suitability of activities between GM and LM were chosen. Secondly, the authors chose the tutorial term which is also found in LM and GM (Arnab et al., 2015). Thirdly, the core of the gamification method was determined. In this step, the researchers implemented several learning activities, such as collecting items and taking items from the enemy (defeating enemies) to enliven the learning atmosphere. Then, the fourth step was the culmination of the activity, namely defeating the king/boss. This was done as a cover of gamification activities in learning. In practice, the researchers made an example of healthy living and defeated germs and killed the king of germs so that children have a healthy lifestyle starting from the activity of knowing oneself. Next, the fifth step was the provision of feedback. This activity will be adjusted to the existing theme at school. Feedback can be in the form of inputs and rewards, and points for children who participate in this game with enthusiasm. Eventually, the step was to end the game by giving the leaderboards to students. This step was carried out to provide information on learning outcomes obtained by children on that day, aiming to reflect the activities of gamification the next day. Figure 2 shows the existence of repetitive cycles, where there is a side of behavioral momentum where children are expected to have measurable changes in behavior. As stated by John (Nevin & Grace, 2000), the stimulus presented by the experimenter, the response of the organism, and the reinforcer that follows the response are fundamental elements in the science of behavior. On the  other hand, the word repetition intended that this gamification method can be done repeatedly to achieve the expected learning objectives: healthy living, defeating germs, and killing the king of germs. It is shown in the figure the term NPC (non-player character). NPC is someone who provides guidance and advice to players (students), but they are not directly involved in the game. For example, a teacher or student who is not playing. Figure 3 Shows the process of acquiring information develops according to a child's age level.
Some children may find it easy to get information from various things, such as games, parents, the environment, and different places and times. The process of acquiring this information in this research is called cognitive style. According to Witkin, cognitive style is a term to define someone's "likes" in processing information (Witkin et al., 1971). Cognitive style is also defined as specific characteristics of decision-makers, and hence it can be considered a factor that influences decision-making performance in obtaining information (Engin & Vetschera, 2017). Mefoh supports this opinion that cognitive style is an important variable that can help problem-solving and information processing (Mefoh et al., 2017). Cognitive style refers to how individuals process information or the different ways they think and learn.
The authors have also published a literature study on the effects of using the method of gamification to stimulate activity in children. This research was conducted by searching for a literature map using the keywords "the effect of the gamification method" on children. The literature map started by searching the journal database from SAGE, Elsevier, and Emerald in 2014-2018. There were 4000 journal titles that match the predetermined keywords which is about the effect of the gamification method. When filtered using Mendeley software, there were some duplications. After that, the process involves a reviewer analyzing the appropriateness of the journal title with the limitations that the authors conducted. From the results of the review, it was found that there were six appropriate titles. The six titles were analyzed using Review Manager software to determine the magnitude of each effect. It was known that the six types of research had a value of 1.01 with a confidence interval of 0.98-1.05 with a significance level of 0.05. The results of these calculations can be seen in the following figure: These findings confirm that the effect of the gamification method for children aged 6 years exceeds the criteria suggested by Cohen, which is 0.8 (Gignac & Szodorai, 2016). Figure 4 shows he 6 studies of gamification methods that have been carried out have a positive impact on children. The research data also provides information about variations in the use of gamification methods in learning.
Initial information was obtained based on some of the opinions above and preliminary studies that researchers have conducted by conducting a meta-analysis and reviewing journals about the use of gamification for education, especially early childhood education. This gamification method could be implemented in the learning process to stimulate language development for children's early reading ability.

Research questions
The three research questions for this research were as follows: (1) Is there a significant difference in influence on the early reading ability between students taught using the gamification method and those taught using the Beyond Center and Circle Times method?
(2) Is there a significant difference in influence on early reading ability between students who have a field dependence cognitive style (CS-FD) and students who have a field independence cognitive style (CS-FI)?
(3) Is there an effect of the interaction between the gamification method and the BCCT method on early reading ability?

Methods
This research aimed to determine the effect of independent variables with the experimental design but because the experimental group and the control group cannot be randomly sorted, this research used a quasi-experimental method with a 2 × 2 factorial design (Tuckman, 1999). The variables in this research consisted of independent variables, moderator variables, and dependent variables. The dependent variable included the learning method which consists of the gamification method and BCCT method. The moderator variable was cognitive style. Meanwhile, the dependent variable observed was early learning ability.
Factorial design is defined as a research structure in which two or more independent variables are discussed together to study the effects independently or interactively on a dependent variable (Kerlinger & Lee, 2000). This factorial design can treat two or more independent variables at the same time. This design divides groups according to the number of groups to be studied. The design of this research refers to a 2 × 2 factorial design involving two groups. The first group as an experimental class used the gamification method and the second group used the BCCT method, the quasi-experimental design in this research used 2 × 2 factorial and the data analysis of the results of this research was ANOVA 2 path.
The sorting factor is the moderator variable of Linguistic Intelligence. The research design matrix is shown in Table 1.
Ethics is very important to ensure the data privacy security of every participant involved in the research. In addition, their condition as children should be prioritized, respected, and protected. We have used the ethics approval and consent form from the deputy of research center Universitas Negeri Malang No:123/23/KK/Ethics for 40 students. In this research, the researchers used two types of instruments, namely 1) the instrument to measure the cognitive style of the Children Embedded Figure Test (CEFT) developed by Witkin and Goodenough (1981) which is a moderator variable; 2) the instrument to measure the early reading ability using the Early Reading Screening Inventory instrument (Morris & Slavin, 2003) which is the dependent variable in this research.
To analyze data based on this factorial design, the data analysis technique was divided into two groups, data analysis for prerequisite tests and data analysis tests to prove the research hypothesis. The analysis was carried out on all the research variables that had been determined.
Test requirements analysis used the normality test and the variance homogeneity test. The data normality test in this research used the Kolmogorov-Smirnov technique while the variance homogeneity test used the Levene's test. Data normality tests and homogeneity tests were performed to meet the parametric assumptions. Data analysis for testing the research hypotheses in this research used Analysis of Variance (ANOVA) with the assistance of IBM SPSS 26 for Mac software and all the above parametric tests were carried out with a significance score of 5%.

Results
Data from the early reading ability pretest between children treated with the BCCT method and the gamification method, with the field dependence cognitive style and the field independence cognitive style were recapitulated aiming to provide general information about the initial conditions of the research subjects subjected to two different methods. From the results of the early reading ability pretest, it can be shown in Table 2 below: Data obtained from Table 2 shows that in the control class column in students taught using the BCCT method, the mean of early reading result with CS-FD reached 59.2 with a standard deviation of 10.7. Meanwhile, for students with CS-FI, the learning outcomes of the early reading ability material reached 69.3, with a standard deviation of 7.8. Whereas the group of students in the class that was treated using the gamification method, the mean obtained for the learning outcomes of the early reading ability with CS-FD was 69.6 with a standard deviation of 5.2. Meanwhile, students with CS-FI for the early reading ability obtained a mean of 73.6 and a standard deviation of 6.3.
From the overall pretest data, it can be seen that the data did not show a significant difference in the students' early reading ability between students with CS-FD and CS-FI, both in the control class and in the experiment. The data also illustrated that the early reading ability of students on the subject of this research did not differ factually.
Data on the results of the early ability of the students above were analyzed by the researchers using the SPSS program to get a significant picture of the children's early reading ability between students in the control class and students in the experimental class. The results of the independent sample t-test with the assistance of the SPSS program can be seen in Table 3 below: Based on the results of the data processing output from SPSS in the table group statistics above, information can be obtained that there were 20 students in the experimental class who were treated with the gamification method having a mean score in early reading ability of 70.1. As for the control class that was subjected to the BCCT method, the mean was 65.8. In addition, when we look at the information in the independent sample test, it shows that it has a score of 0.702 (early reading ability results) in Sig Levene's test, where the value is greater than 0.05. Therefore, it can be concluded that there is no difference in diversity (variance) between the early reading ability in the control group and the experimental group. Thus, according to the rules, it can be tested by an independent t-test by drawing assumptions of homogeneous data.
When it was known that the variety of data obtained was homogeneous/equal variance assumed and to find out whether there is any difference, the following step was to conduct a statistical test using the t-test which is an independent sample t-test. Table 3 shows that the results of the early reading ability (pretest) between the control group and the experimental group had a significance value of 0.441 where (p > 0.05, accepting H O ), which can be interpreted that there is no significant difference in the effect of the early reading ability (pretest) results between the control class and the experimental class. It is intended that, before students are given treatment in the form of the gamification method and the BCCT method, there is a score of early reading ability outcomes in the control class and the experimental class which do not have significant differences or in other words relatively the same.

Research hypothesis testing
There are three hypotheses that have been determined in this research. Hypothesis testing is carried out when all requirements are met. The hypotheses that have been established in this research can be presented as follows: (1) There is a difference in early reading ability between students taught using the gamification method and those taught using the BCCT method.
(2) There is a difference in the early reading ability between students with Cognitive Style-Field Dependence (CS-FD) and students with Cognitive Style-Field Independence (CSFI).
(3) There is an interaction between the gamification method and cognitive style in early reading ability.
Hypothesis testing was conducted by analyzing the results of early reading ability data. The results of the calculations with the help of SPSS software resting on two-way analysis of variance at a significance level of 0.05 can be seen in Table 4: Hypothesis testing was done by grouping similar hypotheses to further facilitate analysis. Hypothesis 1, was grouped based on the learning method used. Hypothesis 2 was grouped based on cognitive style, while hypothesis 3 was grouped based on the interaction between learning methods and cognitive style. In this case, the data were tested by analyzing the pair of hypotheses 0 (H 0 ) and alternative hypotheses (H 1 ).

First hypothesis testing
The hypothesis tested in the first hypothesis testing was hypothesis 0 (H 0 ): There was no difference in early reading ability between students taught using the gamification method and those taught using the BCCT method. Meanwhile, the alternative hypothesis (H 1 ): There were differences in the early reading ability between students taught using the gamification method and those taught using the BCCT method.
Based on ANOVA test results, it shows that the learning method has an influence on the score of students' early reading ability. This is evidenced by looking at the data in Table 4 where the F score was 225.482 with a significance of p = 0.047 which is smaller than α 0.05 (p < 0.05), so rejecting H 0 and it can be concluded that there was a significant difference in the effect of the post-test score of early reading ability between the class given the gamification learning method (experimental class) and the class given the BCCT method (control class). This is reinforced by seeing the mean of students' early reading ability in two groups showing a mean of the results of the early reading ability using the 86.5 gamification method which has a higher mean than the control class taught using the BCCT method which is 81.2. From the two shreds of evidence that have been stated both stands on ANOVA test results and by seeing the mean of each class, where the class which is subjected to the gamification method has a higher mean than the class taught using the BCCT method. It can be concluded that in general early reading ability of students who learn to use the gamification method is better than the class taught using the BCCT method.

Second hypothesis testing
The second hypothesis tested was hypothesis 0 (H 0 ), i.e. there was no difference in early reading ability between students with Cognitive-Field Dependence Style and students with Cognitive-Field Independence Style. The alternative hypothesis (H 1 ), showed differences in the early reading ability between students with Cognitive-Field Dependence Style and students with Cognitive-Field Independence Style.
The ANOVA test results in Table 5, provides information that cognitive style also has an influence on students' early reading ability. This is evidenced by the F score for early reading ability based on cognitive style of 4.531 with a significance level of p = 0.40 which is smaller than the α score of 0.05 (p < 0.05), so rejecting H 0 . With the rejection of H 0 , then H 1 is accepted, which means that there is a significant difference in the effect of the post-test results on the early reading ability between students with Cognitive-Field Dependence Style and students who have Cognitive-Field Independence Style. The data were also strengthened by looking at the mean of early reading ability of students having a Cognitive-Field Independence style having an average reading ability of 86.6, higher than students with a Cognitive-Field Dependence Style of 81.8. Therefore, it can be concluded that the early reading ability with a Field Independence Cognitive style has a better early reading ability than the group of students with a Cognitive-Field Dependence Style.

Third hypothesis testing
The third hypothesis in this research was to test the H 0 hypothesis. There was no interaction effect between the gamification method and cognitive style on early reading ability. Meanwhile, the alternative hypothesis (H 1 ) had an interaction effect between the gamification method and cognitive style on early reading ability.
The results of the calculation of the data to test the third hypothesis about the effect of the interaction between the gamification method and cognitive style on early reading ability can be known by looking at the F count score and the significance score on the source method cognitive style in Table 4 The table provides information that is known in the interaction line between methods and cognitive style that it has an F score e of 4.786 with a significance of p = 0.35 which is smaller than the alpha score of 0.05 (<0.05). This means that H 0 is rejected and H 1 is accepted, so it can be concluded that there is an interaction effect between learning methods (gamification and BCCT) with CS (FD and FI) on early reading ability.
The above data are supported by the mean of the post-test for the early reading ability to read with the gamification method and the CS-FI which was 86.2, having a slightly higher difference compared to students with CS-FD with the gamification method which obtained an average of 85.
The mean of the post-test for the early reading ability with the BCCT method for students with CS-FI of 83 has a slightly higher difference compared to students with CS-FD who has a mean of 80.2. Therefore, the result of the difference between the interaction between learning methods and cognitive style gives information that there is an influence of interaction between the group of students with the gamification method and CS-FI and CS-FD and the group using the BCCT and CS-FI and CS-FD method.
The influence magnitude of learning methods (gamification and BCCT) with cognitive style (FD and FI) on early reading ability is R Squared is equal to 47%, which means that the variability of the post-test results of early reading ability explained by the learning method (gamification and BCCT) with cognitive style (FD and FI) and the interaction between the two was 47%, while the remaining 53% was influenced by factors other than cognitive methods and styles.
The above data are supported by the estimated marginal means of early reading ability based on the interaction of learning methods (gamification and BCCT) and cognitive style (FD and FI) as follows: Based on the estimated average marginal means in Figure 5 above, it can be seen that the mean of the early reading ability of the group of students with both gamification methods that have CS-FD and CS-FI has a higher value than the group of students with BCCT method both CS-FD and CS-FI. In addition, the interaction graph above shows that the two lines have shown interactions showing two intersecting lines (there are interactions), so this reinforces the ANOVA test findings that show that there is an interaction effect between the gamification learning method and BCCT learning method with CS-FD and CS-FI on the early reading ability. This is due to a significant difference in the effect of the early reading ability between the group of students taught by the gamification method and CS-FD and CS-FI, and the group with the gamification method and BCCT with CS-FD and CS-FI.

Discussion
The results of this research are supported by research data conducted by Hong (Hong et al., 2012), taking the title of research on the effects of digital puzzle games on cognitive style for primary school students. This research used the Group Embedded Figure Test instrument to measure the tendency of cognitive styles possessed by students. From the calculation results of the Jigsaw digital puzzle game, it was seen that children with a field independence cognitive style scored better than children with a field dependence cognitive style.

Figure 5. Estimated Marginal Means Post Test
Research on the superiority of field independence in supporting students learning outcomes is also stated by Guisande (Guisande et al., 2007). His research was conducted on 149 primary school children in Portugal. In his research, he measured several learning outcomes, such as mastery of symbols, mastery of literacy, and mastery of art. From the results of tests, students with a fieldindependence cognitive style were superior to those with a field dependence cognitive style.
There was an influence of interaction between gamification learning method and BCCT learning method with field independence and field dependence cognitive style. This can be proven with a significance score of less than 0.05, and the mean of the early reading ability using the gamification method for students with a field independence and field dependence cognitive style was higher than the group of students taught using the BCCT method for the group of students with a field dependence and field independence cognitive style.

Limitations
Limitations of this study include the potential for biases related to participant self-selection. Another limitation was that the group participants had a lower level of health literacy skills than anticipated, which provided linguistic challenges in presenting the educational content and in collecting the quantitative data. This is not a longitudinal study. The Children Embedded Figure Test asks student to reflect on their cognitive styles which may have occurred many years ago.
Finally, while this study has a number of implications for educational practice, caution should be taken in their application, as several limitations exist in the study. The results of this study may not be generalizable to populations that differ significantly from the sample including those at different levels of cognitive styles, or more diverse populations than the sample studied. Moreover, this study just reached in part of small area. Therefore, better yet, the authors can do the testing to a wider area. Then in the future plan, the authors will fix the content and design of the gamification method for the further development if needed.

Conclusion
Based on the scope of the study of experimental research activities that have been carried out and data processing activities and discussion of the results of the research, a conclusion can be drawn as follows: There was a significant difference in the early reading ability between the group subjected to gamification learning methods and the group of students taught using the Beyond Centre and Circle Times (BCCT) learning method. This can be observed from the significance value of less than 0.05, and the mean of the early reading ability using the gamification method was higher than the mean of students using the BCCT learning method.
There was a significant difference in the early reading ability of students with a fieldindependence cognitive style and students with a field-dependence cognitive style. This can be proven from the significance value of less than 0.05, and the mean of students' early reading ability with a field independence cognitive style was higher than students with a field dependence cognitive style.
Gamification in education sectors has increased in popularity, but systematic studies to assess and evaluate its impact on learning are limited in the infancy stage. We encourage practitioners and researchers to use a design science approach to evaluate the impact of gamification in education. Specifically, they can develop gamified educational applications and systematically evaluate them using scientific approaches and methodologies such as experiments and surveys.