Re: Usability Policies - Think Bigger!

> (Re-send with no v-card)

Do we really need Usability policies?  Isn't it the usability police and gurus that gets us all a bad name?

Effective usability processes are impossible without a clear understanding of the value that a system is intended to deliver (plug for my CHI 2004 short paper).  We first moved towards this in our work on structured usability problem extraction when we had to find a way to separate relevant user difficulties from irrelevant/trivial ones (some call this severity rating, but that misses the point as severity depends on impact, and not on the 'strength' of the problem).  So, without a clear view of how quality in use will enhance or destroy product value, one cannot objectively apply severity ratings (and we know how much individual difference there can be in severity ratings, especially when based on naff universal scales).

The same applies to fitness for purpose.  (Poor) fit to usage context can (destroy) enhance product value, but again the impact is not an attribute of the misfit.  It is an attribute of the *consequences* of the misfit during usage.  All problems lie in interaction, not in the product.

Quality in use and fit to context are our fundamental yardsticks in HCI.  However, we cannot objectively or effectively put measures and analyses here to good use without a clear view of intended product value and its implications for quality in use and fit to context.  It is these implications, and not the crusading zeal of the usability police, that tells us how slow is too slow, how much is too much, and how poor is too poor.  Jared Spool has been hinting at a similar line of analysis when he claims that the usability groups with the best performance/credibility/viability are those who clearly address business needs.  Although the original poster works in support for university research and teaching, the same is true. Without a clear view of the value that his on-line resources are intended to deliver, effective *consensus* on usability policies is impossible.

So, in my view, usability policies can only point your car at a brick wall.  If you drive slowly, it will take a while to hit it, but if you move at all, you'll hit it eventually.  The wall is made up of incredulity, mistrust and hostility from colleagues in development and management as they (sometimes rightly) write off reported usability problems as irrelevant and trivial (perhaps tarring the usability specialist with the same brush as we say up North).

What is needed for all product development, whether commercial, public, social, communal or otherwise, is a clear statement of intended product value that can be used to systematically derive targets for quality in use and fit to context.  This implies that measures will be made and analyses of fit will be carried out.  Policies are needed that allow the necessary measurements to be made, and information on usage contexts must be collected and retained to support analysis of fit. At this point we are back into mainstream HCI and (as long as the measures or contextual data are well understood) we know what we are doing (there are even ISO standards at this point).  However, don't start here with your policies.  Put them in a broader context of product value and people will see the need to testing and analysis, and the need to provide the resources to do this properly.

Usability is dead, long live product value ;-)

Cheers
Gilbert

    --------------------------------------------------------------
     POSTINGS (in plain text): [email protected]
           SUBSCRIPTION CHANGES:  http://lists.syntagm.co.uk
            (or send email to mailto:[email protected])
    --------------------------------------------------------------