All entries for Wednesday 19 July 2006

July 19, 2006

The problem of transferable learning technology development – some guiding rules

Development of services/patterns-of-behaviour1 always works best when guided by committed users. But how can we be more assured that such developments will transfer to many other users? The more complex an organisation, the more difficult this is to answer. But there are some useful guidelines.

There is an approach to the development of software known as 'agile' development. Based upon years of painful experience, it priveliges the frequent delivery of quality new features, with constant review and improvement (where strictly necessary), over the master–plans and monoliths that are forever looming up on distant horizons. It is a sensible approach that should perhaps be applied to domains beyond IT. Another of its principles is that rather than trying to satisfy a large number of poorly engaged and un–committed users, it is better to work closely with a small number of closely involved and enthusiastic individuals. This again is a sensible option. Developments thrive on quality feedback, direct from the field, delivered first hand and with the necessary degree of detail.

There is, however, a potential trap in this approach. What if those committed enthusiasts turn out to be mere eccentrics? What if they are unrepresentative of the majority? Perhaps the majority over time will follow their lead? In which case they stop being eccentrics and start becoming early–adopters. Or alternatively, other people may never adopt the new development. In some cases, the correct course of action is to simply seek other enthusiasts, other more promising directions. When one is adding new features/possibilities to an already existing service/pattern–of–behaviours, such change of track is acceptable occasionally. But there is a cost in terms of waisted opportunity, resource, and in people's tolerance of the process. Choosing enthusiasts well is therefore important. When one is initiating entirely new services/patterns–of–behaviours, the problem of who to trust is even more critical. For example, Warwick Blogs was built upon the asumption that a relatively small number of people had a representative and reasonable view claiming that many other people would adopt it once it became available. The innovation seems right to these people. Added to that is some justified belief that it will also transfer beyond that minority, and that this transferability will justify the effort undertaken to get the new service to the point at which a wider range of people can become enthusiasts engaged directly in the development process.

Are then such assumptions of transferability safe? Perhaps sometimes. In some cases the target user population is distributed amongst a very simple ecology – that is to say, there are few differing niches, and similarly few adaptive responses to those niches. Everyone behaves in the same way. Everyone wants the same things. I suggest that the ecology of learning, teaching and research at Warwick is not such a simple ecology. There are in fact multivarious niches, each with a range of overlapping adaptive responses. People here are varied and unique. In fact, as I have argued elsewhere, higher education in Britain is full of such variation. Diversity is encouraged. The result being the demise of the monolithic VLE and the rise of the individually selected mesh of diverse features. Ask a Warwick lecturer in English to describe their ideal learning technology environment and the answer will be different to that of an Oxford lecturer in English, and no doubt different to another Warwick lecturer in English.

The practical result of such diversity is that when given a limited resource and the need to engage and satisfy a large number of people, choosing what services/patterns–of–bahaviour to focus upon is difficult. How is transferability achieved? How do we judge whether we are satisfying an isolated niche or feeding the whole ecology? Difficult, but perhaps there are some guiding rules.

Rule 1

If a smaller number of unrelated users from quite different niches all respond positively to a proposal, or better still come up with the same proposal themselves, then this most likely points to some aspect of the proposal that is transferable across the niches, that has utility and attraction in every case.

This is always more indicative of transferability than when a larger number of users from the same niche back a proposal.

Rule 2

If you want adoption of the proposal to spread more widely than these few diverse people, then a second rule is necessary.

It is quite possible, in an ecology like Warwick, for people to occupy similar or tightly–knit niches, but in significantly different ways. Take two lecturers in philosophy, look at how they prepare lectures, and it's likely that you will see some dramatic differences. Therefore the second rule is: identify the current behaviour of each of the interested users to see just how the proposal fits in, consider how common this is to other people in their vicinity, and ask whether that will make a difference to its adoption within their niche. Just how typical is the relevant behaviour of philosophy Lecturer A in comparison to other philosophy lecturers? Looking at this from another perspective: is there anything that blocks other philosophy lecturers from adopting the proposal that has been adopted by Lecturer A?

Rule 3

In some cases, rule 2 would seem to indicate that transferability is unlikely. Philosophy Lecturer A cannot transfer their new found technique to people in their immediate vicinity. It may however still be worthwhile to other people. In such cases we should seek or create lines of transferability that link up isolated people in separate localities, perhaps with other things in common. For example, the PhD student ePortfolios brings together individuals with isolated needs from many locations.

I am certain that there are other useful guidelines. Feedback would be most welcome.


1. services are always coupled with behaviour patterns, never think of them simply as assemblages of features, as those features are only meaningful when used.

How much do we know about how people really work in academia - researchers, students? How well are the diverse niches, adaptions and mal-adaptions really understood? And how much do we know about the match/mis-match between that diversity and learning technologies? I've recently heard that some people at Warwick are starting to think about these questions. I think that supporting this research is important.