W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2002

Points to consider in writing success criteria

From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
Date: Sat, 26 Jan 2002 18:09:27 +1100
Message-ID: <15442.22055.815833.721161@jpc.local>
To: Web Content Guidelines <w3c-wai-gl@w3.org>
For each individual success criterion, consider the following:

1. Does it apply to all web content? In all circumstances? If not,
   appropriate limitations and qualifications should be included in
   the statement of the success criterion. If you are unable to
   specify these qualifications precisely, then perhaps it belongs in
   the "advisory" category rather than a success criterion that
   everyone implementing the checkpoint is required to meet.

2. Is it expressed clearly and concisely, in a style that is
   consistent with other parts of the guidelines document?

3. Does it introduce new technical terminology? If so, can this be
   avoided without making the success criterion vague, misleading or
   overly lengthy? If new terminology is being introduced, you should,
   if possible, provide a definition that can be included in the

4. Is it testable? In most cases, would informed observers be likely
   to agree on whether or not this success criterion has been
   implemented for different types of pages? 
   Are there specific tests, either manual or automated,
   that could be used to determine whether the success criterion has
   been implemented? If the success criterion is not testable then it
   belongs in the "advisory" section of the success criteria for this

Points to be considered in reviewing all of the success criteria
proposed for a particular checkpoint:

1. Are these success criteria sufficient? 
   If a content developer implemented all of these success 
   criteria, (and nothing else) would the checkpoint always be met? 
   If not, some amendment or addition to the success criteria is needed.

2. Regarding the testable success criteria: are all of these criteria
   necessary in order to satisfy the checkpoint? Specifically, is it
   necessary to implement all of these success criteria, taking into
   account any limitations or qualifications expressed therein, in
   order to meet the checkpoint?

3. Regarding the "advisory", non-testable, suggestions provided along
   with the success criteria: do these enhance accessibility? Can any
   of them be rewritten so as to be made precise and testable? If so,
   then they belong in the "testable success criteria" category.
Received on Saturday, 26 January 2002 02:09:35 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:40 UTC