[Sigia-l] Usability Testing comments from Giga

[Daniel R Drop]

>Joe wrote:
>> You might be interested in a recent article from
>> Giga Research. They recently opined on the issue. In short, they
>> feel heuristic/expert review and guideline following are to be
>> followed, rather than the collection of observed data
>[snip summary of article]
>> Thoughts?
>
>Of course: that's what Giga specializes in (see their Web Site Score
Card
>at: http://www.gigaweb.com/mktg/wssc/)! They would be shooting
themselves in
>the foot if they said (in)formal usability testing yields better results
>than heuristics-based expert reviews.
>
>Peter
>--

This discussion is resurfacing a related question I have about
information architecture.  Will the field of information architecture get
to a point where we have developed sufficient theory with supporting
practical frameworks that usability testing will becoming less important?
 

We often compare IA with building architects.  I am under the impression
that building architects do not have usability testing for their
structures. Surely there are building architects who design horribly
unusable spaces. On the other hand, there are very experienced building
architects who can design pleasant, usable spaces without having to
resort to usability testing. Can our profession get there?

My questioning of whether usability studies will always be necessary was
originally prompted by thoughts from Edward Tufte and Andrew Dillon. As
was mentioned earlier in this list, Tufte is already a critic of
usability testing, stating that with good design testing isn't necessary.
 Dillon recently in the article "Making Progress in a Complicated World,"
the Feb/Mar issue of ASIS&T Bulletin remarked "I would hate to be at
ASIST in 2020 and have to hear again that user studies and
interdisciplinary perspectives are important parts of developing
information systems." 

I generally agree with Lou Rosenfeld's response to why usability studies
are still required today.  According to Dillon's article, Rosenfeld
"remarked the world has become more complicated and the design process
more difficult." To which I might add that we currently don't have a
sufficient theory and framework to deal with these complications without
usability testing.

This train of thought leads me to ask whether there is a level of
complication with which our current IA theories and frameworks can deal
without resorting to usability testing. Perhaps these are the
brochure-ware web sites, the sites IAs generally are not involved in.

Another thought: Maybe when the complication is under control, IAs move
on to more complicated areas leaving the less complicated areas to
non-IAs. Since the IAs tend to the complicated information spaces,
practicing IAs always will require usability testing.

-- Daniel Drop