[Irl-dean] United Nations Global Audit of Web Accessibility

[Paul Walsh, Segala]

Ok, this is pretty long but I've deleted parts of the conversation which I
don't feel were necessary. I'm pretty certain that any deletions I've made
won't change the tone or meaning. This conversation took place on an open
email list that includes a few thousand subscribers with an interest in
usability and accessibility. I hope it makes sense.

The following text was from a reader expressing their concern about lack of
coverage during the test.

And I think it only studied the homepages.  It should have had the decency
to check pages 'one level down' from homepages, this may provide different
answers.

Nomensa's response
When we've tested various pages, the homepage tends to be the best one, as
it's the one with the most attention, both internally and externally.
(There is sometimes an exception when people include unusual content just on
the homepage, but that's rare.) Again, if you are aiming to get good
results, you are best off checking the homepage.

Paul's response
I didn't realise that only the homepages were checked, that's not an
accurate reflection in my humble opinion. Furthermore, who's to say which
page is the homepage? If you chose the top sites, they're likely to be quite
big, which means they're likely to have a number of homepages as determined
by the content and audience. Testing one page on a site is unacceptable in
my opinion, especially if it is the basis for an industry report. 

The following text is Paul's original response to the original email
soliciting feedback about the report.

Yet another report telling us the same thing. Why doesn't someone come up
with a 'positive' report to highlight the companies that are making the
necessary adjustment? 

This negative type of approach worked in the past and it will probably work
in countries where Web accessibility isn't profiled as well as it is in the
UK, but let us start to create some positive vibes.

Nomensa's response
I would agree, but there are two problems with that
1 Reporting. We didn't write the headlines, and wouldn't have gone with "97%
bad, but it is that type of thing that actually makes it more likely to be
reported at all.
2 Shouldn't a sample of top/popular websites be the way to find good Sites

Paul's response
In short, no not necessarily. There are organisations which are dedicated to
accessibility but for technical and/or costing reasons, it may take up to 2
years to make all the necessary adjustments. They shouldn't have to wait
that long before being able to demonstrate their commitment publicly. Yet
they live in fear of being highlighted in negative blogs which appear to be
home to what I call accessibility extremists (or activist), who have nothing
better to do than highlight the inadequacies of others whilst their own
sites don't live up to the standards they express. 

What makes me laugh is that they test one or two things to base their entire
judgement which is technically incorrect. 

I advise anyone who wants to conduct preliminary reviews to follow the W3C
recommended approach - http://www.w3.org/WAI/eval/preliminary.html  

I don't think Nomensa fits in this category as I have a lot of time for your
company and I really like Simon. I'm simply emphasising that some people
have had enough negative connotations associated with accessibility.

Nomensa's response
Agreed, but it's tricky to do. If you make fuss about an accessible website,
there will be someone who tries to shoot it down, no matter how minor the
issue. 

Paul's response
This is the point I'm making. Let the extremists continue and they'll soon
crawl back into their caves.

Nomensa's response
It is partly for this reason (and partly because we can't make assumptions
about development or usage) that we have to be fairly strict when doing
audits. For more
http://alastairc.ac/2006/12/reporting-on-accessibility-issues/

Paul's response
You didn't conduct an audit for your report, you conducted what I call an
incredible quick check on a page. Please don't call it an audit or you'll
give the wrong impression and the companies that come out on top will be
given a false sense of security and the ones who fail, may in fact be much
more accessible than you think.

For example, we audit and certify new sites, page refreshes and landing
pages for O2 every week - yet, o2.co.uk and many other pages don't carry the
trustmark as they haven't been adjusted/tested. 

Nomensa's response
It wasn't quick, and I think 'audit' is the best term. We checked each WCAG
1 checkpoint for each page, mostly manual checks. To bring this down to
man-weeks rather than months, we kept it to the homepage (customer facing
where possible).

Paul.
I didn't think it was necessary to continue the conversation after
expressing my view, so I left it here.

So, if you've reached this far you're probably about to head for the coffee
machine to inject it directly into your veins!

Kind regards,
Paul

-----Original Message-----
From: irl-dean-admin at list.eeng.dcu.ie
[mailto:irl-dean-admin at list.eeng.dcu.ie] On Behalf Of Barry McMullin
Sent: 16 January 2007 11:06
To: irl-dean at list.eeng.dcu.ie
Subject: [Irl-dean] United Nations Global Audit of Web Accessibility

Hi All -

Just to say that the *full* report of the Nomensa study of web
accessibility, on behalf of the UN, is now available (sort
of - read on!). The original announcement of the study is here:

<http://tinyurl.com/y4fhm9>

As of this morning, that still refers only to the "Executive
Summary" being available, and that only "on request".  However, I
had already logged a request for the full thing (by email), and
was advised by email, this morning, that it was available. The
email included a long and complex URL to access the report (pdf
format only, 2.5 MByte download). They did not say anything about
not sharing this URL, so I presume it is OK to pass it on.  I
have made a short version of it, as follows:

<http://tinyurl.com/ykeds8>

On the other hand ... I suspect that the URL they gave me (which
the one above redirects to) may still be a limited time/limited
use sort of thing.  So feel free to try it, but it may well not
work.  If so, the best idea is probably to request your own copy
by email to <info at nomensa.com>. (Actually, they didn't say
anything explicitly about my not sharing the full report either;
but it is marked as copyright by Nomensa, so better not...)

I've had a quick look.  My main interest is in the
methodology. So, some *very* immediate responses:

- They did not rely on pure automated evaluation, which is
  definitely a plus.

- They only looked at the home page of each target site - which
  is a severe limitation.

- Not surprisingly, they found, as usual, that the elementary
  step of providing appropriate alternatives for images is still
  tripping up many many sites.

- They give some examples of what they think is "good practice"
  for text alternatives for images (page 9).  I have to say I
  don't really agree: the examples are all from the UK prime
  minister's site, and in all three cases which they show I would
  classify the image as purely visual decoration and give it an
  empty alt.  I'd be interested in any other opinions!

Anyway, that's as far as I have got - enjoy!

Best - Barry.

_______________________________________________
Irl-dean mailing list
Irl-dean at list.eeng.dcu.ie
http://list.eeng.dcu.ie/mailman/listinfo/irl-dean

-- 
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.432 / Virus Database: 268.16.12/631 - Release Date: 16/01/2007
08:25