Out-Law News | 12 Jan 2006 | 3:05 pm | 3 min. read
The research was conducted by Professor Helen Petrie of accessibility and usability firm Designed for All. Her team studied 500 websites: 250 UK or UK-oriented e-commerce sites and 250 financial sites based in the UK, Australia, Canada, Ireland, New Zealand and the US.
Of the 500 sites, 40 (8%) had an accessibility statement or logo. However, when 20 of these sites were inspected further, only six were found to be accurately stating their accessibility.
Professor Petrie does not suggest that website owners are deliberately misleading people; but she highlighted some of the risks for those who get their statements wrong.
“A company’s accessibility statement is a reflection of its values towards disabled people," she said. "People’s trust will be affected if a company makes a public statement that is not reflected in how it actually behaves.”
Professor Petrie previously led a team that worked with the Disability Rights Commission (DRC) to produce Britain's most comprehensive report to date on the state of web accessibility, published in April 2004.
That report confirmed what accessibility professionals already knew: most sites display woeful levels of accessibility. Of 1,000 sites tested, 81% failed on automated testing to reach Level A of the best known guidelines, the W3C Web Accessibility Initiative's Web Content Accessibility Guidelines, known as WCAG Version 1.0.
More important than the statistics in the DRC report was the impact: for the first time it brought the legal, commercial and ethical arguments for web accessibility to the attention of the mainstream media and board rooms.
The exaggeration of accessibility statements is a known problem. In March 2004, one month before the DRC's landmark report, web testing specialist SciVisum found that 40% of a sample of more than 100 UK sites claiming to be accessible did not meet the WAI checkpoints for which they claimed compliance.
OUT-LAW put the DRC report and SciVisum's research to Judy Brewer, the W3C's Web Accessibility Initiative Domain Leader, in April 2004. She told OUT-LAW that "over-claiming a site's accessibility by as much as a-level-and-a-half is not uncommon."
Brewer acknowledged a problem with the WCAG checkpoints: the way WCAG Version 1.0 is written, it can sometimes be difficult to tell whether various checkpoints are satisfied.
At the time, she was working on WCAG Version 2.0 which promises to make it easier for web developers to know that they have met the guidelines. Version 2.0, Brewer explained, should be more precisely testable. (The final version of Version 2.0 is still awaited.)
But interpreting accessibility guidelines is not the only problem.
Léonie Watson, chair of the Association of Accessibility Professionals (the AAP was formerly called the Usability and Accessibility Working Group) said: "I believe the problem is that there is a great deal of focus given to attaining accessibility targets at the time of launch, but that very little is done to try and sustain those achievements as the site evolves."
Watson continued: "An ongoing quality assurance process is needed to ensure that the level of accessibility is maintained as new content is added and the site expands. The fact that such solutions are rarely implemented makes it very easy for accessibility statements to fall out of date and become inaccurate."
Another problem is that some developers think they know more about accessibility than they really do.
He added: "In my opinion, the vast majority of website developers do not understand the reasoning behind most of the WCAG criteria, don't know how to implement them appropriately and wouldn't know how to test whether they met them."
Suppliers who exaggerate their skills, even in good faith, make the commissioning of accessible sites very difficult for non-experts. To address this, the AAP is developing an accreditation scheme for suppliers. The scheme will make it easier for those commissioning sites to identify qualified suppliers who will have undergone a peer review to carry the accreditation.