>With the control group experiencing a 12% failure rate I think it's pretty
>clear there's a problem with how some of these sites are designed;  it's
>not just an AT issue. Web designers are still not using HTML properly and
>still not testing their sites for usability.

I agree, but would like to add another point to the discussion that has not
yet been mentioned.

The report seems to imply that all web usability issues fall into the "walk
up and use" or ad hoc model when it seems to me from my experiences, that
most of the web pages I encounter are ones that I visit repeatedly.  These
don't change very often and will, in many cases follow some kind of
predictable structure, at least within a given site or company.  So, the
other issue here is one of time on task or *learnability* as opposed to
just bottom-line usability.

Of course, the easier and more accessible a particular page is to navigate
when you first encounter it is important and should be the benchmark, but
many complex or difficult pages will eventually prove more efficient to use
as the user gains a level of familiarity with them.

It would be interesting to see what happens to the numbers when the same
subjects were asked to repeat the tasks a number of times with only minor
variable changes.

I would also like to mention that I do not necessarily think that hard
statistical data can accurately capture and quantify the concepts of
usability and accessibility.  They are simply one tool in the arsenal and
need to be weighted appropriately (most often they are given too much
credibility).

-- Jim