Sunday, September 24, 2006

Science and the HCI Practitioner

I used to be a scientist. After my PhD work, I held research jobs at a couple of universities before joining Logica's R&D centre in Cambridge (the original one). And then I became a craftsman when I began commercial human-computer interaction work.

The thing about HCI is that there isn't much science behind it. In fact, even by the standards of the social sciences, what passes for research in HCI is barely credible. In fact, I'd go so far as to say that most studies in the field are so specific to the technologies and tasks and environments involved that they probably can't be generalised beyond that particular situation. Proper experimental design, let alone replication of a result, is incredibly rare.

Yet the field is full of rules and guidelines, do's and don'ts, as if there was a vast corpus of repeatable research behind them. (I know, I know. Some of you are shouting 'What about Fitt's Law?' and you're right. That is real, solid research. One piece. Count it. One.) So what's going on here? How do people like me get to tell our clients we're experts with a straight face?

Well, partly it's this. Proper research takes a very, very long time. By the time you've done a serious study and published it in a peer-reviewed journal, the chances are you can no longer even buy the display technologies, operating systems and input devices that were used in your experiments. The H in HCI may not have changed in 150,000 years but the C certainly has, and so has the I. And proper science is just no good at shooting at such rapidly-moving targets.

So we experts rely on three things; the experience we gain in working with people and interactions over years or decades, a set of 'rules of thumb' which we know are probably wrong in many situations, and an attitude to the problem that goes ‘the best thing to do is evaluate this with real users’.

That’s why I think of myself as a craftsman, rather than a scientist, or even a technologist – because my real skills are in my approach, my kitbag of techniques, my years of experience, and my attitude to the problem. I wish, sometimes, it could be different but, in this field, it can’t.

Thursday, September 21, 2006

Don't Touch That Dial

The first post on a new blog is always a funny thing. No-one knows you're there yet and, while you're bursting with things to say, you know you really ought to save them for a time when you're not so obviously talking to yourself.

So let me just say that I am a vastly experienced (i.e. old) human-computer interaction specialist - old enough to remember when the field was widely known as 'man-machine interface' (how times change) and experienced enough to know there is nothing new in the term 'interaction design' (been there, Dan that). In fact, one of the things I might get on to in the near future is listing out all the many terms used for what it is I do and, unless I get too bored with the thing, the scope and applicability of each of them.

On the whole though, I just want to dump 25 years' worth of learning on the world in case anyone wants it.

So, stay tuned. (Oh yes, and click on an ad, if you would, please, I'm running short of Kitty Kat.)