Game Development for Computer Science Education Chris Johnson Monica McGill Durell Bouchard University of Wisconsin, Eau Bradley University Roanoke College Claire [email protected] [email protected] [email protected] Michael K. Bradshaw Víctor A. Bucheli Laurence D. Merkle Centre College Universidad del Valle Air Force Institute of michael.bradshaw@ victor.bucheli@ Technology centre.edu correounivalle.edu.co laurence.merkle@afit.edu Michael James Scott Z Sweedyk J. Ángel Falmouth University Harvey Mudd College Velázquez-Iturbide [email protected] [email protected] Universidad Rey Juan Carlos [email protected] Zhiping Xiao Ming Zhang University of California at Peking University Berkeley [email protected] [email protected] ABSTRACT cation, including where and how they fit into CS education. Games can be a valuable tool for enriching computer science To guide our discussions and analysis, we began with the education, since they can facilitate a number of conditions following question: in what ways can games be a valuable that promote learning: student motivation, active learning, tool for enriching computer science education? adaptivity, collaboration, and simulation. Additionally, they In our work performed prior to our first face-to-face meet- provide the instructor the ability to collect learning metrics ing, we reviewed over 120 games designed to teach comput- with relative ease. As part of 21st Annual Conference on ing concepts (which is available for separate download [5]) Innovation and Technology in Computer Science Education and reviewed several dozen papers related to game-based (ITiCSE 2016), the Game Development for Computer Sci- learning (GBL) for computing. Hainey [57] found that there ence Education working group convened to examine the cur- is \a dearth of empirical evidence in the fields of computer rent role games play in computer science (CS) education, in- science, software engineering and information systems to cluding where and how they fit into CS education. Based on support the use of GBL." This is not unique to CS, how- reviews of literature, academic research, professional prac- ever. A review by Papastergiou [95] found limited evidence tice, and a comprehensive list of games for computing educa- to support games for learning, and a systematic review by tion, we present this working group report. This report pro- Graafland et al. [54] also found no empirically validated vides a summary of existing digital games designed to enrich games to support education in the medical field. computing education, an index of where these games may fit Though our reviews were not designed to be comprehen- into a teaching paradigm using the ACM/IEEE Computer sive, our findings support these claims. This lack of evidence Science Curricula 2013 [13], and a guide to developing digi- prevented us from identifying which of these games are most tal games designed to teach knowledge, skills, and attitudes effective in meeting educational outcomes, because little ev- related to computer science. idence exists to make such claims, and it further prevented us from stating which game design frameworks for CS edu- cation might be most effective across various demographics. 1. INTRODUCTION This required us to rethink our approach and to consider As part of ITiCSE 2016, the Game Development for Com- how we would analyze the games and the relevant research puter Science Education working group convened to examine in a way that would provide significant value for the broader the current role games play in computer science (CS) edu- computer science educational research community. The purpose of this working group report, therefore, is: 1. to provide a summary of existing digital games de- signed to enrich computing education and an index of c 2016 Copyright held by the owner/author(s). Publication rights licensed to ACM. This is the author’s version of the work. It is posted here for your personal use. Not for where these games may fit into a teaching paradigm us- redistribution. The definitive Version of Record was published in Proceedings of the ing the ACM/IEEE Computer Science Curricula 2013 21st Annual ACM Conference on Innovation and Technology in Computer Science (CS2013), and Education, July 11-13, 2016, Arequipa, Peru. ITiCSE ’16 July 11–13, 2016, Arequipa, Peru 2. to provide a guide to developing digital games designed c 2017 ACM. ISBN 123-4567-24-567/08/06. $15.00 to teach knowledge, skills, and/or attitudes related to DOI: https://doi.org/10.1145/3024906.3024908 computer science. To narrow the broad scope of games, we have chosen to Though other areas of study may also contain elements of focus on digital games; however, we note that analog games this process, programming offers a unique parallel. Malone's (board games, card games, etc.) can also be an important insight into how self-esteem is intertwined into the process tool in enriching student learning. Games like those pre- is worth noting, as CS educational research supports the no- sented in CS Unplugged [22] provide a meaningful way of tion that learner and instructor self-esteem and self-efficacy implementing active learning within a curriculum [23]. are important to the learning process. We mention this here, Though an initial goal for this report was to also include as it is important to note that although games for CS edu- a guide for evaluating the effectiveness of games for CS edu- cation may be designed to teach disciplinary concepts, such cation, given the expansiveness of such a task, we will focus as programming constructs or computational thinking, they on evaluation and assessment in follow-up work. may have positive or negative unintended outcomes that Game developers, educational theorists, and others may could affect behaviors and beliefs about computing. find value within this report. However, the primary tar- This section is designed to provide a contextual back- get audience is academic researchers interested in develop- ground to readers and describe the important elements of ing games for enriching CS education. This perspective is research relevant to the remainder of this report. We define reflected throughout this report. the value of using games to teach computer science educa- This working group is unconventional in that its work will tion. We also provide a summary of vocabulary for game span two ITiCSE conferences. This first report is divided design, as well as CS educational research and the broader into several sections, starting with a background that intro- educational psychology that is relevant to game design. duces important vocabulary related to game design theory, provides a summary of previous related computer science 2.1 The Case for Educational Games education research, and provides a summary of relevant ed- Games can facilitate learning across a variety of disci- ucational psychology. It also provides a brief review of games plines in multiple ways. Even games designed specifically for created for computer science education and how these map entertainment have been shown to have educational value. to CS2013. Though much has been written previously over the last cou- This is followed with case studies of four games designed ple of decades about games in education, we provide a brief to teach CS concepts. The games are analyzed in two ways: synopsis for the case of using educational games for teaching 1) for their design elements using the Mechanics, Dynam- computing. ics, and Aesthetics (MDA) framework and 2) for how one Educational outcomes and competencies of modern ed- might evaluate their effectiveness. Using information from ucation are changing, and the quickly changing nature of this analysis, our background research, and our previous ex- computing makes it important for educators to keep pace. perience creating games, we provide a set of best practices Learners are growing up with laptops, tablets and cellphones. for creating meaningful games for CS education. Today, people continuously learn and interact daily with in- formation and communications technologies [27]. The mod- 2. BACKGROUND ern workforce needs relevant education focused more on solv- There is a unique relationship between digital games and ing problems individually and in groups. Jobs are changing computer science, since computer science is the foundation and are often characterized by increased technology use, ex- of such games. But more than that, the programming pro- tensive problem solving, networking and complex communi- cess itself contains many of the same elements found in cation [79]. games. In 1980, Thomas W. Malone [84] stated: A recent ESA Essential Facts Report [45] finds that 65% of US households own a device used to play video games. \In some senses, computer programming itself is Games are culturally relevant to today's learners, and pre- one of the best computer games of all. In the vious research shows that learners may feel more engaged `computer programming game,' there are obvi- when culturally relevant tools are be harnessed for educa- ous goals and it is easy to generate more. The tion [72]. Previous research demonstrates that digital games 'player' gets frequent performance feedback (that sustain engagement and motivation across time [52, 104], is, in fact, often tantalizingly misleading about and that this engagement is strongly associated with student the nearness of the goal). The game can be achievement [119]. In addition, students are more intrinsi- played at many different difficulty levels, and there cally motivated [53] and their work can focus on complex are many levels of goals available, both in terms thinking and problem solving through games [19]. of the finished product (whether it works, how Traditional instruction can be improved by games given fast it works, how much space it requires, etc.) that they can foster collaboration, decision-making, problem- and in terms of the process of reaching it (how solving, communication, innovation, production, and proce- long it takes to program, etc.).