Responses and Revisions -- Part 1

***1. In response to Eric Hansen's suggested guideline and checkpoints:
> Suggested Guideline: "If a written text is inaccessible due to the 
> difficulty of the language, provide an alternative representation or 
> version that is accessible.
> Checkpoints: 
> 1. Provide a less complex and less difficult textual version. (Priority 

Daniel Dardailler wrote:

"Note that some of the checkpoints you mention are valid regardless of
this guidelines, because they do relate to access to the information
(e.g. if lang=fr is not used when a voice agent read my last name, you
will get a false reading), but "Provide a less complex and less
difficult textual version" is just not feasable.""

Eric Hansen now writes:

I agree that feasibility is an important consideration.

I had at least two purposes in proposing this guidelines and its 
checkpoints: first, to heighten awareness of issues faced by people with 
language-related disabilities; second, to drive home the point that issues 
of cost (feasibility, practicality, cost-effectiveness) did, do, and should 
play a role in developing the guidelines.

I understand now that priorities (imperatives) are functionally determined 
by impacts, without consideration of other factors such as cost.

However, I have argued that the cost of implementing a solution did and 
should influence the decision of whether or not to include a checkpoint (or 
guideline) in the list. If the costs of implementing a solution seems too 
high to recommend to all Web authors, then it seems to me that the working 
group can assert that the solution must be excluded from the document for 
that reason. In fact, excluding the excessively expensive checkpoint is the 
ONLY thing that the working group CAN do. 

If the conceptual framework for the priorities were different there would 
be a different way out of this problem. You may recall that some while ago 
I advocated making clear that impact and imperative can vary independently. 
If they were allowed to vary independently, one could give an expensive but 
essential checkpoint a HIGH impact rating (e.g., PREVENTS access by people 
with cognitive disabilities) but a low priority rating (e.g., Web authors 
MAY do). But this is not permitted. The guidelines document can ONLY 
include checkpoints that are believed to be economically feasible for 
virtually all Web authors. Otherwise you find yourselves in the untenable 
situation of recommending a solution that you believe is NOT 

At this point, I don't recommend changing the way in which the priority 
(imperative) ratings are produced. I just recommend making explicit the 
grounds for excluding any given checkpoint. "Excessive cost" or "lack 
cost-effectiveness" can and must be justifiable reasons for excluding a 
checkpoint or guideline from the document.

***4. Daniel Dardailler wrote:

"To give you a direct example, I bet some readers of this list will
find your most recent posting inaccessible, semantics wise, because of
its length. Are you willing to rewrite it so that it fits on a single
page (e.g. less than 40 lines) ?"

Eric Hansen now writes:

I apologize for the length. 

If it would be helpful to the working group, I would be willing to try to 
reduce the length.

I know that writing or rewriting good text can be hard labor. Perhaps 
rewriting Web text may only be warranted in Web sites for specific purposes 
or audiences. But as noted in my material in Part 2, the checkpoints and 
the guideline document as a whole is addressed to general audiences 
(especially individuals with disabilities) across a wide spectrum of 
purposes, and therefore may not be sensitive to every specific audience or 
content purpose.

Eric G. Hansen, Ph.D.
Development Scientist
Educational Testing Service
ETS 12-R
Rosedale Road
Princeton, NJ 08541
(W) 609-734-5615
(Fax) 609-734-1090

Received on Wednesday, 13 January 1999 12:48:52 UTC