Important note: This Wiki page is edited by participants of the RDWG. It does not necessarily represent consensus and it may have incorrect information or information that is not supported by other Working Group participants, WAI, or W3C. It may also have some very useful information.
Web Page Sampling
From Research and Development Working Group Wiki
To evaluate the accessibility of websites, it is usually necessary to select "representative samples" of individual web pages. What are the current approaches and best practices to increase the level of inter-rater consistency?
Page author(s): Shadi Abou-Zahra Other contact(s): Chaals
Being able to benchmark various aspects of the Web, prioritising research and development, and many other activities in the scope of the group require the ability to get a "representative sample" of the Web. Determining what is a representative sample, and how to get one, is a complex question.
[Additional background on the subject matter]
- Is it reasonable to use different samples, such as "the front page of the Alexa top 500", or should a fixed collection be used to ensure consistency of results?
- Copyright and access
- What are the legalities of archiving content as a sample? What is the best approach to sampling content that is not openly available, such as the password-protected sections of a web-based banking application, or a corporate intranet?
- Variable or dynamic pages
- Sampling techniques run the risk of collecting a version of a given page that is not the same as a real, typical, or specific user would be served.
[References to related works and activities]
- Brajnik, Mulas, Pitton. Effects of sampling methods on web accessibility evaluations. ASSETS 2007 
Back to the list of topics. Bold text