I really like interactive scenarios, particularly the branching sort which require learners to make decisions in response to an evolving situation or to engage in a dialogue with an on-screen character. Designed well, they provide the learner with the opportunity to engage in a sort of experiment from which insights can emerge into key principles. Although it’s great when these scenarios can unfold as a series of video scenes or as 3D graphics, I’m not too unhappy if the stimulus for each decision is a simple photo or illustration.
So what’s the problem? As I rediscovered this week as I tied my brain in knots constructing a series of branching scenarios for learning designers, the weakness with current scenarios is with the interface.
Unless I’ve missed some major technological breakthrough, all scenarios currently require the learner to interact by choosing from a series of options presented as text. And, as we know from TV quizzes, it’s much easier to answer questions when the options are presented to us (think Who Wants To Be A Millionaire) than when we have to construct them for ourselves (think Mastermind, at least in the UK).
Now scenarios aren’t quizzes. They’re testing the ability to make judgements rather than simple knowledge retrieval, but the effect is similar. When you see a series of options, it’s very hard not to be tempted by the one that looks like the most sensible, even if you’d never have thought of that for yourself. As a result, it’s all too easy to sail smoothly through the scenario, looking cleverer than you really are.
Back in the mid 1980s, Donald Clarke and I worked on a whopping series of video-based scenarios for British Telecom. Each scenario depicted a complete selection interview. The branching was cleverly constructed with the aid of a programmer who was a bit of an AI expert. We tried our best to make the interface more fluid by having the learner construct sentences from a series of phrases. The result was compelling but still not like participating in a real interview. At that time, we would have been amazed to believe that, nearly 30 years later, we would not be using a natural language interface for projects like this. So why is that? Is this ever going to be feasible?
So what’s the answer? First off, I’d suggest that, imperfect as they are, scenarios still provide a valuable opportunity for experimentation and rehearsal, safe from physical, financial or psychological risk. But they’re only a step along the line. Next step is probably the role play, which is all too often an unsatisfactory experience or, if funds permit, a full-scale simulation. But in most circumstances, there is a more practical and logical way to continue the process of learning beyond the scenario and that’s through real-life experience, ideally supported through coaching and/or social learning.
In other words a blend.