What Counts as Learning?

Defining what constitutes game-based learning in DGBL necessitates a recognition of the contested nature of learning itself as a concept. Broadly speaking, there are two general metaphors that have been historically utilized by the educational research community to define and describe learning (Sfard, 1998). The first view characterizes learning as acquisition, be it the acquisition of knowledge, skills, dispositions, behaviors, or other outcomes deemed desirable by educators and institutions. This generally aligns with what scholars call a cognitive perspective on learning, which emphasizes internal mental processes such as memory, decision-making, thinking, and knowledge construction as fundamental to learning (Slavin, 2018). It also speaks to behaviorist conceptions of learning, which tend to emphasize observable behaviors as indicators of learning (Nasir et al., 2021). The second metaphor characterizes learning as participation and emphasizes processes of socialization, mentorship, participation, and social practice. Within the field of education, this metaphor generally informs scholarship identified with a sociocultural perspective on learning, which tends to highlight social practices, identities, and relationships as core to learning (National Academies of Science, Engineering, and Medicine, 2018). In her discussion of these metaphors, Sfard (1998) was quick to note that these metaphors for learning are not meant to be mutually exclusive of one another and instead made the argument that understanding both is necessary for developing a robust understanding of what constitutes learning. And as this article describes later, conceptualizations of DGBL, in research and everyday practice, can generatively be discussed in terms of their positioning on a spectrum defined by these metaphors.

What Is a Game?

Turning to discussions of what constitutes “games” in relation to learning, it is first important to recognize that across a range of disciplines and perspectives, definitions have varied. Should games be defined from the standpoint of their constituent parts, such as the rules, logics, and systems that underlie them? Or should they be defined by the ways they are taken up by the people who play them? And what of the role of games within culture, such as in the formation of gaming communities and the development of affinity spaces around games? Terms that sometimes overlap with games, such as simulations, puzzles, and role play, also add to the challenge of establishing a distinctive definition. Synthesizing a number of these perspectives on games, play, and culture, game designers Katie Salen Tekinbaş and Eric Zimmerman (2003) defined a game as “a system in which players engage in an artificial conflict, defined by rules, that results in a quantifiable outcome” (p. 11). For Salen Tekinbaş and Zimmerman, a game is a system because it involves a set of things (rules, objects, players, etc.) that affect each other within an environment to form a larger pattern that is different from any of the individual parts. A game also involves players—active participants interacting with the system of the game in order to experience the play of the game. Although people may not always associate games with conflict, Salen Tekinbaş and Zimmerman argued that all games involve a contest of powers—though not all games need be a contest between players (e.g., solo games such as solitaire or cooperative games). All games involve rules that provide the structure out of which gameplay emerges, by delimiting what a player can or cannot do. Finally, this definition asserts that all games involve some kind of quantifiable outcome—in its simplest form, a state of victory or defeat. Such a definition is helpful because it distinguishes games from less formalized experiences of play while encompassing analog and digital games. In formulating this definition, however, Salen Tekinbaş and Zimmerman highlighted the limiting cases of puzzles and tabletop role-playing games, depending on how elements such as quantifiable outcomes are approached (for a more complete discussion, see Salen Tekinbaş & Zimmerman, 2003).

Commonly recognized examples of games include board games such as chess and Monopoly, card games such as Go Fish! and Magic: The Gathering, and role-playing games such as Dungeons and Dragons. Such games are sometimes referred to as analog games to distinguish them from their digital counterparts. As time has gone on, however, many games have begun to blur the lines between play that occurs in purely digital spaces and play that occurs in purely physical spaces. Notable examples of this include augmented reality games such as Pokémon Go! which feature a digital overlay of gameplay that invites users to explore the physical world around them in order to interact with the game’s virtual elements.

Digital Versus Analog Games

Just as analog games can take on a variety of forms, so too can digital games appear on a multitude of computer-based electronic platforms. These include games installed as software on personal computers, as well as games played through television-attached gaming consoles such as the Sony PlayStation or Microsoft Xbox. Digital game platforms also include dedicated handheld platforms, such as the Nintendo Gameboy and its various incarnations, as well as mobile smartphones and tablet computers. These games can be engaged as a single player, with multiple players in a single physical location, or with players across distances using online connective features.

Salen Tekinbaş and Zimmerman (2003) suggested there are at least four key traits that distinguish digital games from non-digital ones. Each trait has important implications for DGBL across a variety of contexts. The first is that digital games can offer experiences of interactivity that are generally characterized by both immediacy and focus. This is largely by virtue of the fact that digital games are programmed into computer consoles. Programming provides increased immediacy of feedback because the system is automated to respond to player input. Programming also provides a narrowed range of interactivity because possible player inputs and outcomes are limited by what a digital game developer has programmed into the software. Second, digital games and digital media more broadly make use of a computer’s affordances for collecting, storing, retrieving, manipulating, revealing, or even concealing information in the form of images, text, animations, sound, and even player behavioral data (Murray, 1997, 2017). Third, due to their computational nature, digital games have the potential to automate systems and procedures that may be far more complex for non-digital games to simulate. Examples of these complex systems include simulated environmental features in computer-based role-playing games: terrain–movement interactions, lines of sight, weather cycles, and even interactions with animals or non-player characters in the virtual game world (Dunnigan, 2000). Finally, many (although not all) contemporary digital games have the ability to facilitate communication between players across networks, such as the online networks that connect players in multiplayer, internet-connected games. These communication-facilitating features can include technologies that enable voice or text chat between players within a particular game.

What Is Game-Based Learning?

A critical appraisal of the game-based aspect of DGBL can help better situate the usage of this concept in theory and practice. Building on Sfard’s (1998) learning as an acquisition metaphor, one way to define game-based learning in DGBL is by the use of digital games to promote the acquisition of a variety of knowledge, skills, dispositions, or other educational outcomes. According to this definition, learning is game-based when digital games—or certain elements of digital games such as scorekeeping—are applied toward identifiable outcomes defined by educators or institutions. Examples of this can be seen in games that are explicitly designed for education, such as the Reader Rabbit computer game series (Houghton Mifflin Harcourt, 1983) for addressing aspects of reading and the more contemporary DragonBox game series (WeWantToKnow, 2012) designed to teach about mathematical concepts and procedures. Further examples can be seen through research on COTS games, such as the Civilization series, that are utilized in educational settings (Squire, 2004; Van Eck, 2015).

Another framing of DGBL, drawing on Sfard’s (1998) learning as participation metaphor, characterizes the game-based aspect of DGBL as an inherent part of meaningfully participating in—that is, playing—games, regardless of their connection to some external educational outcome (Koster, 2013). According to this definition, learning is always happening when someone plays a game. Having to solve problems, overcome obstacles, or collaborate with others within and around games are all dimensions of learning. As game designer and theorist Raph Koster (2013) noted in A Theory of Fun, a game is only “good” to the degree that it is “fun,” and a game is only fun to the degree which players are being challenged to learn to be successful in that game space. Scholars who adopt this perspective tend to examine game-based learning as it occurs within video game worlds, such as among player groups in the massively multiplayer role-playing game (MMORPG) World of Warcraft (e.g., Nardi, 2010). In addition to examining commercial games, however, advocates of this approach have also designed games to address real-world challenges, such as facing the everyday challenges of a world in which oil is no longer in ready supply (McGonigal, 2011).

Gamification, often categorized as a subconcept of DGBL, describes the implementation of certain aspects of digital games—visual displays of progress, automated reward systems, scoring leaderboards, virtual avatars, and others—into non-gaming contexts such as education, business management, marketing, and even physical exercise programs (Deterding et al., 2011). Proponents of gamification often cite research demonstrating the positive impacts of these interventions on outcomes such as academic performance, student motivation, customer retention, organizational productivity, or user engagement in online platforms (Hamari et al., 2014). Critics of these approaches, however, argue that gamification applications are not about game design but instead center behavior modification techniques—with some going so far as to label these approaches as “exploitationware” (Bogost, 2011). Audrey Watters (2014), a historian of educational technology, drew a distinction between game-based learning and gamification, citing similar differences between purposeful game design and approaches rooted in previously established approaches to behaviorism in education.

Developing what Salen Tekinbaş and Zimmerman (2003) referred to as a “critical discourse” on DGBL will be an essential part of moving the field forward, as well as addressing some of the empirically supported and philosophically grounded critiques and challenges facing the field. Perhaps just as important is an understanding of some of the historical and ideological roots of the field, to which this article now turns.

Positivist Approaches

Scholarship that applies empirical approaches such as experimental research and statistical analysis to measure the relationship between game-based interventions and learning outcomes can be classified broadly as taking a positivist approach to the study of DGBL. Such research generally models itself after the classical “scientific method” of (a) observing a phenomenon, (b) generating a hypothesis to be tested, (c) conducting experiments, (d) analyzing results, and (e) drawing conclusions based on the findings. Such research also typically focuses on measurable outcomes of learning that are valued in formal schooling, such as content knowledge in a particular school subject, attitudes toward a given academic subject of study, or motivation to engage in school-based learning. An example of this is Siew et al.’s (2016) quasi-experimental study comparing the algebraic thinking of students playing the mobile game DragonBox with a control group, which was “taught algebra using conventional methods involving imitation and repetition” (p. 70). The study found statistically significant differences between experimental and control groups on measures of both algebraic thinking and attitude toward the subject, in favor of the DGBL group.

In addition to individual studies, meta-analyses of DGBL research also reflect a more positivist framing. Meta-analyses use statistical techniques to compare multiple scientific studies and are becoming increasingly more useful as empirical studies of DGBL proliferate. An example of this work is Clark et al.’s (2016) meta-analysis of 69 studies of DGBL providing information on 6,868 unique participants, which found that game-based approaches improved learning outcomes compared to non-game approaches. The meta-analysis also found significant learning outcomes in both cognitive and intrapersonal learning with the use of video games rather than traditional instruction in classrooms. In other words, video games not only helped children learn but also improved motivation, intellectual openness, work ethic and conscientiousness, and positive self-evaluation. The gains are equivalent to having a good teacher but not as effective as strong teaching methods such as reciprocal teaching or problem-based learning.

Games can be used as stand-alone approaches to teaching content, although this does not seem to be particularly effective. Games are not a replacement for teachers but are most effective when paired with teacher-provided scaffolding without heavy internal stories—that is, they work best when well integrated into a classroom environment rather than disconnected. They also do not need to be complex, immersive, or use advanced graphics—in fact, having realistic graphics seems to have a negative effect on learning, as does having a relevant storyline rather than an irrelevant one or none at all. Ongoing research is examining which kinds of digital games and game-based instruction are more effective than others. Other meta-analyses have reported similarly moderate advantages of game-based learning compared to other instructional approaches (Sitzmann, 2011; Vogel et al., 2006; Wouters et al., 2013).

Despite these findings, researchers caution that due to the wide range of conditions, study designs, and quality concerns, it is challenging to make conclusive claims about the effectiveness of DGBL in comparison to other pedagogical approaches (All et al., 2014, p. 6).

Interpretive Approaches

A second broad category of DGBL scholarship is less concerned with measuring academic outcomes and effect sizes and instead focuses on the ways players create and exchange meaning through games, the social practices that occur within and around games, or issues more related to a more participatory view of learning. This body of DGBL scholarship can be considered more interpretive in nature because it typically involves the interpretation of meaning behind different game-play practices, social exchanges, and participatory cultures that develop around digital games (Jenkins et al., 2009).

Interpretive studies of DGBL can focus on the experiences of individual players and player communities within games and virtual worlds. Gee and Hayes’ (2010) work on women and girls playing a computer game called The Sims, for example, highlighted several case studies of female players of all ages developing a sense of meaning, belonging, and creative agency through their engagement with the game and its online community of players. Holmes’ (2015) study of the online multiplayer battle arena game DOTA2 extended the field of DGBL by highlighting the importance of teaching, including design, mentorship, modeling, and feedback sharing within player communities. Importantly, interpretively framed studies of DGBL often explore phenomena beyond the “small g” game experience, providing insights into the broader affinity spaces that connect gaming to other social practices. An affinity space can be understood as a location, physical or virtual, where groups of people are drawn together because of a shared, strong interest or engagement in a common activity (Gee, 2005). In one such study, focused on online discussion forums centered on the MMORPG World of Warcraft, Steinkuehler and Duncan (2008) mapped out the ways that such game-based affinity spaces can foster social knowledge construction within player communities. Tran’s (2018) research on families playing the augmented reality game Pokémon Go! highlighted how participants can bridge affinity spaces to create even broader, more distributed experiences of teaching and learning.

Critical Approaches

A third broad category of DGBL scholarship can be classified according to a more critical paradigm. Looking beyond the measurement of learning outcomes and descriptive interpretation, critical DGBL scholars seek to confront the social, historical, and ideological forces and structures that produce and constrain cultural phenomena involving games and learning. In line with its foundations in sociology, political philosophy, and literary criticism, this smaller but growing area of DGBL scholarship challenges both positivist and interpretivist paradigms by aiming not just to reveal issues of power and ideology within DGBL but also to actively challenge them (Gray & Leonard, 2018).

An example of critically framed scholarship in DGBL is Bogost’s (2015) Why Gamification Is Bullshit, which identified the ways that gamification, or the application of “game-like” elements such as point systems and leaderboards to non-game contexts, works as a way to exploit the behaviors, labor, and psychological responses of participants—whether they be students in school or workers in a factory. These ideas align with the work of educational historian Audrey Watters (2014), which highlighted how the development and application of many educational technologies, including games, have tended to be grounded in principles of behaviorism and paradigms centering control, discipline, and punishment. In response to these issues, Woodcock and Johnson (2018) have called for a reframing of gamification, play, and the denial of labor as acts of resistance to these exploitative pressures and structures.

These structural elements of DGBL are not the only focus of critical scholarship in this area, however. In line with broader trends in critical theory, scholars have also turned their attention to addressing issues of representation, accessibility, and equity with various gaming cultures. The edited volume Woke Gaming (Gray & Leonard, 2018) highlights many of these issues, as well as offering alternatives and challenges to racism, misogyny, stereotyping, and other forms of oppression within the contexts of production, reception, dissemination of games, as well as the design of games. Taken together, this emerging body of scholarship emphasizes the deeply rooted connections between games, learning, education, and society, rather than treating these as separate subjects to be described or analyzed.

As educational structures, technologies, and cultural phenomena continue to evolve, scholarship on DGBL continues to expand across positivist, interpretive, and critical paradigms, highlighting new issues to explore, as well as re-examining prior assumptions in light of new evidence.

Commercial Entertainment Games

The first category of these applications describes the integration of COTS games into classroom learning. COTS games are not necessarily designed for educational purposes that educators can position to support certain educational outcomes. Examples of this include the work of Squire (2004) to teach world history and historical thinking through the real-time strategy game Civilization III. In another example, researchers found that the three-dimensional exploration and puzzle-solving game Portal 2 was able to help participants improve on a variety of cognitive and noncognitive measures, including spatial awareness (Shute et al., 2015). In some cases, educational applications of COTS can lead to the development of educationally oriented versions of an originally COTS game, as in the case of Minecraft EDU (Cózar-Gutiérrez & Sáez-López, 2016). When implementing COTS game-based learning in the classroom, Van Eck (2015) argued for the importance of teachers and instructional designers being technically familiar both with games in general and with the game they have chosen to use in particular. This fundamental orientation then facilitates scaffolded design, in which other course materials (e.g., readings and worksheets), classroom practices (e.g., homework and group activities), and COTS games work in concert to support specific educational outcomes.

“Educational” Games

The case of Minecraft EDU leads us to a second category of DGBL applications—that is, games developed explicitly for the purpose of teaching educational content. An early example of these educational games is the Reader Rabbit computer game series for addressing aspects of literacy; an example of more contemporary games is the DragonBox game series, designed to teach about mathematical concepts and procedures. Games such as Quest Atlantis have been created to teach content through immersive open worlds in which players take on professional roles and have to understand disciplinary content to succeed within the game. In these cases, educators offer these games for play to replace traditional classroom instruction, provide these games as a supplement to traditional pedagogies during the school day, or assign the play of these games as homework to be completed outside the classroom.

Many of the most effective and popular learning games provide brief and engaging opportunities for learners to engage with distinct content areas and discrete skills—rather than within an immersive world—but still in an environment that allows for authentic professionalism. These games should be played over time; multiple gaming sessions show clear gains over single-play sessions. In addition, games that track achievements and progress, such as by awarding points or badges, are more effective than ones that do not. These games can be used as opportunities for focused practice within units. For example, there are games that teach fractions through manipulating geometric shapes to advance through levels; puzzle games in which students explore states of matter; and a game that allows students to take on the role of a virus to learn about pathogens. The work of Filament Games is a good example of this approach.

Games themselves—particularly virtual worlds—can also be used as environments for instruction. When simply approached as a virtual classroom, this misses the point of using games, except when the environment allows for collaboration or feats not possible otherwise. Minecraft, for example, is often used by students to construct or explore historic sites, re-enact scenes from literature, or experience scientific principles such as geology or quantum physics.

Student Game Design

Students can learn through designing video games, which is an especially promising approach given that tools for creating games are increasingly accessible to children. Seymore Papert, one of the progenitors of computer-based educational approaches, developed a specific programming language for this purpose (Solomon & Papert, 1976). In addition to learning design and programming skills, which is valuable, students can also meaningfully grapple with content by designing video games. A similar approach is through “modding” (modification) of existing video games—many games, including Minecraft and Civilization, allow for the redesign of particular aspects of the games for a given purpose. Game-making and modding allow players to gain further insight by engaging directly with the mechanics and content of a game and also more general principles of systems design, multiple literacy practices, and computational thinking (Peppler & Kafai, 2007). Although such approaches may sometimes necessitate technical and design expertise on the part of the teacher or facilitator, research has demonstrated that students can also draw on a variety of other resources and practices—peer relations, prior gaming experience, and online tutorials—to support more independent learning (Aguilera, 2021). Beyond the digital, research and educational projects have also involved students in the creation of physical board games, demonstrating how students can engage in complex procedural, narrative, and ideological design, even without computational technology (Kessner et al., 2021).

Gamification and Gameful Design

The final trend in DGBL applications to be discussed in this article relates to the overlapping practices of “gamification” and “gameful” design. The term gamification was popularized in part through a paper by Deterding et al. (2011, p. 9); it is described by the authors as “an informal umbrella term referring to the use of video game elements in non-gaming systems to improve user experience and engagement.” In this framing, game elements include game interface design patterns (e.g., leaderboards), game design patterns (e.g., time constraints or turn-taking), and game design principles. As educators and/or designers shift across the spectrum from game elements to design principles, Deterding et al. (2011) referred instead to a practice of gameful design, taking inspiration from games to inform the (re)design of contexts such as classrooms and workplaces. Instructive examples of such gameful design include Shute and Ventura’s (2013) work on stealth assessment, which embedded assessment into digital games to evaluate student progress toward a particular goal or measure, facilitating increasingly evidence-based design decisions.

Educational Critiques

Among researchers who reflect a more positivist perspective on DGBL, there is an emerging consensus that the variation in quality of research studies examining the effectiveness of DGBL for improving learning outcomes makes it difficult to draw definitive conclusions about whether DGBL is superior to other forms of educational intervention. Researchers seeking to examine the effectiveness of an intervention typically point to experimental research as the “gold standard” for such studies. One hallmark of this approach is a pre–post test design, which tests performance on a variable of interest before and after an intervention has been implemented. A second key feature of such research is the use of a control group that does not receive the intervention, the results from which can be compared to those of the experimental group. Randomized selection and assignment of participants to control and experimental groups can support the trustworthiness of these studies because these practices can help control “lurking” variables that might otherwise explain change beyond the intervention (Campbell & Stanley, 1973). And despite some claims of proponents of DGBL, many research studies in this area do not employ these methodological features, or they employ them to a limited extent. One methodological review found that out of 54 publications “dealing with effectiveness of DGBL aimed at cognitive learning outcomes,” fewer than half (25 studies) employed a pre–post test design and a control group for comparison (All et al., 2014, p. 6). Of the 25 pre–post controlled studies, there was still much variation across four key aspects of study design: participants, intervention, methods, and measures.

Second, although there is clear potential to use video games for learning, the real-world realities of their use is often quite different. The design of games that achieve a balance between learning and fun continues to pose a challenge. The basic requirement that good games need an artful combination of mechanics and content has meant that many games for learning simply are not effective. Good games take substantial time, resources, and expertise to develop. One emerging solution is meaningful collaboration between professional game designers, software engineers, and learning experts for designing such games. These games often introduce elements such as analytics to adapt game experiences or classroom curricula to gamers’ actions and learning trajectories.

Finally, there are issues of unequal access and application. Because video games are so reliant on particular technologies and infrastructures, less resourced classrooms are at a significant disadvantage in trying to integrate video games into pedagogy, often spending scarce funds for hardware or software in hopes of raising student achievement. However, using video games for learning requires well-educated and supported teachers, although it is still less empirically supported than time-tested social practices. Finally, an increasing number of games for learning are being developed by companies and academic researchers with the goal of making a profit off of quick-fix promises, so making well-informed decisions is important to avoid perpetuating inequalities.

Sociological Critiques: Is All Learning Good Learning?

Some researchers have argued we cannot just talk about what is effective and what is not. We also have to take ethical stances. Learning may happen within and around video games, but is all learning good? Although gaming culture and communities have been examined as rich areas of language and literacy development/exchange, many of these spaces are fraught with problematic interactions, movements, and ideologies. These include ample exploitative, racist, homophobic, and sexist tendencies within games themselves and in the communities that surround them.

In particular, Gray and Leonard (2018) have noted how video games and their communities have become breeding grounds for fascist (“alt-right”) activism. The origins of the popular recognition of these trends may be traced to the Gamergate controversy, a harassment campaign targeting several women in the video game industry, notably game developers Zoë Quinn and Brianna Wu, as well as feminist media critic Anita Sarkeesian. However, in a much broader sense, these phenomena are manifestations of the “culture wars” tied to patriarchal domination and White supremacy, specifically over cultural diversification, artistic recognition, and social criticism in video games and also over the social identity of gamers.

Relatedly, DGBL and the video game industry more broadly often forward corporatized, neoliberal agendas of control and surveillance. This may be most apparent in gamification efforts, but it also exists in newer forms of labor extraction that are part of video games, including educational ones. The gaming industry benefits from players creating the content of games, from “mods” (add-on modifications) for commercial games to massive curated communities around the games. Although it is not labor exploitation, at least in the classic sense, gamers are enrolled into a system that benefits them unequally and hides or poorly compensates their efforts (Ekbia & Nardi, 2017). These include forms of communicative, cognitive, creative, emotional, and organizing labor that we are “pulled” into with video games and gaming communities that engage us more often as a “calling” to creative self-fulfillment or social participation in various projects. These systems both shape and direct our behavior and invisibly extract economic value while not fully acknowledging or rewarding our labor.

This highlights a tension that has existed with digital technologies and video games since their creation, as simultaneously tied to consumerism and the military–industrial complex, while also creating spaces for expansive cultural practices and liberation.