digital media access group
...excellent accessibility research and consultancy
This is a archived version of the DMAG website, but the information remains for reference. Please visit the new website for updated information.
Evaluating the Usability of Online Accessibility Information
By David Sloan, published August 30th 2006
We received funding from Techdis to investigate how useful accessibility pages provided on a selection of UK Web sites were to users - in particular older people without significant impairments, but who may soon benefit from accessibility information. We acted because we believe that too many Web accessibility statements are overly technical, and don't give the necessary support to the people who would most benefit from it.
You can read the full research report (c. 40 pages):
- Usability of Web Accessibility Statements: Final Report (Microsoft Word version)
- Usability of Web Accessibility Statements: Final Report (PDF version)
The work included an expert review of a sample of accessiblity pages taken from UK Web sites, followed by evaluation of the usability of the pages with some participants (all over 60, but with no significant impairments). From this work, here are the key findings and recommendations for Web designers and for researchers:
Key observations - Expert Review
Our expert review revealed the following:
- Accessibility information is often hard to find
- Sites frequently obscured the pathway to the accessibility information provided - often the accessibility link was visually very indistinct, through the way in which it is presented and/or the position on screen.
- Accessibility information is usually presented as technical information
- Most accessibility information was presented from a technical standpoint, focusing on technical description of accessibility features and guideline conformance, and not as advice supporting users in specific accessibility related tasks (like enlarging text size). Information was often limited to a specific impairment.
Key observations - User Evaluations
Our user evaluations revealed the following:
- Limited awareness of the concept of accessibility adjustments
- Participants had significant difficulty with the concept of making accessibility changes, such as adjusting the page appearance. There was also a distinct reluctance to explore how changes might be made. However, those participants who discovered direct manipulation features, for example features that allowed text resizing, generally liked these features.
- Over-optimistic assumption of user knowledge
- Many sites presented information that required a user to know which browser they were using, something unknown to several participants. Several participants expressed frustration at the level of what they saw as 'jargon' used in the advice provided, to the point that some participants appeared to gloss over large amounts of text.
From the key observations of our research, a number of design recommendations can be made:
For Web developers
Key design recommendations:
- Provide a clear and consistent link to the Accessibility page
- This should be at the top of each other page of the site. Use a larger than normal text size for the text of this link; ensure there is high contrast between text and background colours. Use a phrase such as "Help using this site" rather than "accessibility".
- Provide information in a simple, task-focused way.
- Match information to accessibility related tasks rather than descriptions of what has been done. In particular, information should cover how the text can be enlarged, text colours can be changed, and how the keyboard can be used for navigation. Information on alternative formats of information is also valuable. If bespoke accessibility features are provided, then describe them here. Don't assume that the user knows what browser they are using - provide information to help them identify their browser.
- Take advantage of existing accessibility related information on the Web
- Refer to sites like the BBC and Ability Net My Web My Way web site. This encourages users to visit an authoritative and stable resource from a Web site's accessibility page, and enables them to bookmark the site for future reference. Doing this also avoids site providers having to write and maintain their own detailed instructions on how to perform tasks using specific browsers.
- Consider using bespoke accessibility features
- Useful features include text resizing features and audio versions of page content. We accept there is a debate over whether a Web site should attempt to replicate browser functionality (or compensate for a lack of such functionality), but there are clear signs of positive benefit of such features for many users, and they should be seriously considered, at least as an interim feature. The significant challenge is ensuring consistency across different sites - in terms of location on the page of access to these features as well as functionality.
These steps may enable the development of a common template for the format and content of an accessibility page, which can be made available for Web site developers to use, as much of the information will be generic rather than site-specific. However, it is likely that the accessibility page will need to provide additional information unique to the site, in particular a description of areas of the site in question with known accessibility problems, and how users can overcome these problems.
This research was conducted as an initial investigation into this area, and so we recommend further work to validate what we suggest here:
There is a need to increase user awareness of the concept of accessibility in terms of the degree of freedom users have (or should have) in making changes to the appearance of a Web site, or the way in which they interact with the site. The user evaluations indicated that there was a lack of appreciation that it is possible to make changes to the appearance of a Web page.
Related to this, many participants had difficulty understanding the term 'accessibility', which has many other connotations, even within a technology setting. A possible research topic would be to establish whether this is the most appropriate term to use to describe what is perceived by many people as "information to make it easier to use the computer."
In line with exploration of the most appropriate term to use, and following some participant comments, there is scope for identifying an appropriate visual metaphor for Web accessibility information. This could then be used as an icon indicating a link to accessibility advice to be used by Web sites to signpost their accessibility information.
The research also adds weight to the argument that browsers can and should do better in exposing accessibility features to users who may not be aware they exist. There is a paradox in that while open source browsers like Mozilla/Firefox enable enhanced functionality through browser interface improvements or extensions that can be freely installed, such as Accessibar for Mozilla/Firefox, these more capable solutions are likely to be least known by those who may benefit most from them. Conversely, the dominant browser, Microsoft Internet Explorer (IE) has accessibility features that are less immediately prominent, and do not always work - in particular the text resizing facility. Yet this browser is likely to be more widely used by those with less technological awareness, who may not be able to request another browser even if they were aware that alternatives exist.
Until either alternative browsers become more widely available in private and public access PC terminals, or the advances made in open source browsers are reflected in Internet Explorer, many users may continue to be impeded by their browsing technology. Initiatives such as the Web Adaptation Technology from IBM attempt to address browser shortcomings, but for people who are unaware of such technology, browser improvements are the most effective solution. Until these are implemented, how can time spent promoting improved browser usability be balanced with awareness-raising and educational initiatives that encourage users to adopt browsers that better support their needs?
We would be happy to discuss research projects that explore the above topics with any interested parties - please contact us if you're interested!