W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2003

RE: The two models of accessibility

From: Nick Kew <nick@webthing.com>
Date: Wed, 2 Apr 2003 23:13:48 +0100 (BST)
To: Gerard Torenvliet <g_torenvliet@sympatico.ca>
cc: WAI Mailing List <w3c-wai-ig@w3.org>
Message-ID: <Pine.LNX.4.21.0304022249240.1442-100000@jarl.webthing.com>

On Wed, 2 Apr 2003, Gerard Torenvliet wrote:

> 
> I mostly agree with what Graham says here. However, since people continue to
> view 'accessibility' as meaning 'accessible to screen readers' and nothing
> else, we need to become a little more nuanced in our terminology.

I'd disagree strongly with that.  Accessibility means that contents
exist in an accessible form (most usually text) and are sufficiently 
well-structured to be broadly device-independent.

> The idea that I've been working on lately is that both accessibility and
> usability are relational terms. That is to say, a design is not accessible
> (or usable) in and of itself, but only with respect to some target user or
> groups of users. In other words, accessibility or usability are dependent on
> the person or group of persons against whom those terms are being evaluated.

That seems to state rather well the difference - and sometimes
conflice - between accessibility to the severely-retarded (championed by
one of our regulars) and of groups with physical disabilities.  I've
always thought this should be obvious, but judging by some of the
horribly patronising "accessible" versions out there, it clearly isn't
obvious to many (I recollect a few months back complaining to my
own government that their "accessible" examples implied that physically
disabled users should be presented with something that looked like
a picture-book for pre-school children).  If your article can help
educate the authors of those guidelines, that'll be great:-)


> Under this conception, usability and accessibility are actually orthogonal
> (but complementary).  Accessibility is a measure of how much of the data and
> functionality presented by an interface that a user with certain
> characteristics can access.

No, no no!

Accessibility implies that the data are there in the first place, and
are not irredeemably device-dependent.  Knowledge of users and interfaces
is orthogonal to that, and is the business of UI developers.

>	 Usability, on the other hand, is a measure of
> the ease with which that data and functionality can be accessed by the same
> user.

Accepted.

> Note that there are a few implications of this approach:
> 
> 1. A system can be designed that is highly usable to a target group of
> users. However, if that target group only includes those with a certain
> disability, it may not be very usable at all for those who do not have that
> disability.

That seems to me a fairly contrived argument.  OTOH, I guess peepo is a
live example of exactly that.

> 2. A system can be designed that is highly accessible to users with a broad
> set of disabilities, but that is not usable for that group. For example, any
> corporation can take their web site and ensure that it complies with the
> letter of Section 508. Because this is the case, with enough work, all of
> the information and functions should be accessible. However, this does not
> mean that they will be easy to use; in fact, to the contrary - experience
> has shown that to make a web site that is accessible and easy to use cannot
> easily be achieved by retrofitting an existing inaccessible site.

Agreed to a point, but that argument should probably be padded out with
the dangers of checking by rote and without due thought, and the spirit vs
the letter of accessibility.

I should add that mod_accessibility is designed to be a drop-in
solution that *can* usefully be retrofitted, and will make substantial
improvements both in accessibility and usability to many, though by
no means all, existing inaccessible sites.

> 3. Accessibility is a limiting criterion on usability. If data and
> information can't be accessed, it can't be used.

Yep.

-- 
Nick Kew
Received on Wednesday, 2 April 2003 17:13:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:14:09 GMT