How Computer Games Help Children Learn – Chapter 6

Shaffer, D. (2006). How Computer Games Help Children Learn. New York: Palgrave Macmillan.

The final chapter brings out several challenges in designing epistemic games:

  • games are built on simulations which are inevitably simplified (and thus distorted) views of the world
  • simulations without a community of practice and without the opportunity for reflection and feedback offer no real context
  • professions are built on practices which are evolved rather than designed
  • these professional practices do not offer “general principles of learning that can be used anywhere;” instead they provide markers
  • “learning takes place only as part of a coherent system” and thus we will fail if we merely extract professional practices (or game elements) and drop them into existing curricula

The proposal for developing a third place which is neither home (family) nor school made sense although the concept of designing such a place (or space) makes less sense than simply recognizing that the space already exists and has been well-described by danah boyd. How we reach into that space with being “creepy” is the challenge.

The final challenge offers the most fruitful direction for future research: which professions are the most fundamental? Because we cannot know what jobs will even exist in 20 years, this question seems impossible to answer. But it’s not. Every epistemic frame has value; learning how to be a journalist teachers students how to be more than a journalist. Thus, the answer is that we should create a curriculum based on a broad taxonomy of professions and allow students to choose based on personal interests. Some games will model analytical thinking skills with particular application in more quantitative professions; others will model behavioral skills with use in more social endeavors. The unknown professions of tomorrow will require different blends and different ratios of these skills, but if we can craft a new curriculum with the right breadth, the specific job title won’t matter. What will matter is that our students will have learned how to think, how to learn, and how to innovate no matter what the future brings.

Articles for further study from the Notes Section:

Taxonomy – Bartle, R. (1996). Hearts, clubs, diamonds, spades. Journal of MUD Research 1(1).

Games – Donald, M. (2001). A mind so rare. New York: W.W. Norton

Microworlds – Hoyles, et.al., (2002). Rethinking the microworld idea. Journal of Educational Computing Research 27 (1&2).

Schema – Dreyfus & Dreyfus (1986). Mind over machine. New York: Free Press.

How Computer Games Help Children Learn – Chapter 2

Shaffer, D. (2006). How Computer Games Help Children Learn. New York: Palgrave Macmillan.

The Digital Zoo game starts to tie together game qualities with authentic professions (in this case, mechanical engineering). The game is a design simulation with no right answer; goals are realistic tasks (such as build a bridge), and levels are introduced through increasingly difficult requests from clients. Accomplishing each task requires designing, building, and testing various solutions–the basic steps in engineering design–with each solution producing an incremental improvement. Concepts such as center of mass and bracing are learned (and integrated into the students’ deep understanding) during the process.

Students took pre- and post-tests and kept a design notebook which allowed the researchers to investigate numerical and attitude changes; the results were impressive:

  • “players used scientific justifications, on average, five times as often”
  • “design plans became, on average, 55 percent more complex”
  • “Players considered 47 percent more features in making a decision”

Players learned specialized language as part of acting like (and thus learning to think like) an engineer. The language provides labels which improves communication and also highlights what’s important and what’s not.

After describing the research with the game, Shaffer argues that the knowledge gained in the game persists because it’s tied to a particular epistemology. He describes knowledge as more than memory but as the use of symbols which represent knowledge. While words are efficient symbols because they can be written down and studied systematically, computers don’t just store symbols, they process them. While words led to a scientific culture based on symbol storage, computers will lead to as virtual culture based on symbol processing.

Shaffer argues that all games are simulations (microworlds), and that as humans and computers interact in these simulations, players display autoexpressivity. They come to a microworld with a set of beliefs, make decisions based on those beliefs, and and receive responses from the simulation which bring to the surface, challenge, and refine those beliefs. This concept is identical to Wiggins’ idea of eliciting misconceptions as the first step toward deep learning.

Instance-based and rule-based learning

Citation

Taatgen, Niels A. and Wallach, Dieter. “Whether Skill Acquisition is Rule or Instance Based is determined by the Structure of the Task.” Cognitive Science Quarterly: 2.2 (2002). 163 -204.

Summary

The traditional view of skill acquisition as “a gradual transition from behavior based on declarative rules in the form of examples and instruction towards general knowledge represented by procedural rules” is challenged by instance theory. While the latter seems to explain the inability of experts to verbalize rules, research on the directional asymmetry of rules seems to support the traditional importance of rules.

ACT-R’s instance-based architecture is based on 2 key arguments:

  • the strategy to use or the memory to retrieve is based on which has the highest expected gain (optimization)
  • declarative memory is activated (and filtered) by environmental demands and past experience (which has been encoded as production rules by procedural memory)

At the symbolic level of ACT-R, procedural rules are applied to declarative chunks which store information in a proposition; chunks are either new (perceptions) or created internally by prior knowledge/experience. Each rule contains a condition- and an action-part, and declarative items are pattern-matched to the condition and applied in the action. The subsymbolic level of ACT-R deals with the choice of which rule to apply according to Bayes’ Theorem (increases base-level activation each time it is retrieved and decays over time).

The article seems to imply that if instance-based learning fails (because the problem is too time-consuming–no discernible pattern), learners will attempt to derive some sort of rule. The authors argue that instance-based learning works best when the relationships between variables is very difficult; rule-based learning (simplified cases) is more successful with a large number of cases with obvious relationships.  The authors then test this hypothesis with two detailed experiments.

The results:

  • Previous research showed no learning through observation without direct rules (explicit relationships)
  • Subsequent research, indicated that exploratory participants did better than observers, but that observers could better verbalize and construct a causal model.
  • This research shows evidence of learning by observing even without rules; and participants seem better able to answer questions about old systems than new. Both results support instance-based learning.

Response

ACT-R seems to offer tremendous explanatory power. However, instance theory seems related to the concept of expert (tacit?) knowledge:

  • know when to apply
  • gets better over time (more instances)
  • gets better with use
  • additive for community

Because production rules that propose new declarative rules are not accounted for in the ACT-R architecture, this “missing link” may be the elusive transfer element; the authors propose partial-matching in ACT-R retrieval as a solution.

In instance theory, encoding and retrieval seem more closely linked to temporal/spatial reality than to attention as the authors claim; however, the view of memory as evolving from algorithmic processing to memory-based processing succinctly describes expert knowledge.

The recommendations for design strategies seem profound:

  • Instance-based learning takes over from rule-based over time
  • Creating declarative rules is the most important (first) step
  • Analogy works best at the start but declines quickly
  • Declarative rule works well at start and persists
  • Instance continues to improve over time to become best

Simulations in Online Instruction

Citation

Rude-Parkins, Carolyn, et. al. “Applying Gaming and Simulation Techniques to the Design of Online Instruction.” Innovate Online 2.2 (December, 2005 – January, 2006). Retrieved from the Web 02/06/2009 at
http://innovateonline.info/index.php?view=article&id=70&action=article (requires login).

Summary

Characteristics which distinguish an Army training simulation are described: using scenarios, keeping score, and allowing learners to control timing. Each lesson begins with real maps and photos to anchor the instruction “because online … training is already an abstraction.” The transition from concrete to more abstract representations, “is eased by the integration of increased visual cues” and the use of a consistent screen layout. Prior memory is engaged by using familiar real life acronyms and procedures.

Feedback is provided by running scores, consequential feedback in the form of outcomes for sub-optimal solutions, and demonstrations; the latter are provided on neutral terrain (a more trivial subset of the actual problem). Online delivery was selected because the content was so dynamic; however, this delivery choice required a trade-off with limited perspectives (top down or bird’s eye views) because of bandwidth limitations.

Learner satisfaction was high although comparative outcomes were not presented. Future enhancements include the use of “drag and drop” interactions, more consistent rules, and the possibility of adaptive testing as a scaffolding technique.

Reflection

While the authors argue that the training is not a simulation because learner choices are limited, the design of the instruction around a single, optimum solution suggests a simulation rather than a more open-ended game. However, the introduction of competitive teams would transform the experience into a compelling game with the corresponding introduction of competition as a motivating factor. In addition, I disagree with the argument that instruction cannot be considered a game if the training is formal.

The design to anchor the instruction in maps and photos provides authentic immersion, and the link to previous procedural knowledge provides an effective suspension of disbelief and allows learners to concentrate on the content rather than the novelty of the environment.The feedback mechanisms seem particularly effective: consequential outcomes show learners rather than tell them; and simplified demonstrations provide the realistic equivalent of “base case” worked-out problems. I’m curious about the 8 cognitive processes the authors claim can be tested using variations of a drag and drop exercise. And rather than adaptive testing, if the authors move the simulation toward a more game-like design, the introduction of the level-up concept will offer the same scaffolding effect.

Simulations – 7 Examples

Citation

Lunce, Les M. “Computer Simulations in Distance Education”. International Journal of Instructional Technology & Distance Learning 1.10 (October, 2004). Retrieved from the Web 02/06/2009 at http://www.itdl.org/Journal/Oct_04/index.htm.

Summary

The author defines a simulation as an immersive model of the real world that provide feedback from a flexible and dynamic system. An additional advantage of simulations is that they permit learners to experience problems “that would be too dangerous or expensive to explore in reality.” He acknowledges that simulations may oversimplify real-world complexity. The simulations were varied in design and subject matter (although liberal arts and social sciences were conspicuously absent) and contained elements of demonstration, discovery/exploration, practice/experimentation, hypothesis-testing, problem-solving, and coaching (the instructor component). Some simulations used teams while others were solitary.

The 7 simulation articles all reported positive student reactions but only one article measured student outcomes; that article showed higher scores and higher levels of knowledge on post-tests. 4 of the 7 simulation articles reported extensive instructor involvement to situate the simulation instruction; one article reported that use of the simulation without coaching often produced ineffective results, and that use of the the simulation “without first mastering the appropriate problem related skills” often produced null or even incorrect learning. One simulation study reported no greater degree of knowledge transfer than other methodologies.

Reflection

The author’s definition itself is simplistic and misses the primary distinguishing factor of simulations: working (in a virtual community) to solve a simplified but realistic problem. The major findings were equally predictable:

  • instructor guidance/coaching was critical
  • simulation complexity needs to build

The following comparison, an amalgamation from multiple articles,  or indeed any comparative framework should have informed the study; simulations, like games, are designed to teach (with or without direct coaching) students how to solve a general class of problems with a known solution.

Learning Type and Demonstration

Learning Type and Demonstration