Re: Testing web page accessibility by phone

At 08:52 AM 2002-05-29, David Poehlman wrote:

>I cannot talk on the phone and test a web page at the same time.

Is this an intrinsic limitation on the tasks or an infrastructure problem?

For a remote lecture that Jennifer Sutton and I gave to Dan Andresen's class at Kansas State[1], we lashed up a collaboration infrastructure where the remote class had her computer screen projected and had Jennifer, me, and Perfect Paul all on the phone to the lecture room together.

This was all done with commodity technology, so long as you count a cable modem as commodity technology.  To make Net Meeting work fast enough for a lecture/demonstration situation, modem speeds were not enough.  And the speech synthesis had to be done in an external hardware synthesizer.  But it was doable.

The infrastructure that one needs to do this sort of remote collaboration for purposes of usability testing is a subset of what is needed for disability access to virtual meetings.  The people with disabilities have to have enough access to what is happening so that they can function as presenter, recorder, or chair[2].

So the infrastructure issues _have to be_ worked.

As usual, to get it running today takes some special tricks of the trade.  But we're getting closer[3].

Scott:

It may be a little premature to define a standard test protocol for this sort of testing, but perhaps defining an infrastructure kit and doing some usability testing of the kit with multiple users would get us a notch up the ladder.  This would serve usefully to give constraints on experimental protocols that researchers planned to use in this remote mode.  Check with the folks at ATRC in Toronto at least to see what they do with their scattered-site testers.

I don't think the answer is known well enough for you to get an immediate answer here, but if you 

a) put a working body into the Authoring Tools Working Group where they are working on automation of evaluation techniques techniques, and 
b) same for the Evaluation and Repair Working Group for how to blend the usability assertions with the automation results from the rule-checking tools via EARL, 
c) do some actual usability studies in this mode,

..you could come up with a solid contribution to the accessibility knowledge base.

Al

[1] A dip into accessibility on the WWW (K State lecture notes)
 http://www.cis.ksu.edu/~dan/cis726/web/lecture_notes/aDip.html

[2] Notes for planning purposes for the Advanced Collaborative Environments Working Group in the Global Grid Forum
 http://www-unix.gridforum.org/mail_archive/ace-grid/msg00026.html

[3] Modality Translation Services Program
 http://trace.wisc.edu/world/modtrans/


>the kind of validation I am referring to here is inclusive and begins
>with what is usually thought of as validation.  It goes like this.
>Write your pages to speck, validate the pages to make sure they are
>written to speck and while you are writing to speck, include in the
>speck known factors that make pages usable/accessible and when you
>validate the final product, use tools that will check for known
>usability and accessibility factors and also that will allow you to
>examine other issues that may be relevant.
>
>People are important in the mix too but it all starts with a well laid
>plan and when the users finally test the pages, they should find
>insugnificant things that you may have missed or hadn't thought of but
>that wound not necessarily break the accessibility/usability of the
>pages.  I cannot talk on the phone and test a web page at the same time.
>
>----- Original Message -----
>From: "Liz Fuller" <liz.fuller@verizon.com>
>To: <w3c-wai-ig@w3.org>
>Sent: Wednesday, May 29, 2002 8:42 AM
>Subject: RE: Testing web page accessibility by phone
>
>
>
>Validation is all fine and good but it would be quite simple to have a
>site
>that validates and passes all the standards tests but when a user sits
>in
>front of it they can't use it at all.  Perhaps what Scott is testing in
>his
>phone sessions is not accessibility per se but usability by people who
>are
>blind.  This is an extremely valuable activity that will provide way
>more
>insight into how people will access your site than a simple validation
>activity.  Both automated validation testing and real live user testing
>are
>important steps in good site design.
>
>Liz
>
>
>-----Original Message-----
>From: w3c-wai-ig-request@w3.org [mailto:w3c-wai-ig-request@w3.org]On
>Behalf
>Of David Poehlman
>Sent: Wednesday, May 29, 2002 8:03 AM
>To: Jerry Weichbrodt; w3c-wai-ig@w3.org
>Subject: Re: Testing web page accessibility by phone
>
>
>interesting point Jerry!
>
>This harkens back to my point of "vallidation" with which I was putting
>forward the crazy notion that doing it right to begin with makes a huge
>difference.
>
>----- Original Message -----
>From: "Jerry Weichbrodt" <gerald.g.weichbrodt@ived.gm.com>
>To: <w3c-wai-ig@w3.org>
>Sent: Wednesday, May 29, 2002 7:57 AM
>Subject: Re: Testing web page accessibility by phone
>
>
>
>I wonder how much of this depends on the user agent, by which I mean the
>combination of the web browser and screen reader.  As a specific
>example,
>ever since JAWS for Windows discovered how to grab headings (*real*
>headings
>with numbers, not just over-sized type that web designers like to
>pretend
>are headings) and present them in a list with the ability to move
>immediately to a heading, I have suddenly become very enthusiastic about
>proper use of headings to mark off important section divisions in a web
>page.  Prior to the heading support, I have to confess that headings
>didn't
>do much for me because they were essentially indistinguishable from
>other
>stuff on the page.  It makes a big difference, to me anyway, if you can
>gain
>some sort of hierarchical view of a web page rather than just the
>classic
>never-ending linear version.
>
>In summary, how a web page "stacks up" may be surprisingly dependent on
>the
>browser/screen reader used to view it.  Just my two cents, and hopefully
>not
>too far off the mark.
>
>Regards,
>Jerry
>
>----- Original Message -----
>From: "phoenixl" <phoenixl@sonic.net>
>To: <phoenixl@sonic.net>; <poehlman1@comcast.net>; <w3c-wai-ig@w3.org>
>Sent: Tuesday, May 28, 2002 9:13 PM
>Subject: Re: Testing web page accessibility by phone
>
>
>> Hi,
>>
>> First, the goal isn't necessarily validation but looking at
>> accessibility.  Second, different people have different
>interpretations
>> of what an accessible web page is to blind people.  For example, look
>at
>> the various standards that have been or are being developed.
>>
>> The methodology being used was to get subjective experience of various
>> blind subjects.  Rather than taking the perspective that the web pages
>> are accessible because they meet some set of standards, we focused on
>> whether the blind subjects themselves experienced the web pages as
>being
>> accessible.  The feedback that was given was interesting and helpful.
>>
>> Using a comparison strategy can also be helpful, but not always
>> necessary for getting useful information.  The issue of skill is
>> something to consider.  A question though is how much is it a blind
>> person's responsibility to have a certain skill level and how much is
>it
>> the web page's responsibility not to have high expectations for skill
>> level?
>>
>> The Hisoftware person I talked with left me with the impression that
>the
>> software doesn't have mechanisms for measuring such things as how long
>> does it take for the blind subject to understand a web page or
>> determining how accurately the blind subject understands the web page.
>> The software basically is checking syntax against a specified set of
>> standards rather than evaluating the experience of the subjects.
>>
>> Rather than discussing the questions, it might be interesting first to
>> use them when working with a variety of blind subjects.  However, a
>> question to ask is if a sighted person can understand the purpose of a
>> web page in let's say 15 seconds and it often takes a blind person 2
>> minutes to understand the same web page, is that web page accessible?
>>
>> Scott

Received on Wednesday, 29 May 2002 10:09:02 UTC