Digital Game-Based Learning ( DGBL )

After years of research and proselytizing, the proponents of digital game-based learning (DGBL) have been caught unaware. Like the person who is still yelling after the sudden cessation of loud music at a party, DGBL proponents have been shouting to be heard above the prejudice against games. But now, unexpectedly, we have everyone's attention. The combined weight of three factors has resulted in widespread public interest in games as learning tools.

The first factor is the ongoing research conducted by DGBL proponents. In each decade since the advent of digital games, researchers have published dozens of essays, articles, and mainstream books on the power of DGBL—including, most recently, Marc Prensky's Digital Game-Based Learning (2001), James Paul Gee's What Video Games Have to Teach Us about Learning and Literacy (2003), Clark Aldrich's Simulations and the Future of Learning: An Innovative (and Perhaps Revolutionary) Approach to e-Learning (2004), Steven Johnson's Everything Bad Is Good for You: How Today's Popular Culture Is Actually Making Us Smarter (2005), Prensky's new book “Don’t Bother Me, Mom, I'm Learning!”: How Computer and Video Games Are Preparing Your Kids for 21st Century Success and How You Can Help! (2006), and the soon-to-be-published Games and Simulations in Online Learning: Research and Development Frameworks, edited by David Gibson, Clark Aldrich, and Marc Prensky. The second factor involves today’s “Net Generation,” or “digital natives,” who have become disengaged with traditional instruction. They require multiple streams of information, prefer inductive reasoning, want frequent and quick interactions with content, and have exceptional visual literacy skills1—characteristics that are all matched well with DGBL. The third factor is the increased popularity of games. Digital gaming is a $10 billion per year industry, and in 2004, nearly as many digital games were sold as there are people in the United States (248 million games vs. 293.6 million residents).

One could argue, then, that we have largely overcome the stigma that games are “play” and thus the opposite of “work.” A majority of people believe that games are engaging, that they can be effective, and that they have a place in learning. So, now that we have everyone's attention, what are we DGBL proponents going to say? I believe that we need to change our message. If we continue to preach only that games can be effective, we run the risk of creating the impression that all games are good for all learners and for all learning outcomes, which is categorically not the case. What is needed now is (1) research explaining why DGBL is engaging and effective, and (2) practical guidance for how (when, with whom, and under what conditions) games can be integrated into the learning process to maximize their learning potential. We are ill-prepared to provide the needed guidance because so much of the past DGBL research, though good, has focused on efficacy (the message that games can be effective) rather than on explanation (why and how they are effective) and prescription (how to actually implement DGBL).

This is not to say that we have ignored this issue entirely. Many serious game proponents have been conducting research on how games can best be used for learning, resulting in a small but growing body of literature on DGBL as it embodies well-established learning principles, theories, and models. On the other hand, many DGBL proponents have been vocal about the dangers of “academizing” (“sucking the fun out of,” as Prensky would say) games. This is partly the result of our experiences with the edutainment software of the last decade or so, which instead of harnessing the power of games for learning, resulted in what Professor Seymour Papert calls “Shavian reversals”: offspring that inherit the worst characteristics of both parents (in this case, boring games and drill-and-kill learning).Many argue that this happened because educational games were designed by academicians who had little or no understanding of the art, science, and culture of game design. The products were thus (sometimes!) educationally sound as learning tools but dismally stunted as games. Yet if we use this history and these fears to argue, as some have, that games must be designed by game designers without access to the rich history of theory and practice with games in learning environments, we are also doomed to fail. We will create games that may be fun to play but are hit-or-miss when it comes to educational goals and outcomes. The answer is not to privilege one arena over the other but to find the synergy between pedagogy and engagement in DGBL.

In this article, I will outline why DGBL is effective and engaging, how an institution can leverage those principles to implement DGBL, how faculty can integrate commercial off-the-shelf (COTS) DGBL in the classroom, what DGBL means for institutional IT support, and the lessons we can learn from past attempts at technological innovations in learning.

If we are to think practically and critically about DGBL, we need to separate the hype from the reality. Many who first hear about the effectiveness of games are understandably skeptical. How much of the research is the result of rigorous, controlled experimental design, and how much is wishful thinking and propaganda? A comprehensive analysis of the field is not possible here and, in any case, has already been done by others. Several reviews of the literature on gaming over the last forty years, including some studies that use rigorous statistical procedures to analyze findings from multiple studies (meta-analyses), have consistently found that games promote learning and/or reduce instructional time across multiple disciplines and ages. Although many of these reviews included non-digital games (pre-1980), there is little reason to expect that the medium itself will change these results. A cursory review of the experimental research in the last five years shows well-documented positive effects of DGBL across multiple disciplines and learners.

What accounts for the generally positive effects found in all these studies about games and learning? These empirical studies are only part of the picture. Games are effective not because of what they are, but because of what they embody and what learners are doing as they play a game. Skepticism about games in learning has prompted many DGBL proponents to pursue empirical studies of how games can influence learning and skills. But because of the difficulty of measuring complex variables or constructs and the need to narrowly define variables and tightly control conditions, such research most often leads to studies that make correspondingly narrow claims about tightly controlled aspects of games (e.g., hand-eye coordination, visual processing, the learning of facts and simple concepts).

As Johnson says in Everything Bad Is Good for You: “When I read these ostensibly positive accounts of video games, they strike me as the equivalent of writing a story abut the merits of the great novels and focusing on how reading them can improve your spelling.” Although it’s true that games have been empirically shown to teach lower-level intellectual skills and to improve physical skills, they do much more than this. Games embody well-established principles and models of learning. For instance, games are effective partly because the learning takes place within a meaningful (to the game) context. What you must learn is directly related to the environment in which you learn and demonstrate it; thus, the learning is not only relevant but applied and practiced within that context. Learning that occurs in meaningful and relevant contexts, then, is more effective than learning that occurs outside of those contexts, as is the case with most formal instruction. Researchers refer to this principle as situated cognition and have demonstrated the effectiveness of this principle in many studies over the last fifteen years. Researchers have also pointed out that play is a primary socialization and learning mechanism common to all human cultures and many animal species. Lions do not learn to hunt through direct instruction but through modeling and play. Games, clearly, make use of the principle of play as an instructional strategy.

There are other theories that can account for the cognitive benefits of games. Jean Piaget's theories about children and learning include the concepts of assimilation and accommodation. With assimilation, we attempt to fit new information into existing slots or categories. An example of an adult assimilating information might be that when a man turns the key in the ignition of his car and the engine does not turn over, and in the past this has been due to a dead battery, he is now likely to identify the problem as a dead battery. Accommodation involves the process whereby we must modify our existing model of the world to accommodate new information that does not fit into an existing slot or category. This process is the result of holding two contradictory beliefs. In the previous example, should the man replace the battery and experience the same problem, he finds that the engine not starting both means and does not mean a dead battery. Accordingly, our stranded motorist must adjust his mental model to include other problems like alternators and voltage regulators (although perhaps only after an expensive trip to his auto mechanic). This process is often referred to as cognitive disequilibrium. Piaget believed that intellectual maturation over the lifespan of the individual depends on the cycle of assimilation and accommodation and that cognitive disequilibrium is the key to this process.

Games embody this process of cognitive disequilibrium and resolution. The extent to which these games foil expectations (create cognitive disequilibrium) without exceeding the capacity of the player to succeed largely determines whether they are engaging. Interacting with a game requires a constant cycle of hypothesis formulation, testing, and revision. This process happens rapidly and frequently while the game is played, with immediate feedback. Games that are too easily solved will not be engaging, so good games constantly require input from the learner and provide feedback. Games thrive as teaching tools when they create a continuous cycle of cognitive disequilibrium and accommodation while also allowing the player to be successful. There are numerous other areas of research that account for how and why games are effective learning tools, including anchored instruction, feedback, behaviorism, constructivism, narrative psychology, and a host of other cognitive psychology and educational theories and principles. Each of these areas can help us, in turn, make the best use of DGBL.

The positive effects of DGBL seen in experimental studies can be traced, at least partially, to well-established principles of learning as described earlier (e.g., situated cognition, play theory, assimilation and accommodation) and elsewhere by others.This means that DGBL can be implemented most effectively, at least in theory, by attending to these underlying principles. How, then, can we use this knowledge to guide our implementation of DGBL in higher education?

A review of the DGBL literature shows that, in general, educators have adopted three approaches for integrating games into the learning process: have students build games from scratch; have educators and/or developers build educational games from scratch to teach students; and integrate commercial off-the-shelf (COTS) games into the classroom. In the first approach, students take on the role of game designers; in building the game, they learn the content. Traditionally, this has meant that students develop problem-solving skills while they learn programming languages. Professional game development takes one to two years and involves teams of programmers and artists. Even though this student-designed approach to DGBL need not result in commercial-quality games, it is nonetheless a time-intensive process and has traditionally been limited to computer science as a domain. It is certainly possible for modern game design to cross multiple disciplines (art, English, mathematics, psychology), but not all teachers have the skill sets needed for game design, not all teach in areas that allow for good content, not all can devote the time needed to implement this type of DGBL, and many teach within the traditional institutional structure, which does not easily allow for interdisciplinarity. For these reasons, this approach is unlikely to be used widely.

In the second case, we can design games to seamlessly integrate learning and game play. Touted by many as the “Holy Grail” approach to DGBL because of its ability to potentially address educational and entertainment equally, and to do so with virtually any domain, this professionally designed DGBL process is more resource-intensive than the first option. This is because the games must be comparable in quality and functionality to commercial off-the-shelf (COTS) games, which after all are very effective in teaching the content, skills, and problem-solving needed to win the game. The development of such "serious games" is on the rise, and the quality of the initial offerings is promising (e.g., Environmental Detectives, developed by the Education Arcade; Hazmat: Hotzone, under development at the Entertainment Technology Center at Carnegie Mellon University; Virtual U, originally conceived and developed by Professor William F. Massy; and River City, developed by Professor Chris Dede, the Harvard Graduate School of Education, and George Mason University). However, the road to the development of serious games is also littered with Shavian reversals (poor examples of edutainment in which neither the learning nor the game is effective or engaging). Consequently, fewer companies are willing to spend the time and money needed to develop these games, for fear of revisiting their unprofitable past, and so the number of games that can be developed is limited. Although this professionally designed DGBL approach is clearly the future of DGBL, we are not likely to see widespread development of these games until we demonstrate that DGBL is more than just a fad and until we can point to persuasive examples that show games are being used effectively in education and that educators and parents view them as they now view textbooks and other instructional media.

The third approach—integrating commercial off-the-shelf digital game-based learning (COTS DGBL)—involves taking existing games, not necessarily developed as learning games, and using them in the classroom. In this approach, the games support, deliver, and/or assess learning. This approach is currently the most cost-effective of the three in terms of money and time and can be used with any domain and any learner. Quality is also maximized by leaving the design of game play up to game designers and the design of learning up to teachers. I believe that this approach to DGBL is the most promising in the short term because of its practicality and efficacy and in the long term because of its potential to generate the evidence and support we need to entice game companies to begin developing serious games.

This approach is gaining acceptance because of its practicality, and research shows that it can be effective. Entertainment Arts (EA), a game-development company, and the National Endowment for Science, Technology, and the Arts (NESTA) in the United Kingdom have entered into a joint partnership to study the use of COTS games in European schools, and similar initiatives are being proposed in the United States. If the United States is like the United Kingdom, where 60 percent of teachers support the use of games in the classroom, the United States may be well-positioned to begin generating the evidence (through the use of COTS games) that the game industry needs to begin developing serious games.

Integrating COTS games is not without its drawbacks. Commercial games are not designed to teach, so topics will be limited and content may be inaccurate or incomplete. This is the biggest obstacle to implementing COTS DGBL: it requires careful analysis and a matching of the content, strengths, and weaknesses of the game to the content to be studied.

There are ways to minimize these drawbacks, some of which I will discuss later, but the elephant in the room is that in our conversations about DGBL, we rarely acknowledge that the taxonomy of games is as complex as our learning taxonomies. Not all games will be equally effective at all levels of learning. For instance, card games are going to be best for promoting the ability to match concepts, manipulate numbers, and recognize patterns. Jeopardy-style games, a staple of games in the classroom, are likely to be best for promoting the learning of verbal information (facts, labels, and propositions) and concrete concepts. Arcade-style games (or as Prensky and others refer to them, “twitch” games) are likely to be best at promoting speed of response, automaticity, and visual processing. Adventure games, which are narrative-driven open-ended learning environments, are likely to be best for promoting hypothesis testing and problem solving. Many games also blur these taxonomic lines, blending strategy with action and role playing, for instance.

It is critical, therefore, that we understand not just how games work but how different types of games work and how game taxonomies align with learning taxonomies. This is not a new idea. In perhaps one of the most ambitious and rigorous examinations of the use of games to teach mathematics, a 1985 study undertaken for the National Council of Teachers of Mathematics developed eleven games for different grade levels using 1,637 participants. The study authors intended their eleven separate game studies to examine if and how games could be used to teach mathematics at varying learning levels.13 Games, they hypothesized, might be better at promoting learning at some levels than at others. Further, they distinguished between three types of game use: pre-, co-, and post-instructional, based on when games were used in relation to the existing curriculum. The study authors found that there were indeed differences by learning level and by whether games were used prior to, during, or after other instruction and also that there were interactions between these two factors. They concluded that although drill-and-practice-type games at the time made up the vast majority of edutainment titles, instructional games could be effective for higher learning levels if designed and implemented well. Though this seems to support the development of serious games, the core principle—that games can promote learning at higher taxonomic levels—is as applicable to COTS games, which require and promote problem-solving and situated cognition before they are integrated with instructional activities or content.

It is important to understand how the theoretical issues outlined here relate to the use of games to teach. Although this section gives a practical description of the issues, it is meant more as a heuristic for understanding the issues involved than as a prescriptive tool. There are a wide range of other factors that must be considered, such as using the game outside of the classroom (as with all homework), balancing game play and other instructional activities, and rotating students’ use of the computers in classrooms where there is not a one-to-one student-computer ratio. Many of these issues are not unique to DGBL, however, and are adequately treated by authors of texts that emphasize integrating computer technology into the learning process.

Once we have chosen a game, and have analyzed it for content, we have to decide what to do about missing and inaccurate content. What content will have to be created to address gaps? Who will provide this content? Some believe that this is the teacher's responsibility, but current thinking in education suggests that the more students are responsible for in their learning, the more they will learn. Certainly, there is some content that will not be practical for students to address on their own, but wherever and whenever we can maximize student responsibility, we should.

The way we choose to maximize student responsibility is important. Because we are going to have to go out of the game environment and into the classroom, we run the risk of eliminating what is fun and engaging about the game. So, rather than simply providing additional reading or handouts with the missing or accurate information, we should strive to design activities that are logical extensions of the game world. Learning is integral to the story of the game world—players are never asked to step out of the game world to do something (although they frequently do so when stuck). The constant cycle of cognitive disequilibrium and resolution—the engagement—is what leads to the experience that Professor Mihaly Csikszentmihalyi refers to as flow.15 Flow occurs when we are engaged in an activity (physical, mental, or both) at a level of immersion that causes us to lose track of time and the outside world, when we are performing at an optimal level. Good games promote flow, and anything that causes us to "leave" the game world (e.g., errors, puzzles that require irrational solutions) interrupts flow. If we were to simply design “traditional” classroom activities (workbooks, textbook reading, teacher handouts, etc.) that addressed the missing, misleading, or inaccurate content in the game, we would be interrupting the flow experience. Granted, anytime we ask the players to stop the game and do something else, flow will be interrupted. But to the extent that we can keep these additional activities “situated” within the game world (i.e., connected to the problem being solved, the characters solving it, and the tools and methods those characters use or might use), we will minimize this interruption of flow. For the same reasons, we should make sure that students spend enough time in the game to promote flow and, correspondingly, significant time in the extended instructional activities. Even if these extended activities do not promote flow, the more frequently students move from the game to other activities (even those related to the game), the more frequently flow will be interrupted in each activity.

Although it is not possible to stay entirely within the game world (and therefore to keep students in flow) when implementing COTS DGBL, there is another reason we should strive to keep the activities we design situated within that game world. Malone and Lepper identify fantasy (endogenous and exogenous) as one of four main areas that make games intrinsically motivating. 16 Endogenous fantasy elements are those fantasy parts that are seamlessly integrated with the game world and story; exogenous fantasy elements are those that, though in the game, do not appear to have much relation to the story or game world. Endogenous fantasy elements not only help make games intrinsically motivating; in theory, they should also promote flow. So whenever we ask students to not be in the game, we should strive to keep the activities and roles they take on (the fantasy) endogenous to the game.

Thus, the roles we ask them to take on should be extensions of the roles they play in the game. These can be main characters, ancillary characters, or characters that could hypothetically be part of the game. The activities we ask them to perform as these characters should be authentic to the goals of the game world and the professions or characteristics of these characters. Some examples of endogenous activities might be to develop budgets, spreadsheets, reports/charts, and databases; to write diaries, scientific reports, letters, legal briefs, dictionaries, faxes; to design, duplicate, and conduct experiments; to conduct and write up feasibility studies; and to assess the veracity of game information or provide missing data. We should not be so naïve as to think that students will find these activities to be as engaging as the games, but given our need to meet curricular goals and our desire to tightly integrate the games with the learning process, this seems a good way to meet in the middle.

Of the several technology "learning revolutions" during the last quarter-century, most have failed to achieve even half of their promise. Although there are many reasons for this, the primary fault lies with our inability (or unwillingness) to distinguish between the medium and the message. Two examples of such technological learning innovations from our recent past are media technology and computing technology.

In the 1960s and 1970s, audio and video (and later, television) were touted as technologies that would revolutionize learning. We rapidly began implementing media wherever possible, regardless of grade, domain, or learners. Many studies were conducted during the 1970s to compare media-based classrooms to "traditional" classrooms, and some of the more sensational ones found their way into the public eye. By the 1980s, enough studies had been conducted to allow for meta-analyses and reviews of the literature. Most of these resulted in what has famously been called the "no significant difference" phenomenon—meaning that, overall, media made no significant difference to learning. This was not surprising to instructional designers, who argued that the implementation of media was not consistently of high quality and that the quality of the instruction in "media" versus "traditional" classrooms was not controlled. The key to understanding this issue lies in the difference between use and integration of media. Using media requires only that the media be present during instruction. Integrating media, on the other hand, requires a careful analysis of the strengths and weaknesses of the media, as well as its alignment with instructional strategies, methods, and learning outcomes. Weaknesses are then addressed through modification of the media or inclusion of additional media and/or instruction, and instruction is modified to take advantage of the strengths of the media. In cases where there is poor alignment, the media is not used.

Sadly, the history of the use of computing technology in learning parallels that of media use. The personal computer arrived in the 1970s, and predictions of revolutionized learning quickly followed. Schools spent hundreds of thousands of dollars on computers in the early 1980s, vowing to place one in every classroom. Studies comparing classrooms with computing technology and those without proceeded at the same pace as had studies comparing media-rich and media-poor classrooms. Once again, instructional designers and others pointed out that the quality of implementation varied greatly, making comparisons impossible. By the time there were enough studies to evaluate and review, the quality and diversity of the different implementations made it difficult to draw any meaningful conclusions. Once again, it seemed there was "no significant difference" between classrooms that used technology and those that did not. Once again, we had mistaken technology use for technology integration.

Eventually, though, educators learned from this and from prior experience with media. They began developing and testing better-integrated uses of computing technology. Since the early 1990s, educators have been moving toward technology integration and toward pre-service teacher training, emphasizing alignment of the curriculum with the technology. We must take what we have learned forward as we consider how, when, and with whom to implement DGBL in the future.