As mentioned in previous posts, I recently presented a paper at CHI 2011 (available online for free here). At both the academic session itself, and through bumping into people afterwards, one question kept coming up. It is a question that I have often fielded whenever talking about educational games, but which I did not address in the CHI paper (there’s only so much you can fit in a conference paper!) The question is generally along the lines of how to ensure that what people learn while playing a game “transfers” to the real world. I suggest that being in a situation where you are thinking about how to make sure knowledge or skills “transfer” means that you have already made a key games design mistake. You have forgotten why games are interesting – and what they are good at.
In order to explain this position I will refer you first to some research. In my opinion the very best existing example of experimental research on educational game design was carried out by Jacob Habgood at Nottingham University (he’s now head of Serious Games at Sumo Digital). I would recommend interested people to have a look at his thesis or this short paper. Habgood was interested in the relation of the content to be learned and the game play mechanics in the design of educational games. In two studies, he demonstrated experimentally that a game in which the play and learning were not merely placed side by side, but were intrinsically linked, were motivationally and educationally more effective than an almost identical game in which learning was not intrinsic to game play.
In my discussion of how Applied Behavioural Analysis can help guide educational game design, I discuss some very similar points:
”behaviour analysts insist that the “behaviour chosen …. must be the behaviour in need of improvement, not a similar behaviour that serves as a proxy for the behaviour of interest, or the subjects verbal description of the behaviour” (Cooper, Heron & Heward, 2006, p.16). So, whether or not participants have reached a learning outcome should not be judged by their answers to a questionnaire (unless the learning outcome is to improve the learners ability to answer questionnaires). Rather, whether they have reached that learning outcome or not should be observable from the action of the player as they are playing.
Essentially, instruction should not be designed as blocks of play and tests. The play, itself, should be the behaviour that is examined. This is especially important in a computer game, where a teacher is not present to interpret behaviour. So, if the game has been designed to teach industrial chemistry, and a learning outcome is an understanding of the chemical reactions necessary for the extraction of iron from iron ore, the game should take place at the molecular level. It is not appropriate to create a game where people can throw random chemicals into a furnace, then observe the outcomes and answer a multiple choice questionnaire.”
So, why is the question of how to ensure that skills and knowledge learned in a game transfer to the real world the wrong question? Because, if we ensure that educational games are designed so that the skill learned in the game IS the skill we want them to learn – not a proxy thereof – there is no question of transfer. If the game should teach multiplication, then it should involve multiplication as a core mechanic of the game – as demonstrated by Habgood. Multiplication is multiplication!
All of this comes with a caveat – games are not the ideal way to teach everything. Games are interesting, complex systems that allow interaction and trial and error learning. If the thing you want to teach is not an interesting, complex system that allows interaction and trial and error learning – you probably need to find a more appropriate teaching method.
Problems with “transfer” commonly come about due to either not ensuring that the learning is intrinsic to game play, or else because you are teaching something that is not the kind of thing that games are good at teaching.