Improving literacy skills using the serious game model compatible with brain-based learning

Abstract

Conventional learning can be appealing if it incorporates appropriate technological elements. Nevertheless, most teaching aids are unsuitable for students with low cognitive ability. Serious games are some technological mediums that can attract interest and have proven effective. However, education experts do not design all games to meet these students' needs. The recent COVID-19 pandemic has caused teachers to face some difficulties in teaching, as conventional methods were less practical during lockdown. To incorporate games as teaching aids, game designers need a teaching model to help them design appropriate games for students with low cognitive ability. Therefore, this study aimed to validate a serious game model that aligns with brain-based learning (BBL) principles. The design and development method guided the study's conduct. The serious game model was implemented by developing a Mari Membaca (Let Us Read) prototype. This model was validated based on its usability, satisfaction with the BBL strategy, and the model itself. The usability results showed that the developed software has an easy-to-remember, learn, and use interface, a decent content presentation, feedback, an attractive design, and appropriate media integration. The user satisfaction assessment revealed a high level of satisfaction. The final validation ratings for all model components were satisfactory and significant. This study greatly benefits from the serious game model, which implements the BBL strategy and includes a prototype interface design and a consumer assessment instrument.

1 Introduction

Using technology as a teaching aid in teaching and learning activities (T&L) is not a new trend in Malaysia, as discussed in numerous studies since 2013 [1]. However, the recent COVID-19 pandemic has encouraged the international community, including Malaysia, to use digital technology more widely, indirectly leading to the Malaysian digital boom. In an interview on the Teacher’s Day Special talk show with the Senior Minister of Education, Dr. Radzi Jidin [2] mentioned that digitisation methods in teaching and learning require a fairly broad policy to advance the preparation of competency structure training. Teachers require opportunities to learn and must take the initiative to enhance their knowledge and digital skills to convey what they know effectively. Teachers must prepare themselves, as Generation Z and Alpha students are naturally interested in technology. Serious games are among the numerous games that attract the attention of teachers and students. Most researchers agree that games can have numerous benefits since they can increase students’ motivation and engagement [3, 4] if the time spent is controlled to avoid addiction [5,6,7].

Chen and Hsu [73] asserted that previous studies have utilized educationally designed serious games to enhance language and content knowledge. This statement was supported by Siti Norliza et al. [8], Azrulhizam et al. [9], and Nazry and Romano [10], who agreed that serious games have been constructed with the specific purpose of achieving learning objectives rather than focusing on entertainment. The elements obtained of the proposed serious game model in this study using the following methods: The Systematic Literature Review (SLR) and preliminary studies used to gather data through interviews and observations at three schools that offered the Rehabilitation Class. A rehabilitation class is a class of students who failed to pass the construct grade score of 6 out of 12 constructs. These students normally struggle to master basic reading, writing, and counting skills.

Zainal Abidin et al. [11] conducted a preliminary study involving interviews and observations to validate the elements obtained in the literature review and to find suitable elements that would suit slow learner students. The interviews and observations were conducted in four phases. Phase 1 established an objective for the preliminary study by examining game elements, pedagogy, and brain-based learning that align with the serious game model in the literature. Phase 2 involved examining evidence from previous studies to find support for the study's findings. Subsequently, one interview instrument was developed by adapting the instruments from several related journals. In Phase 3, the students were observed in the rehabilitation class, and the teachers were interviewed to collect data. In Phase 4, the collected data were analysed to develop a game model.

In the interviews, this study found that previous game software used by the teachers in rehabilitation classes does not focus on the learning element and does not follow the standard level of students. The students in these rehabilitation classes (7 to 9 years old) have been using game software for pre-school (4 to 6 years old). Previously, the LINUS program stopped using games due to the lack of interactivity and content that did not suit the requirements of rehabilitation classes. The LINUS programme was created to focus on students who failed six constructs, whereas the rehabilitation class is focused on students who failed constructs 7 to 12 during the examination. However, starting in 2019, this program was terminated, and the focus was only on the rehabilitation class. LINUS (Literation and Numeration Program) is under the Ministry of Education Malaysia (MOE), the State Department and the District Office of Education. This program is to help students to read, write, and count starting at the age of seven. No specific game software has been created for students in rehabilitation classes in Malaysia. Most game software is only focused on the commercial elements of the game. Thus, students who play these games would lose interest and focus on learning [12, 13, 14].

Consequently, their learning objectives cannot be achieved [15, 16]. In a preliminary study, Zainal Abidin et al. [11] found that teachers have no specific software to use except for the 32-skill rehabilitation module issued by the Ministry of Education Malaysia in 2012 as a guide. Teachers use this module as a reference for teaching and learning (T&L) in rehabilitation classes. Teachers conduct T&L sessions conventionally, and no standard teaching aids in digital form are needed as long as they follow the modules.

Cognitive ability affects children’s learning activities. Cognitive skills, such as self-management, playing, and schooling, are important aspects of learning. Most games on the market do not emphasize students’ cognitive development and place less emphasis on learning [17]. Wan Rosliza and Faridah [18] stated that cognitive skills are the ability of an individual to think, express opinions, understand, and remember what is happening in their environment. According to Abdollah et al. [76], students who are slow learners can be divided into two categories. Students in the first category do not excel academically and may have difficulties and weaknesses in mastering three basic skills: reading, writing, and counting. These students are not included in the special learning category.

The second category consists of students who have to be placed in special education groups for learning disabilities and are taught by trained teachers. Their learning disabilities may include low cognitive ability, language, speaking, reading, counting, behavioural, and social development. Doctors must confirm these disabilities as the cause of their learning process being disrupted. This current study was focused on students with low cognitive ability in the first category. These students were facing learning difficulties in mastering basic reading, writing, and counting skills due to various factors and not facing disabilities. The selection of rehabilitation students was done carefully using two methods. First, the students were selected by their class teacher or subject teacher based on their achievement in reading, writing, counting, and their test results in the Malay and Mathematics subjects. Second, all students were asked to undergo a screening test to determine their eligibility to enter the rehabilitation class. Students who failed six out of 12 constructs were admitted to the rehabilitation class. These students also underwent a diagnostic test to detect learning problems and identify the cause of their failure to master a skill. Most of these children needed more consideration, and the intervention of technology and aiding tools to enhance their learning ability to be comparable to typical children [19].

Muppudathi [20] and Zakarneh et al. [77] stated that if appropriate approaches and methods are used, these underprivileged students have the potential to improve their academic skills. Usually, the approach of using games is very close to a child’s soul. This play approach in education for underprivileged students can prevent them from feeling isolated and marginalised [21]. Cognitive, psychomotor, emotional, social, and psychosexual developments are closely related to children’s cognitive development, and thus, using games as part of their lessons is highly appropriate [22]. Research findings by Drigas, Kokkalia, and Lytras (2015) showed that software applications can help improve children’s literacy skills. In addition to using serious games as an interactive medium for students with low cognitive ability [23], the brain-based learning (BBL) approach also considers how the brain best learns. Teachers are encouraged to use this information when planning teaching strategies to increase motivation (Connell [78]) and student academic achievement.

The BBL approach is the most comfortable way of learning, as it does not burden the user's cognitive ability [24]. The BBL approach is closely related to cognitive ability involving mental activities, such as memory, categorisation, planning, reasoning, problem-solving, creation, and imagination [25]. BBL strategies can optimise left and right brain utilisation and activate the neuron cells. A serious game medium and the BBL approach can be combined to help students master the three lowest skills in Anderson's taxonomy: remembering, comprehension, and application. It is important for students with low cognitive ability to improve their cognitive skills before mastering medium and high-level skills. This statement was supported by Edelenbosch et al. [26], who stated that the Brain-based learning technique improved students' academic performance and helped them retain information longer. Outlining this paper starts with the introduction section, followed by the purpose of this research, literature review, methodology, results and discussion, and conclusion.

2 Purpose of study

The overall goal of this study was to contribute towards developing a better reading and learning model for students with low cognitive ability. The Serious Game and Brain-based Learning (SG-BBL) Model was developed using the design and development research (DDR) method. According to [27], DDR can be used to test the theory and applicability of a model or product in practice. Meanwhile, Richey and Klein [28] divided design and development studies into product and model reviews. The current study has focused on the model validation method through the Mari Membaca prototype development process, i.e., implementing the SG-BBL Model into a software application. Subsequently, this study aimed to answer the following research questions:

• RQ1: Are there any significant mean score differences in the level of early literacy skills among the rehabilitation students in the control group before and after undergoing conventional learning?

• RQ2: Are there any significant mean score differences in early literacy skills among the rehabilitation students in the experimental group before and after learning using the Mari Membaca prototype?

• RQ3: Is there a significant difference in mean scores of early literacy skill improvement between rehabilitation students who used the Mari Membaca prototype and those who used conventional learning?

Subsequently, three hypotheses were developed and tested in this efficacy evaluation test based on these research questions.

• H1: There is a significant difference in the mean score of the Rehabilitation students’ early literacy skills in the control group before and after undergoing conventional learning.

• H2: There is a significant difference in the mean score of the early literacy skills of rehabilitation students in the experimental group before and after learning using the Mari Membaca prototype.

• H3: There is a significant difference in the mean score of the early literacy skills improvement between rehabilitation students who used the Mari Membaca prototype and those who used conventional learning.

3 Literature review

This section will discuss the literature review of studies related to the effect of interactive computer applications on the reading skills of students with low cognitive ability, brain-based teaching approaches in learning, and the serious game conception model, with brain-based learning (BBL).// This section will discuss the literature review of studies on the effect of interactive computer applications on the reading skills of students with low cognitive ability, brain-based teaching approaches in learning, and the serious game conception model with brain-based learning (BBL).

3.1 Interactive computer applications on literacy skills of low cognitive students

Most computer application content, such as serious games, has been developed to help students master basic literacy skills. Good designs in serious game development can maintain interest and enhance students' theoretical learning abilities and skills. Appropriate selection of strategies, game components, and learning techniques for these students should be considered when developing game software. Previous studies have discussed the following positive aspects of using digital game software in learning: 1. Improves cognitive and motor skills and helps improve Information and Communications Technology (ICT) skills [22], 2. Learning is more effective [29], and 3. Increase students' attention to short-term convergence problems (Gillies and Boyle [79]). According to previous studies, serious games [30,31,32,33,34], Siti Syafawati and Farhana, [74, 35,36,37], interactive storytelling software applications [38], and interactive multimedia software [39] can be used as teaching aids by teachers to help students with reading problems. The development of computer software, especially games containing Malay reading literacy skills for students of low cognitive ability, needs to be improved. Most technologies are not designed and adapted to fit users' needs [40]. Several examples of serious game applications for teaching and learning are available on websites, applications, or compact discs, such as MyLINUS, Cerdik BM Series 1, LINUS Literacy Application (LiLIN), My Computer, and MENTOR Malay reading courseware using persuasive techniques. However, some strategies or approaches can be improved for students with low cognitive ability. Sangeeta [41] stated that besides having low cognition, some students might also have poor memory strength, short attention span, and difficulty expressing ideas or opinions.

The preliminary study (Zainal [11]) revealed several weaknesses in the game software for rehabilitation students. The game software does not use the KPM module as a reference in content development or selecting appropriate game elements. The software also has a long storyline, and the lack of interaction in the game can make students lose their focus. Other existing games also need storyline narratives,some have inappropriate narrative writing fonts. Selecting game elements to achieve learning objectives can be difficult [42, 43]. Alternatively, teachers must resort to various preschool game software on the market. The content and design of these game software are more focused on their commercial value rather than the need to meet the ministry syllabus. Students would focus more on the game, failing to meet the learning objectives [44]. Thus, this study aimed to contribute towards developing a better reading and learning model for students with low cognitive ability.

3.2 Brain-based teaching approach in learning

Brain-based learning (BBL) is an educational approach that considers how the brain acquires, processes, and interprets information and connects, stores, and recalls messages. BBL is based on neuroscience research on how human brains learn best naturally [45]. BBL is also based on the theory that learning will always occur if the human brain is not blocked from everyday processes compared to traditional teaching [46, 47]. Every learning involves the use of the brain since it is the most important organ in supporting daily life. The BBL approach implements the ‘Principles of Brain-based Learning’ developed based on theories and the latest findings in brain-related research. It was designed to be compatible with the optimal structure, tendencies, and functions of the human brain to ensure the effectiveness of students’ learning processes [46, 47]. Caine and Caine [47] constructed a list of 12 Principles of Brain-based Learning that serve as the theoretical foundation for BBL and offer guidelines and frameworks for teaching and learning (T&L). Many studies have shown that brain-based teaching approaches have effectively improved students’ conceptual understanding and learning motivation ([48, 49], Nik Zarini and Salmiza [53]). The BBL approach can vary depending on the time, the students, and the suitability of the learning environment. This T&L method allows students to improve their overall brain function based on three key elements: calmness and sensitivity, coordination of different experiences, and active processing of Physics subjects [50].

Most studies on BBL were implemented in the classroom, with teachers implementing their BBL methods or techniques. Fazrin and Salmiza [50] found that the BBL approach can effectively improve students’ motivation to learn Science compared to a group of students who followed the conventional instruction in the Biology subject for Form Four students focusing on Concepts and Mechanisms. Yulia Pratitis [51] stated that students’ potential can be optimally polished when learning the Japanese language by applying BBL. When the 12 principles of brain function in learning can be understood by the Japanese teachers, then, learning Japanese will be effective, fun, meaningful, and most importantly, both sides of the students’ brain can be optimally developed. Kiedinger [52] studied the effects of BBL on two elementary schools that were conducting Wisconsin exams. Their study confirmed that teachers who applied the BBL approach in the classroom have positively affected their students’ achievement in The Wisconsin Knowledge and Concepts Examination (WKCE). Teachers have been using various teaching techniques, pre-viewing information, practising proper hydration, frequent movements, creating a calm and sensitive environment for students, and continuous application of BBL strategies. Sani, Rochintaniawati, and Winarno (2019) have investigated the effect of BBL on students’ motivation in electrical circuit learning and found positive results. Their results were in line with the results of other studies, whereby BBL can increase students’ motivation in Physics (Nik Zarini and Salmiza [53]), Biology [50], and Chemistry [53]. Saleh and Mazlan [80, 54] studied the application of the BBL approach module, with i-Think and Brain Gym Integration (BBL- iTB), to improve the conceptual understanding and the motivation to learn Physics among matriculation college students in northern Malaysia. In addition to using appropriate teaching aids for students with low cognitive ability, the approach or strategy used is equally important. Most educators in numerous countries, including in Malaysia, are more comfortable using the traditional T&L approach, and less interested in the BBL approach. To have a positive effect on students, the implementation of the main components of learning and appropriate activities requires teachers with specific knowledge. In addition, educators also need to design appropriate learning activities. The time and energy required to perform these tasks are making most of them more comfortable with using traditional teaching approaches. Some teachers are also unaware of how to combine this strategy with technological teaching aids and of the impact of using this strategy in helping to improve the cognitive ability of rehabilitation students. Therefore, educators’ readiness to use this teaching approach can be a major constraint.

3.3 SG-BBL model

The software was built based on an appropriate model for applying BBL and is suitable for students with low cognitive ability. This study aimed to develop a Serious Game and Brain-based Learning (SG-BBL) model using the design and development approach, or the Design and Development Research (DDR) by Richey and Klein [28]. According to [27], DDR can be used to test the theory and applicability of a model or product in practice. The study of this model consisted of a design formation method and a game model validation method. The first category of serious game model design methods included the development of the SG-BBL Model, proposed component analysis consisting of gaming and learning components, brain-based learning strategies, and the initial SG-BBL Model validation by experts using the Delphi Technique. The second category was the SG-BBL Model validation method through the Mari Membaca prototype development, which was to implement the SG-BBL Model into a software application [11]. This study has focused on model validation methods through software prototype development and evaluation. The validation method consisted of two parts: developing the prototype and validating the prototype through four types of evaluation. The initial model has guided the prototype development through four phases: analysis, design, development, and evaluation. The first phase in the SG-BBL Model consisted of a needs analysis to identify the needs and content specifications of the model to be developed. During this phase, important information was collected to determine the design and development of the next phase of teaching materials [55]. The second phase was the design phase, whereby the researcher determined the learning objectives. During the design phase, the SG-BBL Model component was adapted with the instructional design element (ID) to produce the Mari Membaca prototype ID concept model. The third phase was the development phase.

During the development phase, low- and high-fidelity prototypes were produced. Content, storyboards, and flow charts were produced for the low-fidelity prototypes. The prototype interface was meticulously developed based on the criteria and content agreed upon during the design phase. Then, the authoring and programming of prototypes were produced for the high-fidelity prototypes. In the final activity of the development phase, alpha and beta testing were conducted to ensure that the prototypes worked. The last phase in the software development process was the evaluation phase. A serious game evaluation was performed to validate the model. This study performed a formative assessment to build a useful, serious game model. Dick and Cary [81] stated that a developed product should be tested with at least one user, as making improvements will make a significant difference. While researchers conduct formative assessments for developing serious games to ensure no unexpected problems, changes must be made according to user demand to solve problems users face when using software [56]. Several evaluations were performed to test the effectiveness of the SG-BBL Model implemented in the Mari Membaca prototype. The evaluations were conducted at Sekolah Rendah Kebangsaan Semeling, Kedah to validate the developed model:

• Prototype usability evaluation

• Satisfaction evaluation of BBL strategy use

• Final validation evaluation of the SG-BBL Model components

Figure 1 shows the validated SG-BBL Model, with the BBL approach added to the environment and the game. The seven teaching phases of BBL, as adapted from Sousa [57], have been listed outside the core model environment for teachers to practice before, during, and after their teaching session using this serious game environment. The BRAINBASED strategy [58] is a component of PBO that is applied in serious games together with learning and game components. The BRAINBASED teaching strategy was introduced to create a teaching session to stimulate the brain and improve memory. The brain may struggle to remember words, names, equations, vocabulary, and facts. These ten strategies can strengthen the neurons between the right and left brain for a longer memory span. The game and learning components obtained from the systematic literature review and preliminary study have been confirmed as being appropriate by expert teachers through the Delphi Technique and the final model validation testing. Each element and learning component in this game has also been adapted with BRAINBASED strategies to produce a balanced and collaborative model for maximum impact. Adding BBL strategy into the game model was a novel attempt to stand out compared to other serious game models. Section 5, Results and Discussion, will explain every element of this model further.

Fig. 1

Validated SG-BBL Model

4 Method

This section will describe the development of the research instrument, respondents’ involvement, and the methods used in data analysis.

4.1 Development of research instrument

This section will describe the development of the research instrument, the respondents involved, and the methods used in data analysis. This study has conducted four assessments: usability, satisfaction using BBL strategies, SG-BBL final validation model, and effectiveness. Before the commencement of the actual study, the completed questionnaire was tested during the pilot study, and a Delphi test was performed by experts consisting of nine LINUS and rehabilitation teachers. The actual study involved six Rehabilitation teachers and rehabilitation students. Components that scored less than the mean of 2.33 were deemed the low category [59] and were rejected.

4.1.1 Usability assessment instrument of Mari Membaca Prototype

The development of the questionnaire instrument involved two steps, namely, the instrument construction process and the content validation process. A 5-point Likert scale was used in the instrument, and the respondents had to select the following responses: “1 = Strongly Disagree”, “2 = Disagree”, “3 = Not Sure”, “4 = Agree”, and “5 = Strongly Agree”. The construct elements have been identified before developing this questionnaire instrument. Several questionnaire instruments from previous studies were compared, for example, the Questionnaire for User Interaction Satisfaction 7.0 (QUIS) [60], Nielsen’s Attributes of Usability (NAU) (Nielsen [75]), Practical Heuristics for Usability Evaluation (PHUE) (Perlman [82]) and USE Questionnaire [61]. The developed instrument contained seven main design constructs, and the evaluation items for each construct were modified to suit the usability of this instrument. All constructs and items were processed from previous research questionnaires. Before it was used in the final evaluation, this instrument went through several processes of revision and modification in terms of text, wording, layout, and the addition and subtraction of items.

Among the constructs assessed in the usability questionnaire were rememberability, learnability and ease of use, content presentation, feedback, interface and screen design, media integration, and perception of overall satisfaction, as adapted from Nielsen [62], Brinck, Gergle, and Wood [63], Hollins [64], Nor Azan, Azizah, and Yue [65], and Perlman [82]. Subsequently, seven constructs were included in the usability instrument, as follows: i) Easy to remember (2 items)—If a player stops playing for a while and then, plays again, the player does not have to relearn the game because the software design helps the user to remember easily [62],ii) Learning ability and ease of use (8 items)—Users can easily perform activities, even if it is their first time using the software [62], which allows new users to learn quickly Brinck, Gergle, & Wood [63],iii) Content presentation (14 items)—The desired content and information delivered in the application is effective [64],iv) Feedback (4 items)—The software design can help users interact when they make mistakes (Perlman [82]); v) Interface and screen design (5 items)—An attractive interface design can help convey the content easier for better understanding and effectiveness [64],vi) Content media integration presentation (4 items)—The form of content and information conveyed in the software was effective [62], and the use of various media in the software design can strengthen the concept of learning [66],and vii Perception of overall software satisfaction (4 items—Users were satisfied when using the software design [62].

4.1.2 Content validity test of questionnaire instrument

Five experts were involved in the instrument content validation process in this study. Previous studies involved at least three experts for content validity [66, 67]. Validity means using different test methods to measure the same feature to the maximum [68]. The Content Validity Ratio, better known as CVR, is often used to empirically measure the content validity of more appropriate items in terms of time and money (Tojib and Sugianto, [83]). Local and foreign researchers have used the CVR method, as introduced by Lawshe in 1974, as the first step in the instrument preparation process (Doustmohammadian et al. [84]; Matore et al. [85]). This method uses calculations to measure content validity, which helps researchers decide whether to retain or drop each item in the instrument to ensure that the retained items represent the main construct appropriately. According to Lawshe [69], in CVR, the expert must evaluate each item based on a three-point scale according to the level of importance, namely, (1) very important, (2) useful but not important, and (3) unnecessary. To perform CVR calculation, the following Eq. 1 can be used, where CVR is the value of the item content validity ratio. It represents the number of expert panels involved, which gives the item evaluation importance, and N is the total number of expert panels. Once all CVR values have been calculated for all items, the Content Validity Index (CVI) values can be calculated for the entire instrument [69].

$$\text{CVR}=\frac{\text{ne}-\left(\text{N}/2\right)}{\text{N}/2}$$

(1)

The calculated CVR value must be in the range of -1 to + 1. In the content validation process, + 1 refers to any value close to + 1, which means that the expert panel agrees that the item is very important. If the CVR value equals zero (CVR = 0), half of the expert panel believes that the measurement item is very important, while the other half agrees that it is unimportant. CVR values higher than zero (CVR > 0) indicate that half of the expert panel believes most measurement items meet facial validity [70]. The CVR value for each item was then compared with the critical value. Lawshe [69] stated that if more than half of the experts rated an item as very important, then this item was considered to have met the validity of the content. According to Almanasreh et al. [86], the critical value of this CVR method is based on the number of experts involved. Table 1 shows that the critical value set for five experts is 1.00. Therefore, if the CVR value of an item acquires a value that is lesser than the critical value, this item needs to be modified or dropped from the instrument.

Table 1 Minimum CVR value for one-tail test at p = 0.05

4.1.3 Satisfaction assessment instrument for the use of BBL strategies in Mari Membaca Prototype

As previously discussed, a questionnaire instrument was used to measure the respondent’s perception of the methods practised in the Mari Membaca prototype in a serious game environment. These items have been constructed based on the validated preliminary studies. Nine aspects of this strategy for learning language literacy skills have been assessed. Respondents among the teachers have been asked to answer using the following 5-point Likert scale to evaluate each item: “1 = Strongly Disagree”, “2 = Disagree”, “3 = Not Sure”, “4 = Agree”, and “5 = Strongly Agree”. For respondents among the students, this questionnaire uses pictorial icons to represent points in the Likert scale, as shown in Fig. 2. This Likert scale that used 1–5 pictorial icons made it easier for the respondents to express their feelings.

Fig. 2

Likert scale represented by pictorial icons

Questions were asked of the respondents among the students using simple, easy-to-understand language, and they were required to select the preferred icon scale that represented whether they strongly agreed, agreed, did not know, disagreed, or strongly disagreed. Figure 3 shows an example of the usage of pictorial icons when this study was conducted.

Fig. 3

Pictorial icons used by the rehabilitation students

4.1.4 Instrument for the final validation assessment of the SG-BBL Model

Questionnaire instruments were used to evaluate and validate the components in the SG-BBL Model. These instrument items were constructed based on the literature review and validated by experts, including nine LINUS and rehabilitation teachers. The Component Validation Questionnaire for this model was distributed to six teachers (two rehabilitation teachers and four language teachers). This instrument used a 5-point Likert scale to evaluate each item, with feedback options as follows: “1 = Strongly Disagree”, “2 = Disagree”, “3 = Not Sure”, and “4 = Agree”, and “5 = Strongly Agree”.

4.1.5 Effectiveness evaluation instrument for the Mari Membaca Prototype

The effectiveness evaluation used the Mari Membaca prototype and the pre-test and post-test question sets. These screening test questions have been edited to select students from the rehabilitation class. Then, the rehabilitation teacher edited and validated these questions to suit the prototype content taught over the study's five weeks.

4.2 Research sample

The study sample can be divided into two categories: teachers and students in rehabilitation classes. The sample of teachers involved six teachers, which consisted of two rehabilitation option teachers and four Malay language teachers, who were experienced in teaching the LINUS programme before it was discontinued in 2020. A total of 22 respondents were in the Grade 2 rehabilitation class (8 years old) who had undergone the student selection screening test. Students who failed the screening test and the marks set by the Ministry of Education are placed in this rehabilitation class. The rehabilitation respondents consisted of seven boys and four girls. The assessment began at the beginning of September until the beginning of October 2020, which was a period of five weeks: one week for the pre-test, three weeks for the teaching and learning intervention, and one week for the post-test. The question set contained four sections: letter modules, open phrases, word phrases, and simple sentences. The next four assessments involved other teachers and students. A total of six teachers and 11 students in the experimental group participated in the BBL strategy satisfaction assessment. As for the usability assessment, 22 students participated.

Only six teachers were involved in the usability evaluation and final validation of the model. They were asked to assess the prototype’s level of usability using the constructed instruments. The limited cognitive ability of the rehabilitation students made it difficult for them to understand the instrument's content. On the other hand, the satisfaction assessment involved teachers and remedial students who learned using the Mari Membaca prototype. Teachers were fine with the task of answering the questionnaire. However, the rehabilitation students found it difficult to understand the questions. Thus, the researcher used the pictorial icon initiative to replace the Likert scale. The final validation assessment involved only six teachers to validate each component and element in the final model to match the students’ cognitive abilities.

Meanwhile, to evaluate the usability of this model in the experimental study, 22 students in Year 2 who attended the rehabilitation class at Sekolah Rendah Kebangsaan Semeling, Kedah, were chosen. They were selected based on their Malay language assessment test. Students who failed the 12 constructs were placed in the rehabilitation class. Usually, the number of students is 15 in one class. Nevertheless, the relatively high number of students in this class during this study was due to the home teaching and learning factor that had to be implemented during the pandemic that hit Malaysia starting in February 2020. This school is located in a rural area, and most students experienced various problems, especially the lack of tools and internet access to run a perfect T&L at home. Roscoe [87] and Sekaran [88] stated that at least 10 to 20 study samples are required for a simple experimental study involving comparisons. All students were placed in one class to facilitate the test. The sample population was selected using the purposive sampling method, and these students were randomly divided into the experimental and control groups. The sample size for both groups was 11. This study conducted two types of t-tests, the paired t-test and the independent sample t-test, which meant that both datasets were normally distributed, with p-values higher than 0.05.

4.3 Data analysis

This section discusses the data analysis methods for all assessments. Pre- and post-tests for semi-experimental studies were performed using data collection methods for efficacy evaluation. Statistical Package for Social Science (SPSS) version 16.0 was used. This study used two data analysis methods, namely descriptive and inferential statistical data analysis. For the usability assessment, satisfaction was assessed using the BBL strategy, and the final model validation was done using descriptive statistical data analysis to determine the mean and standard deviation. Usability evaluation involved analysing inferential statistical data using a t-test. The t-test is often used to compare, test, and predict data. The mean score was interpreted using the three categories proposed by Mohammad Najib [59]. The first category was Low, with a mean of 1.00–2.33; the second category was Middle, with a mean of 2.34–3.67; and the last was Top, with a mean of 3.68–5.00.

This study used two types of t-tests: the paired-sample t-test and the independent sample t-test, where both datasets were normally distributed, with p-values higher than 0.05. Paired t-tests were used to obtain the overall achievement scores of both groups by measuring the differences in the pre-test and post-test scores after the intervention was conducted. The study sample was relevant because they were the same respondents. An independent sample t-test was used to compare the overall scores between the experimental and control groups in the pre-test and post-test. Figure 4 summarises the analysis of the usability of reading tests.

Fig. 4

Analysis summary of the usability of reading tests

4.3.1 Analysis of usability assessment of Mari Membaca Prototype

Once the respondents had used the Mari Membaca prototype, a usability assessment questionnaire was distributed to the teachers. The completed questionnaires were collected, and the descriptive data collected from these questionnaires were analysed using mean and standard deviation.

• 1. Data Analysis of BBL Strategy Satisfaction Assessment.

These data were taken in the last week of the intervention sessions with the teachers and students in the rehabilitation class. Respondents among the teachers received the questionnaires as usual, while respondents among the students were called to answer the satisfaction questionnaire to determine their level of satisfaction. Using this strategy, two rehabilitation teachers assisted the researcher in completing the data collection for the satisfaction assessment. These respondents must answer by holding up a pictorial icon to indicate their level of agreement. The researcher recorded these choices on their behalf because they were not proficient in reading and understanding the questions in the given instrument. Quantitative analysis through SPSS was performed after all data had been collected from respondents who attended the teaching and learning interventions. The obtained quantitative data were analysed in the form of means and standard deviations.

• 2. Final Validation Evaluation Data Analysis of the SG-BBL Model.

Each respondent among the teachers was given a questionnaire for the assessment and was asked to score the components of the SG-BBL Model. Similar to the usability analysis, quantitative data from these respondents were collected and analysed to obtain and validate the final components of the SG-BBL Model. The data were analysed in the form of mean and standard deviation.

• 3. Effectiveness Assessment Data Analysis of Mari Membaca Prototype.

The quasi-experimental design used in this study was a pre-and post-test design for non-equivalent groups (Chua [89]; Creswell [90]; Wiersma and Jurs [91]). This quasi-experimental design was based on unequal groups (Campbell and Stanley, [92]) and did not involve a random selection of respondents (Fraenkel and Wallen [93]; Lim [94]). The results assessed the level of reading skills of the rehabilitation students by comparing their pre-and post-test scores. The pre-and post-test questions were the same, i.e., taken from the teacher's monthly tests and modified according to the content in the serious game software. A descriptive analysis of mean frequency was conducted to determine the distribution of literacy skills based on the comprehension scores obtained by the experimental and control groups in the pre-and post-tests. Data were analysed before the intervention session to determine the students' literacy skills. Pre-test sets were distributed to the students, and the intervention was initiated after completion. According to Pallant [95], through pre-and post-tests, an independent sample t-test was conducted to determine significant mean differences in the students' literacy levels in the experimental and control groups. A paired sample t-test was performed to compare the same group's mean pre-and post-test scores. As a non-directional test, the probability level for these tests was α = 0.05.

5 Results and discussion

This section discusses the prototype's development, the SG-BBL Model's implementation, and validation based on tests on the Mari Membaca prototype.

5.1 Development of Mari Membaca Prototype and implementation of the PS-PBO Model

This prototype was developed through a design and development research approach (DDR) within four comprehensive phases: analysis, design, development, and evaluation. Applying the 10 PBO strategies and the seven teaching strategies has been coordinated with integrating multimedia in serious games. The main focus of this prototype was to apply the PBO strategy to a serious game to optimise the use of students’ brains. The concept of learning using this prototype was independent, meaning that students could learn according to the appropriate place and time. However, student development must be regulated by teachers, parents, or guardians. Explanations of every element used in this model are listed as follows:

• 1. Activation

  • The activation method can activate the student’s memory and prior knowledge to stimulate information transmission. This technique can be effective in preparing students to receive new knowledge. Teachers can ask students (Granado et al., [96]) about the topic that will be taught in general before starting the class.

• 2. Learning outcomes

  • The learning outcomes that need to be achieved and the learning process involved can be explained as follows: student’s existing knowledge or conceptions are activated to stimulate learning of a new concept. Teachers can inform students of the objectives and what needs to be achieved from learning this topic.

• 3. Making connections between old and new knowledge

  • This strategy connects newly acquired knowledge with what was previously known. In this game application, the developed modules were based on the progression from low difficulty levels (Letter Module) to difficult (Simple Sentence Module). Each of these modules is related to each other.

• 4. Conduct learning activities

  • Direct student involvement in every learning activity can stimulate active processing and generate learning effectiveness. The seven-phase teaching strategy and BRAINBASED strategy have been applied to the learning and game components to keep students interested and focused on learning while playing.

• 5. Demonstration of students’ understanding of new skills

  • Demonstration activities allowed these students to use new knowledge or skills in new situations. This process gave the students time to adapt to the newly acquired concepts and can indirectly strengthen their conceptual understanding.

• 6. Evaluation of students’ understanding

  • To assess the student’s level of understanding and acceptance of the learned concepts, the teacher can conduct evaluation and closing activities. The teacher can perform a simple test to evaluate their understanding and memory and openly ask questions about what has been learned.

• 7. Preview new topics

  • Previewing the topic to be studied can prepare the students’ brains for the upcoming learning task and help improve the effectiveness of the learning process. Before the class ends, the teacher can inform the students of the new topic discussed in the next lesson.

• 8. Learning Objectives

  • Learning objectives explain what the students need to achieve at the end of a lesson unit. These learning outcomes include literacy skills and student motivation.

• 9. Cognitive theory of multimedia learning and constructivism

  • Adapting learning theory in serious game applications can provide an appropriate learning environment and meet students’ diverse learning styles. The cognitive learning theory of multimedia learning and constructivism are two theories to be included in the game after the initial study was conducted.

• 10. Instructional Content Design

  • The involvement of teachers as educators and students as recipients of knowledge was considered extremely important when designing appropriate content to ensure that the software development goals can be achieved. According to Boton and Forgues [71], the course content needs to meet students’ needs, which can lead to a loss of interest and demotivation to continue learning among students if these needs are not met.

• 11. Organisational Structure Design

  • The learning objectives for each activity were displayed on the front page of each activity. The selected learning content has been organised into small units. The objectives at the beginning of each activity coincided with the BBL strategy, which explained the learning outcomes to be achieved.

• 12. Content Delivery Design

  • Interface and screen design were among the activities involved in the content delivery design. According to Harith Azidin [72], interface design has less to do with the aesthetic value, look, and feel of the design elements, such as colour, text, images, and layout, and instead more of a communication medium that acts as an intermediary between users and systems.

• 13. Serious Game Technology

  • The Mari Membaca software is an interactive multimedia product where information is displayed through a screen, and users can see, hear, interact, and control how to display the information. This software presents links to allow users to switch from one interface to another. Users only need to click on certain areas with links to see other interfaces at any time.

• 14. BRAINBASED Strategies

  • The SG-BBL prototype was designed based on three techniques related to brain-based learning (BBL): i) immersive orchestration, ii) a quiet environment, and iii) active processing. The SG-BBL prototype was also included with the seven phases of BBL teaching [57] and the B.R.A.I.N.B.A.S.E.D BBL Strategy by Hileman [58]. This strategy would be suitable for increasing the use of the brains of the rehabilitation students.

• 15. Game Objective

  • The learning objective for this game was to improve the knowledge and literacy skills of students with low cognitive ability. Students should know what they are going to learn. Therefore, in this prototype application, the formulated objectives were clearly stated to the students on each front page of the game before the start of the activity.

• 16. Feedback and evaluation

  • The use of the Mari Membaca application can give feedback answers immediately. In this serious game, corrective feedback can aid learning and provide guidance. This corrective feedback would inform the students whether their actions were wrong, provide knowledge of the correct response, and show the correct answer to the student.

• 17. Interactivity

  • Interactivity involves a dialogue between the computer and the user, and the user can interact with the teaching material. In a game, the requirement to solve a problem or quest involves interactions between the player and the game. The game allows players to socialise either via player-to-player or player-to-computer.

• 18. Rules

  • In this game, students will be guided by the teacher, which is a quick way to start the tutorial to start the game. Chee (2006) stated that there are two types of rules: the Paidea and Ludus rules. This game has incorporated the Paidea rules, i.e., no results are obtained. Although these rules do not state who wins or loses, an assessment is available to determine whether the students have improved their literacy skills after playing the game.

• 19. Challenge

  • Challenging games will usually attract more students to get involved. Students will quickly feel bored if the challenge is too easy and quickly give up if it is too difficult. In this Mari Membaca application, students will be assisted with audio to maintain their motivation to continue playing.

• 20. Fantasy

  • Fantasy refers to the representational aspect of a game, regardless of whether the depiction is realistic or not; it is used to distinguish it from abstract games. In this application, the theme and storyline were based on fantasy stories where the characters were anthropomorphic animals that could speak, have emotions, and think like humans.

In this Mari Membaca game, students were exposed to the following learning modules: Letter Module, Open Syllable Module, Word Phrases, and Simple Sentences. The evaluation component of student performance was developed through these learning modules, i.e., each answered module will be given a score. The teachers were also monitoring the students during the evaluation. Story and Song Modules were included to encourage students to enjoy learning using this game software. Students can sing the ABC song and watch the included short story. A brief introduction to this game software can be found for teachers and parents in the About M2M section.

Figure 5 shows the structure of the teaching organisation and navigation for the Mari Membaca prototype. The use of navigation as a flow of presentation can help students to choose the desired learning module. Meanwhile, Fig. 6 and Fig. 7 show examples of the game interface.

Fig. 5

The structure of the teaching organisation and navigation of the Mari Membaca prototype

Fig. 6

The main menu interface of the Mari Membaca prototype

Fig. 7

Interface of the drag and drop activity

6 Findings of the instrument-based usability assessment

Table 2 shows the results of the usability assessment based on the questionnaire data obtained from a sample of teachers.

Table 2 Mean scores of the usability assessment construct

Overall, the usability assessment of these seven prototype constructs showed positive feedback from all respondents, with a mean for the entire construct exceeding 4.00. According to Najib [59], a mean score of 4.00 is a high category where almost all respondents agreed that the item is important.

7 Findings of the satisfaction assessment on the use of instrument-based BBL strategies

This section describes the findings of the satisfaction assessment of using brain-based learning strategies. This assessment was conducted among six teachers and 11 rehabilitation students in the experimental group who studied using the Mari Membaca prototype in a serious game environment. The satisfaction results of the BBL strategy usage were based on the analysis of the questionnaire instrument that measured the respondents’ perceptions of the overall methods practised in this prototype. Nine aspects of learning language literacy skills were assessed in this assessment. The respondents among the rehabilitation students and teachers responded positively to items 2, 4, 5, 6, 7, 8, and 9, with a mean score of 4.00 and above. This score belonged in the high category, where almost all respondents agreed that the item was important. However, for item 1, both groups of respondents gave the lowest mean, with the group of students giving a mean of 2.91 and the group of teachers giving a mean of 3.00, which were in the moderate category. For item 3, the group of students gave a mean of 3.91, and the teachers gave a mean of 4.33, which was in the high category. The complete analysis results are listed in Table 3.

Table 3 Perceptions of satisfaction among students and teachers on the use of the BBL strategy in the Mari Membaca prototype

8 Findings of the questionnaire-based final confirmation evaluation

The analysis results for the final validation study of the SG-BBL Model were obtained using questionnaires distributed to six teachers involved in the usability evaluation of the Mari Membaca prototype. This final validation assessment was required to validate the components of the SG-BBL Model, which were necessary and appropriate for developing game applications for rehabilitation students. The findings of the final instrument analysis are listed in Table 4.

Table 4 Final validation analysis of the SG-BBL Model components

According to these findings, the teachers responded incredibly positively, as the mean score for all elements was higher than 4.00. Their evaluation of the SG-BBL Model components applied in the Mari Membaca prototype gave mean scores that ranged between 4.00 and 5.00, which were very important. Almost all respondents agreed that these items were important. For the learning component, five of the seven elements, namely, setting learning objectives, adapting learning theory, teaching content design, teaching content structure design, and serious game technology and teaching media, recorded a mean score of 5.00, which belonged in the very high category where all respondents agreed that these were very important items. The element of brain-based learning strategy recorded a mean score of 4.67, while the content delivery design recorded a mean score of 4.83, which belonged in the high category. Thus, the respondents agreed that these items were important. In the game component, the objective setting and feedback elements recorded the highest mean score of 5.00, followed by the interactivity and rules at 4.67, which were high scores. The mean scores for the challenge and fantasy elements were 4.33 and 4.50, respectively. For the BRAINBASED component, six of the ten elements, namely, the brain’s period, repetition, active learning, image, colours, and vibrant appearance, recorded mean scores of 5.00, which were in the high category. Thus, all respondents agreed that these items were very important. The elements of something new and the social brain each scored 4.67. The indirect learning element recorded a mean score of 4.33, while the thinking skills development recorded a mean score of 4.00.

Based on the final validation analysis results, the SG-BBL model was refined. Thus, the refined SG-BBL components consisted of three main components, namely, learning components (learning objectives, learning theory, instructional content design, instructional content structure design, brain-based learning strategies, content delivery design, and serious game technology), serious game components (game objectives, feedback, interactivity design, rules, challenges, and fantasy), and BRAINBASED components (brain’s period, repetition, active learning, images, novelty, colourful, indirect learning, brain social, emotional outwardness, and developing thinking skills).

9 Findings of the pre-test experimental case study-based effectiveness evaluation study

A case study involving 22 Year 2 rehabilitation students was conducted to evaluate the effectiveness of the Mari Membaca prototype. Effectiveness assessments were conducted by comparing and evaluating the levels of literacy skills between two groups of students who underwent different learning methods. The pre-test instrument was adopted from the edited Level 1 reading screening test. The post-test instrument used the same set of instruments. The teachers agreed with the edited questions used in these instruments. This evaluation was also performed to support the research hypotheses. Table 5 shows the level of each early literacy skill set to determine the early literacy skills of rehabilitation students.

Table 5 Levels of Literacy Skills

The increase in literacy skills in the control group was 1.96, from 5.35 (pre-test) to 7.31 (post-test). According to Table 5, the early literacy skills of these rehabilitation students remained at the weak level (scored between 4 and 7), as no chance of increase or decrease was detected. Meanwhile, the increase in literacy skills in the experimental group was 6.14, from 5.20 (pre-test) to 11.34 (post-test). Table 5 also shows that the early literacy skills of the rehabilitation students have increased from a weak level (scores between 4 and 7) to a moderate level (scores between 8 and 11). Both the control group and the experimental group experienced an increase in skill levels. However, the descriptive statistical analysis findings showed that the average increase in the experimental group score of 6.14 was higher than in the control group score of 1.96.

However, these descriptive statistical analysis findings cannot be accepted and confirmed lightly (Kamarudin et al., [97]). Thus, to confirm whether there was a significant difference in the improvement of early literacy skills between the experimental group and the control group or vice versa, inferential statistical analysis (t-test) was conducted. The overall achievement scores in the pre-and post-tests between the experimental and control groups for the Reading Screening Test are shown in Tables 6 and 7.

Table 6 Overall achievement scores of the control group

Table 7 Overall achievement scores of the experimental group

According to these tables, the average score gained by the rehabilitation students for their level of early literacy skills is 5.35 for the control group and 5.20 for the experimental group. These results indicated that the rehabilitation students’ early literacy skills were weak (scored between 4 and 7).

Hypothesis 1

There were significant mean score differences between the rehabilitation students' early literacy skills in the control group before and after undergoing conventional learning.

• H₀1: There was no significant difference in the rehabilitation students' mean scores of early literacy skills in the control group before and after undergoing conventional learning.

• H₀1 µpost = µpre

• There was no difference between the mean scores of the post-test and the pre-test.

• H₀1 µpost > µpre

There was a difference between the post-test mean score and the pre-test mean score; the post-test mean score was greater than the pre-test mean score.

The analysis of paired t-test data for the control group’s mean scores (sample size, N = 11) showed that the mean pre-test score was 26.73, while the mean post-intervention test score was 37.09, an average difference of -10.364. The paired sample correlation table also showed that the pre-test and post-test scores were positively and significantly correlated (r = 0.862). The paired sample t-test table showed that the test results of the control group gave a significant value (2-tailed) (p = 0.001), and this value was smaller than alpha = 0.05 (p < 0.05). Statistically, it was rejected, and H_a was accepted, which showed a difference between the achievements in the post-test mean score and the pre-test mean score. The post-test mean score was greater than the pre-test mean score after these students completed their conventional learning.

The descriptive analysis also showed that the mean score of students’ early literacy skills has increased by 1.96 from 5.35 (pre-test) to 7.31 (post-test). This improvement was also significant when analysed using the paired t-test, which showed that rehabilitation students who underwent conventional learning had shown a significant increase in early literacy skills but remained at the weak level (scored between 4 and 7).

Hypothesis 2

There were significant mean score differences in the early literacy skills of rehabilitation students in the experimental group before and after learning using the Mari Membaca prototype.

• H₀2: There was no significant difference in the overall scores between pre-and post-test for the experimental group’s reading literacy skills after using the Mari Membaca prototype.

• H₀2 µpost = µpre

• There was no difference between the post-test and pre-test mean scores.

• H₀2 µpost > µpre

There was a difference between the student’s achievements in the post-test mean score and the pre-test mean score; the post-test mean score was greater than the pre-test mean score.

A paired t-test of the early literacy skills for the experimental group (sample size, N

p = 0.00) that was less than alpha = 0.05 (p < 0.05). Therefore, statistically, H₀ was rejected, and H_a was accepted, showing a difference between students’ achievement in the post-test and pre-test mean scores. Thus, the post-test mean score was greater than the pre-test mean score.

There was an increase of 6.14 in the mean score of students’ early literacy skills, from 5.2 (pre-test) to 11.34 (post-test). The paired t-tests have also shown a significant improvement in literacy skills. Thus, implementing the game method was shown to help improve the students’ initial literacy skill level from weak (scored between 4 and 7) to medium level (scored between 8 and 11).

Hypothesis 3

There was a significant difference in the mean scores of early literacy skill improvement between rehabilitation students who used the Mari Membaca prototype and those who used conventional learning.

• H₀3: There was no significant difference in the mean score of early literacy skill improvement between rehabilitation students who used the Mari Membaca prototype and those who used conventional learning.

The overall achievement differences between these two study groups based on the pre-test and post-test scores were observed using an independent t-test. The results of this analysis are listed in Table 8.

Table 8 Differences in overall achievement of the experimental and control groups in the pre-test and post-test

The comparative analysis of the pre-test for the experimental group showed an average score of 26.00 (n = 11, sd = 10.770) compared to the pre-test for the control group, which was 26.73 (n = 11, sd = 10.364). These results showed that p = 0.873 was greater than p = 0.05 for the pre-test. Therefore, H₀ was accepted, and Ha was rejected. This means there was no significant difference between the overall achievement of students who used the Mari Membaca prototype and those who used conventional learning during the pre-test conducted between experimental and control groups.

The comparative analysis of the post-test for the control group showed an average score of 56.73 (n = 11, sd = 20.145) compared to the post-test for the experimental group, which was 36.55 (n = 11, sd = 15.102). These results showed that for the post-test, the value of p = 0.015 was smaller than the value of p = 0.05. Therefore, H₀ was rejected, and Ha was accepted. This showed a significant difference between the overall achievement of students who used the Mari Membaca prototype and students who used conventional learning during the post-test conducted between the experimental and control groups.

Overall, the findings of this study only involved pre-test and post-test scores. These results did not consider other factors, such as student demographics, gender, age, religion, ethnicity, place of residence, parental education level, and socioeconomic status. These findings only showed that the improvements in reading literacy skills were significant for the experimental and control groups in the post-test and the pre-test.

10 Conclusion and contribution

This study was conducted to facilitate relevant parties in educational software development and students with low cognitive ability. This serious game software model can be a source of reference and guidance for other researchers to develop educational software for rehabilitation students. The findings of this study represented only one school. However, it can be a benchmark and a trigger for other researchers to research on a larger scale to describe better the state of cognitive ability among students throughout Malaysia.

Using serious games as teaching aids allows instructors to apply different teaching techniques, and students can exercise control over their learning sessions. In addition, serious games can help teachers diversify the use of technology in T&L as an appropriate alternative in the curriculum. This study could be an example of how early and ongoing assistance through a fun, serious game model could help improve students' literacy mastery and motivate rehabilitation students. For the upper echelons, including the Ministry of Education and schools, their goal of seeing zero illiterate students is getting closer when student achievement increases.

This study has contributed to the SG-BBL Model, which is important for designing a serious game application for students with low cognitive ability. The SG-BBL Model enables the production of quality serious game applications and is equipped with appropriate brain-based learning, games, and learning elements. The seven phases of brain-based learning in a play environment can be used through the concept of activation, explaining the learning outcomes to be achieved and the processes involved, making connections and constructing meaning, conducting learning activities, demonstrating student comprehension, re-evaluating student comprehension and memory, and pre-viewing topics that are newly implemented by teachers on students before, during, and after carrying out learning activities in a serious play environment. These seven phases of brain-based learning instructions can be advantageous for self-learning, educating students to solve problems while improving their cognitive skills. This software design can help other researchers produce effective and useful educational software for self-directed learning and online or offline collaboration with other teachers and students based on brain-based learning strategies.

In addition, the prototype interface design of Mari Membaca can guide other software designers when developing a serious game application specifically for rehabilitation students. The interface design that refers to the components of the SG-BBL Model allowed this serious game software to be developed according to the specifications and needs of the teachers. The consumer assessment instrument was developed through a comprehensive questionnaire, which included the instrument’s construction, validation, and reliability assessment. Therefore, this instrument can help other researchers apply it in their consumerism evaluation studies. The SG-BBL Model can be a reference and guide for software developers to design serious game applications to improve literacy skills effectively. However, this SG-BBL Model can be improved to produce fully functional software, such as game components.

The delivery method can be improved by adding game activities and game levels to add user fun. In addition to improving the activity module, it is recommended that tutorial modules or notes be added as a reference for players before starting any activity. The service component can also be considered by providing additional learning resources to students, such as administrative and technical support. For example, this software was designed with the support of different resources, i.e., the library, the Ministry of Education Malaysia, technical staff assistance, and other services.

Among others that can be improved is the study period. This study proposed that the final evaluation study should be conducted within at least six to 10 months to obtain more significant data. Long-term studies can assess the level of early literacy skills of rehabilitation students at the initial, intermediate, and final stages according to the student syllabus by the Ministry of Education. Further research should focus on factors other than literacy skills influencing literacy mastery, such as health, social, emotional, physical, and behavioural factors (Denham [98]; Mashburn et al. [99]).

11 Limitations and future directions

Several limitations were encountered during this study. The number of respondents was limited because this study involved students experiencing difficulty mastering basic reading skills. This study has involved rehabilitation students as study respondents. Thus, the researcher should have taken all respondents involved in rehabilitation for level 1 for final testing. This study was also conducted during the Malaysian Movement Control Order. Therefore, several procedures had to be followed. As planned, students must stay in the classroom to go to the computer lab. With the school’s consent, the researcher was given an empty classroom next to the Rehabilitation Room to accommodate all rehabilitation students and was allocated from 8:00 to 10:00 am to conduct this study. The researcher could not use the computer lab due to the prescribed procedures. Thus, the school and rehabilitation teachers helped prepare three laptops for this study. The SG-BBL Model was generated once several models and frameworks had been compared. Among those considered, the SG-BBL Model can be a reference and guide for software developers to design serious game applications for effective literacy skills. However, the SG-BBL Model can be further improved to produce software applications that work comprehensively by adding game components to provide positive learning. In addition to improving the activity module, it is recommended that tutorial modules or notes be added as reference for players before starting the selected activity. In addition, the service component can be considered by providing additional learning resources to students, such as administrative and technical support.

For example, this software was designed with the support of several resources, such as the library, the Ministry of Education Malaysia, technical staff assistance, and other services. The prototype development of Mari Membaca was an adaptation of the Richey et al. [100]). Richey et al. [100] have developed several products and models that other researchers can utilise as guides that can be modified as appropriate, such as adding activities. The Mari Membaca prototype can be improved by adding an element of reinforcement to each activity through revision quizzes that can be added to the game app. Prototypes can be improved by storing scores in a database to make it easier for teachers and students to review quiz scores and for the teachers to monitor students’ overall performance. The final evaluation study was conducted within at least six to 10 months. Long-term studies can assess the level of early literacy skills of rehabilitation students at the early, intermediate, and final stages according to the syllabus prepared by the Ministry of Education. Further research should focus on factors other than literacy skills affecting literacy mastery, such as health, social, emotional, physical, and behavioural aspects.