Bodong Chen

Crisscross Landscapes

Notes: Game-like learning



Citekey: @gee2008game

Gee, J. P. (2008). Game-like learning: An example of situated learning and implications for opportunity to learn. Assessment, equity, and opportunity to learn, 200–221.


Game-Like Learning (p. 1)

knowledge: as noun and verb (p. 1)

The theory of learning in many schools today is based on what I would call the “content fetish” (Gee 2004). The content fetish is the view that any academic area (whether physics, sociology, or history) is composed of a set of facts or a body of information and that the way learning should work is through teaching and testing such facts and information. (p. 1)

“know” is a verb before it is a noun, “knowledge” (Barsalou 1999a, 1999b; Bereiter and Scardamalia 1993; Clark 1997; Glenberg 1997; Glenberg and Robertson 1999; Lave and Wenger 1991; Rogoff 1990). (p. 1)

Any actual domain of knowledge, academic or not, is first and foremost a set of activities (special ways of acting and interacting so as to produce and use knowledge) and experiences (special ways of seeing, valuing, and being in the world). (p. 1)

The fruitful patterns or generalizations in any domain are those that are best recognized by those who already know how to look at the domain and how the complex variables at play in the domain interrelate with each other. This is precisely what the learner does not yet know. (p. 2)

An interesting comment KB will not agree with. (p. 2)

Yet as we have already said, simply turning learners loose to engage in the domain’s activities won’t work either, because newcomers don’t know how to start, where to look for the best leverage, and which generalizations to draw or how long to pursue them before giving them up for alternatives. Of course, we can hardly expect learners to reinvent for themselves domains that took thousands of people and hundreds of years to develop. (p. 2)

This paradox has lead some educators, over the last few years, to search for what I would call “post-progressive pedagogies”; that is, pedagogies that combine immersion with well-designed guidance (e.g., Brown 1994; Lehrer 2003; Lehrer and Schauble 2005; Martin 1990). (p. 2)

good video games and related types of simulations can play in learning inside and outside schools (e.g., Barab et al. 2005; Barab et al. in press; Gee 2003a, 2005; Jenkins and Squire 2004; Shaffer 2007; Squire 2005, 2006; Steinkuehler 2004, 2006). (p. 2)

I will point out that the dilemma we discussed earlier – between knowledge as information and knowledge as activity and experience – is related to another dilemma familiar from recent research on cognition: the dilemma between general, abstract, and verbal understandings, on the one hand, and situated understandings, on the other. (p. 2)

A situated understanding of a concept or word implies the ability to use the word or understand the concept in ways that are customizable to different specific situations of use (Brown, Collins, and Dugid 1989; Clark 1989, 1993, 1997; Gee 2004). A general or verbal understanding implies an ability to explicate one’s understanding in terms of other words or general principles but not necessarily an ability to apply this knowledge to actual situations. (p. 3)

Research in cognitive science has shown, for example, that it is perfectly possible to understand Newton’s laws as formulas, realizing their deductive capacities in a general way, but not be able to actually draw these deductions and apply them to a concrete case in actual practice to solve a real-world problem (Chi, Feltovich, and Glaser 1981; Gardner 1991). (p. 3)

all human understandings are, in reality, situated. (p. 3)

Verbal and general understandings are top-down. (p. 3)

Situated understandings generally work in the other direction; understanding starts with a relatively concrete case and gradually rises to higher levels of abstraction through the consideration of additional cases. (p. 3)

knowledge as activity and experience before knowledge as facts and information (p. 3)

the possibility that what we might call “game-like” learning through digital technologies can facilitate situated understandings in the context of activity and experience grounded in perception (Games-to-Teach Team 2003; Gee 2003a; McFarlane, Sparrowhawk, and Heald 2002; Squire 2003). (p. 4)

game-like learning: andy disessa (p. 4)

Andy diSessa’s (2000) work is a good example, in science education, of building on and from specific cases to teach situated understandings. (p. 4)

DiSessa has successfully taught children in sixth grade and beyond the algebra behind Galileo’s principles of motion by teaching them a specific computer programming language called Boxer. (p. 4)

The students write into the computer a set of discrete steps in the programming language. (p. 4)

Once the program starts running, the student will see a graphical object move one meter per second repeatedly, a form of uniform motion. (p. 4)

The student can keep elaborating the program and watch what happens at every stage. In this process, the student, with the guidance of a good teacher, can discover a good deal about Galileo’s principles of motion through his or her actions in writing the program, watching what happens, and changing the program. (p. 5)

an embodied way, tied to action, how a representational system that is less abstract than algebra or calculus (namely, the computer programming language, which is actually composed of a set of boxes) “cashes out” in terms of motion in a virtual world on the computer screen (p. 5)

DiSessa does not actually refer to his work with Boxer as game-like learning, though some people pushing the design of actual games for learning have been inspired, in part, by his approach to learning and science education (p. 5)

Supercharged! (p. 6)

Kurt Squire and his colleagues (Squire et al. 2004; see also Jenkins, Squire, and Tan 2003; Squire 2003) have worked on a computer game called Supercharged! to help students learn physics. (p. 6)

an electromagnetism simulation game developed in consultation with MIT physicist John Belcher by the Games-to-Teach project at MIT (p. 6)

placing charged particles and controlling a ship that navigates by altering its charge. (p. 6)

The game play consists of two phases: planning and playing. (p. 6)

Each level contains obstacles common to electromagnetism texts. These include points of charge, planes of charge, magnetic planes, solid magnets, and electric currents. (p. 6)

Squire et al. (2004) (p. 6)

In this study, the experimental group outperformed the control group on conceptual examination questions. Post-interviews revealed that both experimental and control students had improved their understanding of basic electrostatics. However, there were some qualitative differences between the two groups. The most striking differences were in students’ descriptions of electric fields and the influence of distance on the forces that charges experience. (p. 6)

the teachers came to realize that students were initially playing Supercharged! without a good deal of critical reflection on their play. The teachers then created log sheets for their students to record their actions and make predictions, which reinforced the purpose of the activity and encouraged students to detect patterns in their play. Later, the teachers provided even more structure, using the projector to display game levels, encouraging the class to interpret the events happening onscreen and make predictions about how they thought the simulation would behave. (p. 7)

Full Spectrum Warrior (p. 8)

There are a plethora of people today who want to make “serious games” for learning (for more information, see or www. (p. 8)

Good commercial games are more or less forced to incorporate good principles of learning (Gee 2003a). Today’s video games are long, complex, and hard – and avid players would not have it any other way. (p. 8)

deep learning involves, first and foremost, activity and experience, not facts and information. (p. 8)

A large body of facts that resist out-ofcontext memorization and rote learning comes free of charge if learners are immersed in activities and experiences that use these facts for plans, goals, and purposes within a coherent knowledge domain (Shaffer 2004). (p. 8)

simply turning learners loose to engage in the domain’s activities won’t work either. (p. 8)

Unfortunately, our schools are still locked into endless and pointless battles between “traditionalism” and “progressivism,” between overt teaching and immersive learning, between skill-and-drill and activities, as though these were the only two alternatives. (p. 8)

Many good commercial video games are based on a theory of learning I will call “distributed authentic professionalism,” a theory that resolves our paradox quite nicely (see also Shaffer 2004, 2007). (p. 8)

Full Spectrum Warrior (p. 9)

this game is ideologically laden. It carries messages, beliefs, and values about war, warfare, terrorism, cultural differences, the U.S. military, and the role of the United States and its army in the modern, global world. (p. 9)

Full Spectrum Warrior has its origins in a U.S. Army training simulation, but the commercial game retains only about 15% of what was in the Army’s simulation (Buchanan 2004, 150). (p. 9)

Full Spectrum Warrior teaches the player (yes, it is a teacher) how to be a professional soldier. It demands that the player think, value, and act like one to “win” the game. (p. 9)

These additional skills are a version of the professional practice of modern soldiers – the professional skills of a soldie (p. 9)

soldier commanding a dismounted light infantry squad composed of two teams. (p. 9)

“Everything about your squad … is the result of careful planning and years of experience on the battlefield. Respect that experience, soldier, since it’s what will keep your soldiers alive” (p. 2). (p. 9)

The virtual characters in the game (the soldiers in the squads), on the one hand, and the real-world player, on the other hand, control different parts of the domain of professional military expertise. We get the whole domain only when we put their knowledge together. The knowledge is distributed between them. A human being (the player) shares knowledge with a virtual reality (the soldiers). (p. 9)

The game only works when the two different bits are put together – thought about and acted on – as a whole by the player who uses the virtual soldiers as smart tools or resources (p. 10)

The player is also scaffolded by some quite explicit instruction given “just in time, (p. 10)

The learner is not left to his or her own devices to rediscover the foundations of a professional practice that took hundreds of years to develop. Our paradox is solved. (p. 10)

“authentic professionalism.” Authentic professionals have special knowledge and distinctive values tied to specific skills gained through a good deal of effort and experience (p. 10)

Finally, professionals welcome challenges at the cutting edge of their expertise (Bereiter and Scardamalia 1993). (p. 10)

What will be the values in a KB game? (p. 11)

What all of these games exemplify, though, is that there is no real learning without some ideology. Adopting a certain set of values and a particular worldview is intimately connected to performing the activities and having the experiences that constitute any specific domain of knowledge. Physicists hold certain values and adopt a specific worldview because their knowledge making is based on seeing and valuing the world in certain ways. (p. 11)

A good school-based learning experience that followed the Full Spectrum Warrior model would have to pick its domain of authentic professionalism well, intelligently select the skills and knowledge to be distributed, build in a related value system as integral to learning, and give explicit instruction only (p. 11)

“just in time” or “on demand.” David Shaffer’s “epistemic games,” one of which we will discuss below, exemplify this approach. (p. 12)

augmented by reality: madison 2020 (p. 12)

Madison 2020 project (p. 12)

a game-like simulation that simulates some of the activities of professional urban planners (Beckett and Shaffer 2004; see also Shaffer et al. 2004) (p. 12)

Shaffer and Beckett’s game is not a stand-alone entity but is used as part of a larger learning system. Shaffer and Beckett call their approach to gamelike learning “augmented by reality,” because a virtual reality – that is, the game simulation – is augmented or supplemented by real-world activities; in this case, further activities of the sort in which urban planners engage. (p. 12)

As in the game SimCity, in Shaffer and Beckett’s game, students make landuse decisions and consider the complex results of their decisions. However, unlike in SimCity, they use real-world data and authentic planning practices to inform those decisions. (p. 12)

based on David Shaffer’s theory of pedagogical praxis, a theory that argues that modeling learning environments on authentic professional practices – in this case, the practices of urban planners – enables young people to develop deeper understandings of important domains of inquiry (Shaffer 2004). (p. 12)

The emphasis, however, is not on professions as vocations but as domains of expertise that recruit important ways of knowing and producing knowledge; thus, Shaffer calls his games “epistemic games” (Shaffer 2007). (p. 12)

Shaffer and Beckett’s Madison 2020 project situated student experience at a micro level by focusing on a single street in their own city (Madison, Wisconsin): (p. 13)

The high school students Shaffer and Beckett worked with had volunteered for a ten-hour workshop (run over two weekend days) focused on city planning and community service. At the beginning of the workshop, the students were given an urban planning challenge: They were asked to create a detailed redesign plan for State Street, a major pedestrian thoroughfare in Madison, a street quite familiar to all of the students in the workshop. (p. 13)

Students then watched a video about State Street, featuring interviews with people who expressed concerns about the street’s redevelopment aligned with the issues in the informational packet (e.g., affordable housing). (p. 13)

students walked to State Street and conducted a site assessment (p. 13)

MadMod – the “game” in the learning system – allows students to see a virtual representation of State Street. It has two components, a decision space and a constraint table. (p. 14)

As students made decisions about changes they wished to make, they received immediate feedback about the consequences of changes in the constraint table. The constraint table showed the effects of changes on six planning issues raised in the original information packet and the video: crime, revenue, jobs, waste, car trips, and housing. (p. 14)

MadMod functions in Shaffer and Beckett’s curriculum like a game much in the way SimCity does. In my view, video games are simulations that have “win states” in terms of goals players have set for themselves. (p. 14)

In this case, the students have certain goals, and the game lets them see how close or far they are from attaining those goals. At the same time, the game is embedded in a learning system that ensures those goals and the procedures used to reach them are instantiations of the professional practices and ways of knowing of urban planners. (p. 14)

Better yet, perhaps, Shaffer and Beckett were able to show transfer: Students’ responses to novel, hypothetical urban planning problems showed increased awareness of the interconnections among urban ecological issues. (p. 14)

assessment: a game example (p. 15)

real-time strategy video games (p. 15)

“Why shouldn’t learning school subjects be more like playing Rise of Nations? Why shouldn’t assessment work in school the way it does in Rise of Nations?” (p. 16)

Why, then, would we need any assessment apart from the game itself? One reason – indeed, a reason Janie herself would – is that Janie might want to know (p. 16)

Of course, the game is very complex, so this won’t be any particular score or grade. What Janie needs is a formative or developmental assessment that can let her theorize her play and change it for the better, and this is what the game gives her. (p. 16)

At the end of any play session in Rise of Nations, the player does not just get the message “you win” or “you lose,” but rather a dozen charts and graphs detailing a myriad of aspects of her activities and strategies across the whole time span of her play (and her civilization’s life). This gives Janie a more abstract view of her play; it models her play session and gets her to see her play session as one “type” of game, one way to play the game against other ways. It gives her a meta-representation of the game and her game play in terms of which she can become a theoretician of her own play and learning. From this information, she does not learn just to be faster or “better”; she (p. 16)

learns how to think strategically about the game in ways that allow her to transform old strategies and try out new ones. She comes to see the game as a system of interconnected relationships. (p. 17)

Janie will see herself compared, at each stage of the game play, with the other players (real people or the computer) in each chart and graph: (p. 17)

Janie uses these charts and graphs – they are part of the game play, part of the fun of the game – to understand where things went right and where they went wrong, where things can be improved and where no change is needed (p. 18)

She can even look at the charts and graphs and conclude, not that there were weaknesses in her performance, but that she won by a certain style and would like now to try another one. This is formative or developmental assessment at its best. (p. 18)

implications (p. 18)

Of course, “fair” may not be the right word here, as Pullin has pointed out to me (personal communication). As she points out, “things can be fair; that is, equitably distributed, but offered at a very low level.” (p. 19)

She prefers the term “meaningful opportunity to learn” rather than “fair opportunity to learn.” (p. 19)

when students engage in situated learning of the sort discussed here, facts and information eventually “come free” (e.g., Gee 2003a; Shaffer 2007). (p. 19)