Designing a Mobile Serious Game for Raising Awareness of Diabetic Children

Mobile serious games are very important tools for self-learning, training, and edutainment. They can be used to improve the children’s learning skills to grasp new concepts rapidly in an enjoyable manner. Fun, motivation, and engagement are the most important features in designing attractive serious games. Therefore, designers should always find the best strategies and techniques to keep children playing the game with motivation and enthusiasm. The process is very long and requires different expertise from all the stakeholders. Arcade games are the most played games for fun and amusement in the last decades. Their approach is based on simplicity, easiness, repetitiveness, and soft challenges. The players of arcade games need only to improve their scores in every episode and avoid frustration and hard challenges as in other games. Hence, this arcade-based approach can be used in designing attractive serious games for young children and adolescents. We propose in this research a new serious game for diabetic children to teach them about the disease in a very easy manner. The game uses the arcade approach coupled with useful information about diabetes where the children discover them incrementally during the play. The longer time the child spends on the game, the most she/he can learn and enhance her/his knowledge on the disease. The game success came from its design and implementation with the presence of diabetes children, advisors, and nutritionists. We compare the game with the well-known DEX pet-based game for diabetes awareness on different aspects including user experience, usability, and learning outcomes. The study is conducted on twenty children from a diabetes camp who accepted voluntarily to participate in this research. We evaluate their educational aspects, using an improved version of the MEEGA+ model. We demonstrate the importance of the proposed game in enhancing the children knowledge on diabetes and adopt a new healthy lifestyle.


I. INTRODUCTION
Many entertainment sources have emerged over the last decades. One of the most used entertainment channels is the multimedia video games [66]. Their prevalence is widespread among the young generation due to the appellative side of the content and the messages they disseminate. A survey on two thousand children and teenagers aged between 8 to 18 years old showed that they spend seven hours and 38 minutes in average daily time playing video games [58]. Another study conducted by the Qatar Childhood Cultural Center (2016) showed that schoolchildren in the country spend more than The associate editor coordinating the review of this manuscript and approving it for publication was Mauro Gaggero .
3 hours daily on mobile games. They have also demonstrated that games significantly affect the behaviors of the children. Reference [56], conducted a survey on how the children's behaviors are changing negatively or positively through games. They shown that games can have impacts on violence and aggression attitudes of the players as they can improve their problem-solving skills and help to establish social relationships. The video games are not only designed for fun and entertainment. They are also intended to spread a wide range of messages and ideas throughout this digital medium. Robinett [59] defines the video game as: ''a simulation, a model, a metaphor'' [60]. Other definitions of games are: ''A game is an interactive structure that requires players to struggle toward goals'' and ''Game is the set of positions from among which the player can choose in a given state of the game, and by extension, in mechanics for example, the set of possible positions and thus movements of a system, of an organ, of a mechanism that has furthermore been subjected to certain constraints'' and ''A game is a problem-solving activity, approached with a playful attitude'' [28]. These definitions include many features; one of which is that video games are constructed by principles to give thoughts, and feedback such as serious games. They have dual faces or functions; this means that while the player is getting entertained, he/she receives messages that intend to affect his/her knowledge and beliefs [45]. Educational games can be considered serious for being interactive and efficient tools for learning and training [3]. Several research studies found that children would learn better while playing educational games. For instance, Petri and Wangenheim [53] showed that learning through play is a very successful approach to improve the learning skills of children. They can get benefits from the mistakes they commit while playing and acquiring new knowledge [26]. The most challenging aspects in designing educational games are fun, motivation and engagement. Thousands of educational games are available, however, the player get bored after a while and quit them. Therefore, the game designers need to work closely with the potential users, understand their needs and find ways to keep them motivated and evolved [29].
Nowadays, health awareness games are considered one of the emerging technologies that can be used to educate people effectively [34]. For instance, they can be used in adopting a healthy lifestyle, enhance prevention, self-care to a disease or self-management [12]. In addition, for people who are specialized in the healthcare field, health games offer them the chance to learn clinical skills, disease diagnosis, and appropriate treatment. Such experiences could be acquired by immersing and indulging players in real challenges and interactive environments like augmented or virtual reality [18]. Indulgence may lead to gaining new attitudes, motivation, skills, and other factors in the health field. Video games could also be considered as a tool that raises awareness about diseases like diabetes [38] and [41]. Medicine has started to depend significantly on video games that represent health content. For instance, Thompson et al. [67] have discussed the effects of serious games on diabetic and obese people. They found that video games could intervene, improve, and change bad habits and lifestyles.
Health education can be combined with other game genres to form an edutainment game [32]. Video game genres include action, puzzles, adventure, strategy, casual, and simulation [24]. Each game genre influences the learning experience of different learners as everyone processes and understands information in different ways. For instance, some learners like to learn based on deep thinking, so strategy games are the best choice for them, while other players might learn better through adventure or fighting games [11]. Combining educational games with other genres can also suit the characteristics, skills, interests, cognitive traits or learning styles of learners. Rapeepisarn et al. [57] showed that selecting the appropriate game genre for educational games is a critical factor to reinforce effective learning. The works in [25] and [49] emphasized the importance of designing an easy to use and engaging educational genre game to improve the players learning skills and keep them engaged.
This work consists of merging the arcade approach with educational features in games design to motivate the diabetic children to adopt a new healthy lifestyle and rely on themselves for self-management and learning. The game is very simple, easy to understand and informative. Diabetes children found it very useful and attractive. They felt the game is theirs as they participated in all phases of the game design. We evaluate the game performance through users' experience, satisfaction, and learning experience using an improved version of the MEEGA+ model.
The structure of the paper is organized as follows: in Section 2, we give an overview on video games. In sections 3 and 4, we describe the proposed game and its technical design. In section 5, we discuss the improvement of the MEEGA+ evaluation instrument while in section 6 we describe the experimental design. Results and discussions are presented in section 7 and finally, we conclude the paper in section 8.

II. BACKGROUND AND MOTIVATION
Games have the capability to influence human behavior, especially children and youth [56]. Cognitive research has found that better usability user experience in games can help children improve their literacy and develop essential skills such as reading three-dimensional image visualization and ability to track images in a simultaneous manner. Subrahmanyam et al. [65] showed that children who play educational games with better usability tend to perform well academically and have better mathematics and reading skills than their peers.
Human-computer interaction (HCI) refers to the ways, in which human and computers interact [20] and [68]. The scope of HCI has expanded to include social, organizational, cognitive aspects linked to computer use because of increased infiltration of computers in businesses and homes. HCI can help in the prediction of child behavior and development of social and cognitive skills. Moreover, improving technology in virtual simulation offers an opportunity to create games that will enhance children's organizational skills. Children are also able to develop other skills through video games like problem-solving and reasoning. The more they fail to complete a task in the game, the more they repeat it to succeed.

A. SERIOUS GAMES
Serious games which are games with fun activities and useful objectives like learning and training, have gained the attention of the game industry in the world. They are used successfully in different domains. We can find them games in education, health, simulation, environmental sciences, ecology, rehabilitation, business economics tourism, marketing, and psychiatry. These games help in reasoning, predication and VOLUME 8, 2020 problem solving [17]. The military sector was the first to be highly interested in serious games as they can be used for training and simulation. The United Nations has developed several serious games to raise awareness about sensible world issues, like war consequences in Darfur (e.g., ''Darfur is Dying''), HIVs (e.g., fast car game), and COVID19 pandemic. Muratet et al. [48] have developed a serious game for students to learn concepts of computer programming. Barbosa and Silva [8] proposed a serious game to learn about the human circulatory system. Allegra et al. [2] presented a serious game to manage a touristic company. Other business and management serious games are learn2work and Sharkworld. The Techforce, is a training game in the field of electro and metal industries. Roma Nova is an immersive serious game to teach the history of Rome using 3D and virtual reality. We can find serious games in different scientific field like the Ludwig and PhysikusHD in physics, and the Feon's Quest game in geography. The KickAss and Zirkus Empathica for autistic children. Games like SuperBetter and Blanket Space push the players to concentrate mainly on their own personal betterment. The iThrive Games Foundation (iThrive, 2019) ses games to address problems like empathy, mental health, and emotional skills [26]. The game-based learning platform scoyo covers the complete curriculum of subjects in Germany up to Grade 7 for math, German, English, biology, physics, chemistry, and. Other games like Tetris, Bejewelled, and Pac-Man can be used as therapeutic games to reduce stress and build skills [14] and [37]. The survey of [23] discusses several recent serious games with applications. Çiftci [13] did a survey from 2007 until 2017 on serious games publications and shows that the scientific community is highly interested in the topic with an increased number of publications every year.

B. ARCADE GAMES
In the game industry, a genre can be described as a set of interface standards and game mechanics shared by a collection of titles [15]. Moreover, a game genre is referred to as a specific group of games connected by similar gameplay features. However, genres are not defined by their medium of play or the actual content of the game, rather these are defined the, gameplay, user interface, key mechanics a common challenge, and other features. Besides, genres may comprise a wide range of games, resulting in more specific categorizations called subgenres [39]. For instance, an action game can be categorized into several subgenres including fighting games. The most defined video game genres include action games, sports games, and educational games genre [5]. Educational games are special types of games designed mainly to introduce new concepts or ideas through multimedia, hence classified as an educational based strategy game or educational action arcade game [46].
An arcade game used to be found in public places and operated by inserted coins. Arcade games include electromechanical games, pinball machines, or video games. The golden age of the arcade games was the 1970s through the 1980s.
However, the prevalence of these games started to decline at the end of the 1990s as computer and console games became dominant. The arcade games features are repetitive structures that require hand-eye coordination and fast actions with respect to time [46]. Usually, arcade-style games are organized by levels, best score or missions and do not have the concept of victory challenges (Connor, 2015). Therefore, their usage can make the serious game simple and attractive and can keep them the player engaged for a longer time.

C. VIDEO-GAMES TO RAISE DIABETES AWARENESS
The number of diabetes children exceeds 79 thousand every year [38]. They regularly visit the health centers and hospitals to undergo several assessments including blood glucose test, weight, and eyes health. The follow up cost is very high and technological solutions should be adopted to reduce the strain on hospitals and centers. We should then find solutions to reduce as much as possible the dependency of diabetic people on health professionals and motivate them to control their bodies' health [30]. Note that diabetes is a chronic disease and may generate other symptoms like visual impairment, nerve damage, self-isolation, and psychological issues. The control of the diabetes diseases requires monitoring blood glucose, diet, insulin intake, and lifestyle for a long-term. One approach for raising diabetes awareness is through educational games. These games can influence the behavior of diabetes children, improve their knowledge, and motivate them to eat healthy, and apply physical activities and self-manage their health [9] and [38]. The diabetes children need to learn about the disease and adapt a new lifestyle to save their lives. It is highly important to understand the correlation between, diabetes and food intake with physical activities. Kerfoot et al. [54] show the suitability of diabetes awareness games to both diabetic and non-diabetic players. In their study, both types of players received a drop in their average blood sugar, of 0.4%. Accordingly, the non-diabetic children who played these games exercised healthy feeding habits to avoid obesity that can lead to diabetes. The game players offered also psychological support to their diabetic friends because they understood much better their conditions and played the same games with them. Further, the study showed that diabetic children succeeded to reduce their blood sugar (HbA1c) by 8.3% after one year of playing diabetic games, and hence playing these games has lasting benefits on the blood sugar level. Several diabetes games have been developed in the literature. For instance, Captain Novolin is a diabetic game that develops a superhero who suffers diabetes and needs to control the insulin level [31].
Another diabetic game is the Packy & Marlon (PM). The work in [42] studied the effectiveness of this PM game in raising diabetes awareness amongst children, and the results showed an important reduction of 75% in urgent care visits to health centers. The game motivates diabetes children to follow up a predefined diet schedule, prescribed by nutritionists, to improve their health conditions. The diabetes game DAILY (Daily Automated Intensive Log for Youth) [36] helps diabetic children to monitor their blood sugar through insulin and carbohydrates. The game increased the knowledge of children and incited them to control their health regularly. The Power Defense game [6] uses a simulation approach to teach learners about the diabetes and assists them to control their health through food and insulin.
Several other apps educate children about diabetes and help them understand how to manage their disease on daily basis. For instance, mySugr Junior app features a creative diabetes monster character that encourages children to daily record notes about their blood sugar, meals, carb intake, and insulin doses in a funny manner. The app has attractive interface and parental reporting features. The game, Coco's Cove (Bradley University in Peoria, Illinois), offers entertainment and puzzles with actions to leverage the awareness of children about diabetes control and self-management. The Monster Manor game is designed to help diabetic children of type 1 for self-management and self-care. The game is intended to be used by supervised children so that they can input correct measurement values of their blood glucose and continue the play. Parents have a copy of the app on their iPads to control their children and follow up their actions.
The DEX (Georgia Regents University 2015) is a virtual pet diabetic mobile app that prompts the healthy habits and educates the player about the disease. While playing the game the player gets daily rewards while managing the glucose level for the character through selecting the right food type. The game aims at improving health outcomes and encourage behavioral change. Many companies and research institutions develop games and tools for diabetes people. For instance, the Juvenile Diabetes Research Foundation, in UK has developed many online games for children and adolescents to control their blood glucose and adapt a healthy lifestyle. The Ayogo Health, Inc, in Canada (https://ayogo.com/tag/canada/) proposed several genre apps for children and adults with type 1 and 2 diabetes. Boehringer-Lilly Company (www.boehringeringelheim.com) propose an educational digital game for type 2 diabetes called Complications Combat. A very good survey of the topic is given be [30].

III. THE PROPOSED GAME
According to Schema theory [1] and [62] repetitive learning helps children learn quickly by using all their senses and body features. Repetitive knowledge to children can be an area of strength, where they choose to engage in specific behaviors, tasks, and activities as a means of reflecting on their preferred intelligence. It is vital to watch children play to identify their schematic repetitive actions to plan for their development in the future by this knowledge. Memory exercises and arcadestyle games also exhibit the same recurring patterns. The insistent character of arcade-style games makes them typical conditions for embedding a type of recurrence, which helps in memory retention and recall. Similarly, arcade games have an intensely coherent trial-by-trial structure, which contin-ually encourages players to overcome new challenges [17]. It is therefore imperative for children aiming to learn new skills to keep repeating the same knowledge several times to understand new concepts. After the multiple repetitions of the same practice, students will understand what action to take next, such as which of their responses they find suitable in a specific stimulus.
In various methods of learning, behaviorism plays a critical role in learning because it makes students learn from mistakes and learn by doing [63]. In learning by doing, students play an active role as well as interacting with computers while finding a solution to complex problems. After playing the game for a while and knowing how to solve some tasks efficiently, their problem-solving skills tend to improve significantly. The computer provides various tasks based on the students' level of the game, and the students will respond correctly and effectively. Learning by doing resembles learning from mistakes in some ways; though in the latter, students obtain feedback regarding the wrong steps they might have taken. Consequently, they repeat the tasks until they correctly do them. According to Rapeepisarn et al. [57], learning by doing techniques is related to action arcade game styles. Realizing the best techniques of learning and the style of game is more likely to promote some strategies for learning that enhance the development of new skills, knowledge, and encourage positive behavior. Since those who play video games regularly are more likely to learn to play arcade action video games than their counterparts who employ a strategy of repetition to learn to play. Understanding how children learn to play video games has helped to establish the games' genre children like plating.
Recent surveys on educational games [70] revealed that intrinsic engagement by people leads to more advanced learning because they not only accomplish a task willingly, but also devote more of their efforts to learn and will make use of its future [4] . Apparently, engagement influences learning and motivation skills since researchers established the fact that such video games involve multisensory settings as well as stimulating the ability of the players to think and make meaning of what they do. To date, the major findings stress the significance of both motivation and enjoyment to sustain players' participation.

A. HEURISTICS FOR EVALUATING GAME USABILITY
This section discusses the usability heuristics used in the mobile games, which will later be applied during the design and implementation of our game. Usability in educational games is a significant element given that if the game has issues affecting user entertainment or experience value, then there is a risk of the user abandoning the game and choosing a different activity. Table 1 of game usability heuristics adapted from [33] shows the game controls and the interface for usergame interaction. They allow the player to manipulate the game efficiently without frustration. We use these heuristics in the design of our game through a long cycle of testing and improvement to reach the current stage of the prototype. Usually, the game interface is the first item players come across when they start playing a new game. Indeed, successful game usability will ensure that players enjoy the game and use it. There are several groups of game usability heuristics: Firstly, heuristics GU 1 -GU 5 have a connection to the presentation of the information and visual design. Secondly, GU 6 -GU 8 deals with the navigation arrangement as well as the use of controls when controlling and navigating a game character. The remaining groups of heuristics deal with other essential aspects such as receiving feedback, ways of helping or guiding the player to concentrate on the game. Game usability also covers similar elements to the traditional usability matters of software productivity although there is an adjustment to individually include game-related issues. All these aspects have been used in the implementation of our game after a heavy testing with users of different categories.

B. GAME OVERVIEW
We propose a new and simple health mobile serious game called QDG (i.e., Qatar Diabetes Game) that uses only the best-score approach and avoid hard challenges. The player needs only to improve her/his score at each episode while information is injected during the play. The game help diabetes children to learn about the disease and motivate them to control their bodies by moving, eating healthy food and control their blood sugar. We designed a local character called Abood who is affected by the diabetes disease. The children should monitor Abood's blood glucose to progress in the game. They should learn about diabetes and distinguish between healthy and unhealthy food impacts on Abood's health. Since the traditional arcade games have, almost in totality, disdained the notion of attempting to ''tell a story'' of any type, the proposed game does that in the same manner. The designers of traditional arcade games never felt a sense of requirement for fleshing out their game-worlds, for example, by providing reasons as to why the players focused on shooting a particular target or even eating a given form of a dot; as a result, these games suffered from such learning actions.

C. GAME OBJECTIVE
The objective of this work consists of developing an educational game that should link the gameplay with predetermined learning outcomes. Human-machine interaction research, experts, and patient codesign of the game and its content are essential factors to its success [44]. We use different methodologies including experts in instructional designs to set the learning outcomes and nutritionists to approve the used food items in the game. For instance, one way to set and review the game requirements is by soliciting the expert's advice. Ms. Katie Al-Nahhs, a nutritionist from the Qatar Diabetics Association who points out some important learning concepts needed by children to know more about diabetic people and how they are affected by eating different kinds of food. She also highlights the importance of eating local healthy food and discourages indulgences in unhealthy food publicized largely in the country. In fact, it is highly important to adapt a different lifestyle for diabetes children and avoid as much as possible fast food which are easily accessible in malls and shopping areas. Additionally, the Qatar National Diabetes Strategy 2022 team has underpinned the importance of raising awareness among diabetes children about different food types and their impact on the body and health. These comments have shaped the overall objectives of the gameplay. We give below the list of learning outcomes used in our game: • The impact of junk food on the health conditions. • The control of the blood sugar level. • The critical situation for a diabetic person. • The role and importance of insulin. • The importance of eating regularly and moderately. • The importance of sport and movement to avoid overweight or obesity which may lead to diabetes.

D. GAME PRINCIPAL
In this game, different types of foods and insulin injections will be dropped, and the player must move the main character Abood to the left or the right side to pick the appropriate food. Accordingly, each type of food will affect the sugar level of Abood that is visible to the player and hep her/him to comprehend the correlation between the taken food the blood sugar changes. Our game does not have a winning condition following the Arcade game approach in which the best score concept is utilized. Thus, the game increments the best score every time the player picks the healthy food and decrements the score by 4 points whenever the player picks up unhealthy food items. Learning and mastering how to play traditional arcade games become easier as the learning is mainly done by making trial and error. The player in the QDG game has multiple lives, which provide them with more chances to learn how to play the game. On the other hand, conventional arcade games allow players a fixed number of attempts or ''lives'' before they end. The multiple lives in this proposed arcade educational game give beginner players more chances to explore the mechanics of the game before it ends. Keeping eating healthy food contribute for a better health and the shape of the character Abood will be improved. In addition, the player gains extra life that can be used to extend the time play of the game. This motivates the child to continue with the game, particularly, when progressing from one life to another. Moreover, additional lives act as incentives for players because one death does not necessarily mean the end the game. This notion motivates the players to undertake risks they would have otherwise avoided had the situation been different. In fact, getting diabetes is not the end of the world and one can return to a normal life by changing her/his style of living. The game demonstrates that we can extend or healthy life through simple control on food items.

E. USER-CENTERED DESIGN
Commonly, edutainment games design adopts the Usercentered design (UCD) methodology [44]. Pagulayan et al. [50] mentioned that the essentiality of UCD is set during game development by utilizing several techniques entailing playtesting usability and initial experience. It also involves some surveys and user groups to continuously pretest the game and get the proper feedback. In this work, the UCD is adopted to design and develop the QDG game. To test the effectiveness of our game, we have conducted a qualitative and quantitative research study with children of 8-11 years at a diabetes winter camp in Doha, Qatar after signatures of appropriate parental consent forms. There were diabetic and non-diabetic children. The study involved specialists from Qatar Diabetes Association and other members from Qatar National Strategy organization. The aim was to maximize the engagement between potential users and the game developers. The study was carried out for seven days focusing on group discussions, face-to-face interviews, and explanation to design and test the game with the children. The study included children with diabetes of types 1 and 2. Every child was interviewed individually to assess her/his knowledge on diabetes. The intention was to apply human-computer interaction methods to design an effective diabetes-learning game, and to incorporate the diabetics' ideas and feedback for motivated diabetes education. The interviews were used in designing the prototype, selecting the character Abood, the food items, and setting the appropriate messages. In fact, working closely with children to co-design the game, allowed us to build the best suitable and workable diabetes game for them. This strategy ensures that the children will play the game and informed their peers to use it as well. The second part of the work is to apply MEEGA+ [52] measurement survey tool for children to gather data about the game usability, player experience and perceived learning.

IV. GAME DESIGN
The development of the educational game requires deep understanding of the domain of discourse and the learning outcomes [7] and [61]. The player should enjoy playing with the game, perform simple actions and learn by consequences. Otherwise, she/he may get bored and drop the game. In our game, we follow the same principle, where the child keeps moving the character Abood and the action applied is eating the dropped food and avoid as much as possible the junk items. As a consequence of these actions, the body of Abood get healthy with movement and healthy food items while Abood blood sugar is increased with junk food, change body shape, get obese and may lose lives. The player learns then by consequences and keep having fun while playing. We worked very closely with the children who are the potential players of the game in all phases of the design. Experts in learning have participated in this process, by approving the character of the game and the foods. Their involvement allowed to build a very successful and tailored game for diabetes children and can ensure its adoption and sustainability. Initially, a paperbased prototype consisting of the game requirements, learning objectives, messages to disseminate to the players and assumptions were sketched and discussed with the stakeholders in deep details (i.e., children and specialists). Our goal was to gather initial feedback on the design through face-to-face interactions rather than design a random interface or based on exiting games of the genre. We wanted to demonstrate that the game is designed exclusively of the children so that they can feel it is their game and accept to play with it and learn. The first prototype expresses the main ideas of the game including basic navigations between windows and the content view. Several designs were implemented and refined until the game interface satisfied the children and the experts. For instance, during the initial testing phase, the design shown in Figure 1 (left) was found to have an interaction issue when holding the mobile phone with two hands in a portrait screen portal. Accordingly, this problem was resolved in another design version where no virtual control buttons are used, and so provides the freedom of gripping the device. The design process was iterated taking into consideration all feedback and comments as much as possible. The final version of the game shown in Figure 1 (right) completely tackled all technical issues and limitations of the other earlier designs and delivered the needed information about diabetes which have been approved by nutritionists. The process was very long following a top down approach with lots of patience from the game designers to reach a satisfactory user interface. In fact, the experts changed their minds several times on messages to be sent and food items to be selected. We accepted to change our design even after finalizing the game whenever something seemed inappropriate to experts or the children. In fact, we wanted to gain the sustainability of the game. Therefore, we did not ignore any remark to improve the game. Figure-1 below shows the initial and final design. Note that more than 15 different designs have been proposed and discussed.
For the player to obtain a good score reward and learn the basics of diabetes self-management, she/he should keep blood glucose in some acceptable ranges. Thus, the player has three possible chances to score higher and consequently to discover certain understanding of changes in blood glucose resulting from healthy/unhealthy food consumption and insulin injections. To start the game, the user logs into the game app and gets access to its contents. Once, the user pushes the play button, she/he will begin collecting food items and insulin injection by moving left and right within the gameplay. While playing the game, the user gets points, rewards, and feedback messages from the system accordingly. In addition, the user has several options during the gameplay, such as 1) pauses the game and continue later, 2) progresses to the next second/third life, 3) compares points of the three lives, and 4) pushes ''exit'' button to close the application. Like other games in this domain, the QDG requires players to balance insulin and food intake to keep the HbA1c within the normal range. Our algorithm, in Figure  2, maintains this balance by calculating the player's score based on the collected food items. Inspired by the quality factors Q 13 -Q 18 given in Table 1, the algorithm allows understanding the impact of being in a status of hyperglycemia versus hypoglycemia as well as being in a normal situation while consuming foods or taking insulins. The algorithm runs every time the player starts the game, and the challenge is to get as many rewarding points as possible. Note that during the diabetes camp, the children have been introduced to all the diabetes notions, concepts and jargon used. During our face-to-face interviews with the children we found that they know the importance of healthy food and how they can affect their health. However, we noted that most of the children like to eat fast food from restaurants as they found them very delicious and their peers go regularly to them. In addition, they have money to try all types of fast food with their friends. Advertisements are also highly used by fast-food restaurants to attract children and adolescents to come and degust them.

V. EVALUATION INSTRUMENT
There are many methods used by educationalist and humancomputer interaction (HCI) specialists to evaluate educational games. Based on the literature review, MEEGA+ model (Model for the Evaluation of Educational Games) considered the best tool for gathering data about player in terms of learning outcomes, user experience, and usability. MEEGA+ supports collecting data through a pre-test, post-test, and a questionnaire [52]. Hence, for this work, the MEEGA+ model measurement tool was used as an evaluation instrument of the game and its impact on children's knowledge. The application of the MEEGA+ model in any given context requires the use of questionnaires to facilitate the collection of data on the perceived reactions of learners after they have interacted with the game. As shown in Figure 3, MEEGA+ provides a measurement instrument tool to study children perception in terms of user experience and perceived learning following different criteria. The first part of this section discusses the limitation for the MEEGA+ model and the suggested improvement. Then the following section discusses the process applied for adapting MEEGA+ measurement tool in our evaluation.

A. IMPROVEMENT IN THE MEEGA+ MODEL
This subsection discusses some improvement in the MEEGA+ evaluation model. In MEEGA+, the two main quality factors are the player experience and the perceived learning quality factor. The player experience consists of several dimensions, such as fun, challenge, and usability. The perceived learning consists of short-term learning and learning goals dimensions. Nevertheless, based on the principles of HCI, the user experience should not be fragmented into usability. According to the definition in ISO 9241-11, usability measures the efficiency and effectiveness of users to accomplish specific goals in particular environments.
In contrast, the user experience (UX) focuses on the user's feeling and emotions triggered by the app or the system. In HCI, the UX and usability factors are interlinked, but they are different regarding the impact of user satisfaction. In fact, UX and usability complement the entire experience once interacting with a system (Lewis, 2013). To let the users, continue to be motivated to use the system (or the game), they must be stimulated. We refine the MEEGA+ model by considering the usability as a quality factor that covers a deep involvement and contains the defined dimensions result in the theoretical evaluation model presented in Figure 4. Therefore, the MEEGA+ should then contain three quality factors, namely usability, user experience and perceived learning. The usability should assess four factors in the QDG game, accessibility, learnability, operability, and user error protection.

B. ADAPTING THE MEEGA+ MODEL FOR CHILDREN
From the player-centered design process discussed earlier, specifically during the testing phase, the children were not able to understand some of the questions in the current MEEGA+ model measurement tool. Many questions were quite tricky for eight-year-old children to understand. This made the model problematic to use for young users. The MEEGA+ questionnaire contained sections for selecting participant age ranges, and the lowest was under eighteen years old. This indicate that the questionnaire was not suitable for young children to assess; otherwise, there would be age groups such as 5-10, 11-15 or similar. In addition, and according to our knowledge, none of the previous research works used MEEGA+ as a tool for evaluating children perceived learning. To address these issues of children assessments, two adaptations in the MEEGA+ model were made to better match the children needs. The first step was to convert the 5-point Likert scale to a 3-point Likert format. Wright and Asmundson [69] converted an original 5-point Likert scale response format to make it more easily understandable by young children and adolescents. In addition, Gelman and Baillargeon [21] argue that younger children cannot respond on a 5-point scale because it is beyond their mental capacity and reasoning. So, in the adapted measurement tool for children, the 3-point Likert system was used in which the three answers available were simple agree, disagree and neutral (e.g., either agree or disagree). We explained the meaning of neutral to the children through different examples during the fac-to-face discussions before applying the questionnaire. Once we felt that all the participants understood the questions and how to answer them based on their conviction, we started the assessment by distributing the questionnaires and collecting the data for analysis.
The second change in the adapted measurement tool MEEGA+ was as to reduce the item descriptions because the children were frustrated with answering the first 34 user experience items through the player-centered design process. Accordingly, the questionnaire was minimized to fit the children's understanding skills in collaboration with Qatar Diabetes Association experts. Table 2 presents the adapted part of the MEEGA+ evaluation instrument for the children using the QDG serious game.

VI. EXPERIMENTAL DESIGN
Studying users is a commonly HCI method used to understand users' behavior and needs. In this research, quantitative VOLUME 8, 2020 methods are applied to collect and analyze data. The analysis is related to hypothesis testing and provides a basis for comparing groups. Moreover, qualitative methods are essential in these types of studies to deeply understand the users' interaction and their needs. Indeed, they are heavily used in the field of HCI as they provide views about different situations. In this work, we are investigating the effect of arcade game approach in terms of engagement, player experience, usability, and learning outcomes. The study shows a comparative study between the proposed diabetic awareness mobile game and a well-designed diabetic mobile game called DEX. We conducted qualitative and quantitative studies that involve interviews and surveys. The quantitative instrument is proposed by MEEGA+ tool that mainly focuses on evaluating educational games. The MEEGA+ instrument analyzes educational games to evaluate the perceived quality with respect to the player experience and perceived learning from the player's point of view. The MEEGA+ model considers a method of ascribing quantitative value to qualitative data. Interviews are used to collect qualitative data to observe and understand the player feedback and behavior on the game developed, compared to the DEX mobile game.

A. CASE STUDY SELECTION
To perform the experiment, we need to select the case study. Therefore, a qualitative and quantitative usability test is to be conducted for the QDG game to collect the data needed to study the impact of an arcade approach. On the other hand, another usability test is to be conducted to a diabetic game case study, and eventually, the data results from the case study will be compared with our game. The goal of the comparison is to study the impact of arcade game approach as compared to another non-arcade diabetic mobile game approach used in different diabetes game. According to Miles and Huberman [47], ''qualitative samples tend to be purposive, rather than random''. It is important to select the scenario in the case study very carefully. Selecting the case study wisely will help in obtaining the most relevant, valid, and potentially generalizable evidence. The selection process follows the advice suggested by Robert's two criteria [64]: 1) Select a case study that is more likely to maximize the learning outcomes, 2) Select a case study that is available, and likely to yield evidence. We set the following ten conditions for selecting an appropriate case study: • The game genre should not be an arcade style. • The game is developed based on research aspect. • The game should be a mobile diabetic game. • The game should be a single-player game. • Both games should deliver the same learning outcomes. • The game is very simple to use and easy to learn. • The game is 2D with an animated design. • The game targets the age group of 8 years and above. • The game uses simple vocabulary. • The play time given to finish the game is unlimited. We found that the diabetic game that matches these ten conditions is the DEX game (Augusta University, and Georgia Regents University 2015): a virtual pet diabetic mobile game as shown in Figure 5. The player in DEX game takes care of a diabetic virtual pet called DEX. The game prompts the healthy habits and educates the player about the disease. While playing the game the player gets daily rewards while managing the glucose level for the character through selecting the right food type. The game aims to improve health outcomes and encourage behavioral change.

B. PARTICIPANTS
We have invited twenty schoolchildren of 8-11 years old to test the games during a winter camp organized in Doha, Qatar over a one-week period. Initially, we invited all the children to participate in this work. However, only 20 of them responded positively and accepted to participate. We could not enforce the others to be involved in this research against their will and desire even though their advisors encouraged them to participate with us. They preferred to perform other activities, like riding horses, playing football, running, or playing cards. Therefore, we limited the current evaluation study with 20 children only. All consents forms have been approved and signed. We split the participants children randomly into two groups of consistent distribution and sizes: (Group A: 5 boys and 5 girls), (Group B: 6 boys and 4 girls). For collecting demographic data, the children in each group were asked, ''How often do you play digital games?'' As most children play the digital game every day and for several hours, these children either attended a mobile development event or were nominated by the Qatar Diabetes Association.

C. DATA SOURCE
We have used the MEEGA+ model version 2 to collect the quantitative data. We have introduced the game to the children and explained its usage and main features. We asked every child to test the game on a mobile device and start playing for a short time. We have interviewed the children to assess their understanding of the game and observe their behavior and reaction to compare it with the DEX mobile game. We ensured that all participants play with the games easily.

D. PROCEDURE
For the first experiment, two groups of children were formed: Group A and Group B. Each group tested a different game for seven days. After the elapse of this period, the groups switched the games that they were playing. So, Group A started playing Game 1: QDG; and Group B played Game 2: DEX. In the beginning of the study, a pre-knowledge questionnaire was performed with the participants in each group to ascertain their knowledge about diabetes. Knowing the children's prior knowledge helps in evaluating the impact on perceived learning after playing the two games. Each group was handled separately to perform the training. In the training, a brief introduction to the disease and the game interface was explained. The training contained instructions on how to perform the main actions in each game. Then, the players had the chance to download the game in their mobile device to play with it for seven days before switching to the second game. On the seventh day, after finishing the first part of the experiment, a post-survey using the MEEGA+ measurement tool was performed. Then, in the second part of the experiment, the two groups switched playing the games, so Group A began to play the DEX game, and Group B began to play the QDG game. On the fourteenth day, after finishing the second part of the experiment similarly, a postsurvey using the MEEGA+ version 2 measurement tool was presented to each group.

VII. RESULTS AND DISCUSSION
For comparing QDG with DEX, the Mann-Whitney test [43], a Likert scale quality measurements, is performed on  table 3, the three quality factors addressed are: usability, player experience (UX) and perceived learning (PL). For each quality factor, a set of dimensions is used. The experiment statically studies the difference in children's answers to each question called item description to determine the game with the better quality in usability, UX and PL. The MEEGA+ contains 18 questions in the ''Likert Scale'' described as positive answers (high = 1), negative answers (low = −1) and neutral (neutral = 0). To calculate the mean score and the standard deviation of each question, we give three points to positive, two points to neutral and one point to negative.

MEEGA+ version 2. As shown in
The experiment resulted in data that presents the impact of the arcade approach comparing to a virtual pet-based game. From the first part of the experiment, we can observe that  the UX and perceived learning quality factors for the arcade approach received higher average scores than in Game 2 (e.g., DEX), as shown in table 4. After applying the Mann-Whitney test, a significant difference appeared between Game 1 and Game 2 in UX and perceived learning quality factors. Game 1 received a better gameplay experience and a better perceived learning experience than Game 2. Both games received a similar level of usability in terms of aesthetics, operability, accessibility and error prevention and recovery.
To get more insights on why some dimensions rated negatively, an interview was conducted with the children. From the experiment, the children rated item description (Q 11 ), ''I forgot about my immediate surroundings while playing this game,'' negatively in both games. To understand this result, the interview discovered that the students felt the question showed a weak point in their personal awareness. In addition, in order to understand why Game 1 was less usable than Game 2, it was highlighted from the qualitative data that it is difficult at the beginning to learn how to control the game, and the participants needed to develop their motor skills to learn the gameplay. Similarly, to understand why Game 2 rated negatively in most of the UX dimensions, we found that children wanted more challenging game rules and a more competitive game world, which Game 1 offers.
In the various explorations of the prototype interview and conversations with children and experts, we found that children and experts appreciate those visual effects that represent the culture and the surrounding environment. Most of the children liked the local character Abood and the Corniche environment (e.g., coast road in Doha, Qatar). One of the user's opinion was, ''I like that character call him Rashid same as my name.'' Another young diabetic girl said, ''I love that game, the character is wearing traditional clothing, the same as my brother.'' Similarly, a seven-year-old boy asked, ''Is the main character located at Corniche? I know those towers in Qatar''.
As for using arcade in an educational game, Mrs. Lama from the Qatar National Diabetic Strategy team said, ''I am sure that my child will play the game; the gameplay is similar to Subway surfers, but this is going to teach my child how to deal with diabetic children in his school.'' In addition, Dr. Maryam Al-Hidos, a general doctor from the Ministry of Public Health in Qatar, highlighted that this type of game is needed because children prefer fast-paced games. In addition, she emphasized the need for a diabetes-awareness mobile game that reflects the culture and tradition of the country. Dr. Al-Hidos said, ''There is a need for a short, attractive, simple game derived from the culture that can deliver the knowledge so that we can use it in the campaigns.'' The game needs to have features to select different characters and food items where the player can select them and play. For instance, the child should select the desired characters either male or female, the type of clothes to wear and the preferred food items for the meals. Sounds should also be associated with the selecting items describing them briefly, like for instance, you have selected an apple; you missed an orange which is a healthy food; only few food items remain to gain a new life. We can also generate a report on time spent, scores obtained, number of healthy and unhealthy food items selected, for further analysis and improvement. The game has only one episode which can lead to a boring aspect of the player. More scenarios should be developed and added to the game with some soft challenges. Another limitation of the game is the lack of exercising instructions which are highly important for diabetic children. Several assessments should be done on a larger number of students and for a longer period to confirm the usefulness and effectiveness of the game.

VIII. CONCLUSION
We have proposed an easy-to-use mobile game using the arcade approach to raise the awareness of diabetic children about their disease in a very easy manner. It motivates them to change their eating behavior and adapt a healthy lifestyle. We have shown the effectiveness of the game in keeping the children playing with enthusiasm and motivation as the participated on all phases of the game design and felt that its designed for them. The simplicity of the game coupled with its learning tips are the main characteristics of its success and long-term adoption and sustainability. The game followed many design steps including guidelines from the literatures, advices from nutritionists, learning experts, and feedback from children. We have tested the game with children during a winter camp in Doha, Qatar and the obtained results were promising towards improving the learning concepts of diabetes and its consequences on health. We have explored different evaluation techniques for games-based learning and highlighted their main limitations. Specifically, our work used the enhanced MEEGA+ model to evaluate games from different perspectives including user experience, usability, and the learning outcomes. The limitation in the MEEGA+ model was also addressed to take into consideration the children's skills and capabilities. We adopted the player-centered design to test different scenarios with the arcade game-based learning approach. We have compared our game with the DEX diabetic game in terms of usability, player experience and perceived learning. The obtained results showed a strong correlation between the user experience and the arcade approach. The players (Group A and Group B) enjoyed learning while playing and felt motivated and excited without frustration. These findings outline the importance of the arcade approach to design attractive educational games that match children's needs as well as motivate them to play for a longer time and keep them engaged. We plan to test the game with a larger number of children with diabetes in the next cycle of the camp conducted yearly in December. In fact, long term testing is needed to evaluate the performance of the game in educating the children. Two groups should be identified and use the app for several weeks or months and get assessed accordingly. We do hope that the world can get rid of COVID19 pandemic and return to normal so we can improve the game through deep testing and experiments. We intend to add more features to the game, like adding personalized food items, recording the daily time spent for further analysis, and team-playing where two or more children can play the same episode. They can then compete in selecting the healthy food items and get better scores. We can also add soft exercises and puzzles with specific learning outcomes, like collecting healthy food to solve a puzzle or getting out from a maze to continue the play. These mini games added to the main game allows to add more learning objectives and can enhance the knowledge of the diabetes children. We can add rewarding items, like gaining virtual currency that can be used to ''buy'' in-game life. Some physical activities can also be added to the game to burn calories and improve the player mental health. We plan to make the game available in Google play or Apple store for further testing, feedback, and improvement.
NOOR AL-MAADEED (Member, IEEE) received the Ph.D. degree in computer engineering from Brunel University, U.K., in 2014. She is currently an Associate Professor with the Computer Science and Engineering Department, Qatar University. She participated in many regional and international conferences and published an important number of research articles in prestigious peerreviewed journals, book chapters, and conferences proceedings. She has improved the relationship between academia and the industry by leading many research projects, both domestically and abroad, totaling over eight million QAR in her fields of specialization: image processing, speech, speaker recognition, intelligent pattern recognition, video-surveillance systems, and biometrics. She is a member of the first batch Qatar Leadership Center, the Current and Future Leaders Program, and the Qatar University Senate, as well as other committees. She is also a member of various international associations, such as IET, IEEE, BA, and IAENG. She participates in activities which connect her to the community, such as working with charities and volunteering in sport events. He is currently an Assistant Professor with the Department of Computer Science and Engineering, Qatar University. He is also a Cisco Certified Internetwork Expert (CCIE) and a Cisco Certified Instructor Academy (CCAI) serving for the Qatar University Cisco Academy. His research interests include text-to-image synthesis, assistive technology, eductaional multimedia systems, natural language processing, network scheduling, and distributed and real-time systems. He has published over 50 refereed articles in reputable international journals and conferences. He has served as a Technical Program Committee for Elsevier Computer Communications, EURASIP Wireless Communications and Networking, and IEEE ACCESS,  IEEE LCN, IEEE ICT, IEEE AICCSA, IEEE APCC, IEEE GCC, IEEE  ICC, IEEE  SOMAYA ALI AL-MAADEED (Senior Member, IEEE) received the Ph.D. degree in computer science from Nottingham, U.K., in 2004. She has excellent collaboration with national and international institutions and industry on different research project. She was a Visiting Academic with Northumbria University, U.K. She is currently a Full Professor with the Computer Science and Engineering Department, Qatar University. She is also the Coordinator of the Computer Vision Research Group. She has published extensively in computer vision, information engineering, and pattern recognition. She organized several workshops and competitions related to biometrics and computer vision. She was selected as a participant in Current and Future Executive Leaders Program at Qatar Leadership Center from 2012 to 2013. She leads several NPRP research projects on special needs assistive technology, games development, and learning platforms for children.
JIHAD MOHAMED ALJAAM received the B.Sc., M.S., and Ph.D. degrees from Southern University (The National Council for Scientific Research, CNRS), France, in 1989France, in , 1990, and 1994, respectively. He worked on the connection machine CM5 with 65000 microprocessors in the USA to solve hard problems. He was with IBM-Paris as a Project Manager and with RTS-France as an IT Consultant and with Qatar University as a full professor for several years. He is currently working in a research project for children with learning difficulties. He organized many workshops and conferences in France, USA, and the GCC countries. His current research interests include multimedia, assistive technology, learning systems, human-computer interaction, stochastic algorithms, artificial intelligence, information retrieval, and natural language processing. He is also a member