Verification and validation of accessibility

In an appendix to the Guidelines, various techniques for validating and
verifying the accessibility of documents are proposed. These range from
the use of validation tools, to displaying the document with user agents
that have different capabilities. Should these techniques be prioritised?
The existing list seems rather lengthy and, while I would not suggest
omitting any of the items, it would seem reasonable to provide content
developers with guidance as to which approaches have proved most effective
practice as means of detecting coding and design shortcomings that hinder
access. However, it can also be argued that quick solutions such as
evaluation tools should not be used as substitutes for the longer but also
more revealing process of examining different renderings of documents by
various user agents, under diverse conditions. This point suggests that an
attempt to prioritise the evaluation and verification techniques would
lead to a contest between (1) speed and efficiency of application through
validation tools; and (2) comprehensiveness of assessment, via more
reliable methods of human checking.

Should we simply leave these considerations to the discretion of content
developers as in the Proposed Recommendation, or seek to classify the
verification techniques, perhaps even indicating which should be tried
first, or which should receive first priority?

Received on Tuesday, 13 April 1999 19:13:11 UTC