Electronic Performance Support Systems are a fabulous concept, as pioneered by Gloria Gery back in the early 90’s. The notion is that as you use a system, and have entries or decisions to make, there are tools available that can provide guidance: proactively, intelligently, and context-appropriate. Now, as I heard the complaint at the time, this would really be just good interface design, but the fact is that many times you have to retrofit assistance on top of a bad design for sad but understandable reasons.
The original were around desktop tasks, but the concept could easily be decoupled from the workplace via mobile devices. One of my favorite examples is the GPS system: the device knows where you are, and where you want to go (because you told it), and it gives you step by step guidance, even recalculating if you make a change. Everything from simple checklists to full adaptive help is possible, and I’ve led the design of such systems.
One of the ideas implicit in Gery’s vision, however, that I really don’t see, is the possibility of having the system not only assist you in performing, but also help you learn. She talked about the idea in her book on the subject, though without elaborating how that would happen, but her examples didn’t really show it and I haven’t seen it in practice in the years since. Yet the possibility is there.
I reckon it wouldn’t really take much. There is (or should be) a model guiding the decisions about what makes the right step, but that’s often hidden (in our learning as well). Making that model visible, and showing how it guides the support and recommendations that are made, could be made available as a ‘veneer’ over the system. It wouldn’t have to be visible, it could just be available at a click or as a preference for those who might want it.
Part of my vision of how to act in the world is to ‘learn out loud’. Well, I think our tools and products could be more explicit about the thinking that went into them, as well. Many years ago, in HyperCard, you could just use buttons and field, but you could open them up and get deeper into them, going from fixed links to coded responses. I have thought that a program or operating system could work similarly, having an initial appearance but capable of being explored and customized. We do this in the real world, choosing how much about something we want to learn (and I still want everyone who uses a word processor to learn about styles!) about something. Some things we pay someone else to do, other things we want to do ourselves. We learn about some parts of a program, and don’t know about others (it used to be joked that no one knows everything about Unix, I feel the same way about Microsoft Word).
We don’t do enough performance support as it is, but hopefully as we look into it, we consider the possible benefits of supporting the performance with some of the underlying thinking, and generating more comprehension with the associated benefits that brings. It’s good to reflect on learning, and seeing how thinking shapes performance both improves us and can improve our performance as well.
I have a few cool examples of simple performance support with just in time concept / procedure learning support. We’ve been focusing more than 50% of our development energy on the development of EPSS products for over a decade. We think it pays off and are expanding our models to include more formalized lesson.
Our EPSS products aren’t really fancy. They don’t have links to logistics systems. The wizards included in procedure support sections are relatively low-tech. Our audience targets include both novice and expert users. However, the model we use allows us to provide packaged support at a very reasonable cost (support topics are generally less expensive than an equivalent training product) and we provide the bonus service of validated tasks and capture of practice through the AP network.
I’ll see if I can get a few of these examples posted.