Tuesday, 19 August 2014

What should a simulation simulate?

Some of the lecturers in my faculty have become interested in simulations as a way of bringing more practical experience into their learning materials. As I prepare to advise them on the possibilities and pitfalls, I'm remembering the best piece of advice I ever received on simulation design.

Computer simulations for professional practice have been around almost as long as I have, especially in business and management where the funding for high-specification products is more forthcoming; I can remember my school contemporaries in the 1970s playing a computer-based business simulation against teams from other schools. They sent their decisions off by post, and received back lineprinter output with tables of figures, showing how all the teams' simultaneous decisions affected their shared market. By contrast the management simulation which I saw demonstrated at a conference in 2001 was stupendously impressive in its production values: graphical realisation of the business premises, video inserts of actors playing the various characters, a sophisticated "co-pilot" or animated mentor to give you in-simulation access to all the business management concepts and theories you might need. Truly, an embodiment of the notion of theory-in-practice.

As I watched the demonstration, I wasn't surprised to learn that the simulation had been the brainchild of a former McKinsey management consultant, nor that the cost of the simulation licence was £750 per single copy. I tried to estimate what the development and production cost must have been, and ran out of zeros in my mental calculator. I was working at the time for LMD Learning Solutions, where most of our clients were public sector organisations, and I knew that though they would drool over the prospect of such a product there was no way that any of them would be able to afford to develop one.

I'd also done enough game design (adventure games) to know the basic problem which you quickly encounter: you can't simulate everything. The idea of a simulation - a virtual world, in which you can lose yourself, like the holodeck on Star Trek: The Next Generation - is very attractive. But outside of science fiction, I knew, you have to be selective: you can't create every detail of the environment a user might choose to inspect, or program a response for every action they might think of doing. So I realised what my question to the speaker from the simulation company needed to be: not "How do you produce this kind of thing on a budget?" - although I did want to know that - but the more important and deeper question "How do you decide what things to simulate?"

Her answer was so blindingly obvious in retrospect that I felt embarrassed at not having thought of it myself. You simulate those things which a learner is most likely to get wrong. There's no point in spending effort and resources in simulating the things which they can do without any trouble; you need to simulate the possibility for learners making the kind of mistakes which they're likely to make in real life, so that they can make them safely and learn from them without penalty.

If you want to theorise this, I'd say that a teaching simulation needs four components: Situation, or the practical setting and challenges in which the learner is going to find themselves; Action, or the possibility for the learner to do things in the setting; Feedback, or the results of those actions, whether immediate or long-term, made evident to the learner; and Reflection, or some kind of debriefing, to turn the simulation experience into learning.

Of these, Situation is the most immediately attractive and the one on which salespeople for simulations tend to focus - as in the Imparta video above. It speaks to the fantasy or game-like aspect of a simulation: you can be a management consultant to a blue-chip multi-national corporation; you can be a newly-qualified social worker thrown into the deep-end of a potential child abuse case; you can be a starship captain leading a covert mission in the Romulan neutral zone; and so forth. But if we're designing simulations which are going to be learning and not just fun - which isn't to say that they can't or shouldn't be fun, just that we're not going to spend the time and resources unless they make a pretty good contribution towards learning outcomes - then we need to plan also for Action, Feedback and Reflection.

(As a contrast to the Imparta video, see this simulation, about which I've already blogged, in which you play a scientific expert on a commission investigating a high-profile accident. Technically, it's very simple indeed; there's no video, no graphics, just very-well-written text, on which I suspect the authors spent a good many hours. But it's very thought-provoking, which surely needs to be one of the characteristics of a simulation at university-level.)

No comments:

Post a Comment