Perfect Circle

One of the most frustrating yet fascinating challenges to achieving an optimally accessible Web is trying to complete a perfect circle. Three parts of that circle are covered by the W3C Web Accessibility Initiative’s triumvirate of guidelines: conformant content (WCAG), created by conformant authoring tools (ATAG), and accessed using conformant user agents (UAAG). This approach has some practical flaws – not least in that there’s a missing part of the circle: conformant users.

As awareness amongst web designers grows of the need to create accessible web content, there’s a natural frustration that despite the designer’s best efforts, people can still experience accessibility problems. In the WAI model of accessibility, this might be defined as a user agent problem. A lack of conformance by browsers with UAAG means it still seems to be extraordinarily difficult to make basic,  effective and persistent changes to web page display (for example changing text size, colour schemes) or interaction style (efficient keyboard shortcuts). Following UAAG, especially Guideline 11 and Guideline 12,  would – the theory suggests – make it easy for a user to make the necessary adjustments to meet their accessibility requirements.

The short term, but unsatisfactory, workaround to this problem, has been for web sites to provide accessibility aids as part of their content – features like text resizing widgets and alternative style sheets. Embedding accessibility features in web content has obvious and significant shortcomings – they can only every apply to one web site; so a user can’t rely on them being available across all the sites they want or need to visit.

This ‘give a man a fish or teach a man to fish?’ conundrum has been a focus of attention for many accessibility and standards advocates. In terms of raising awareness, efforts have been made to make it easier to apply accessibility changes through browser widgets (e.g. Firefox text size extension), scripts for accessibility (e.g. AccessMonkey), or user-definable access keys. There are web sites such as the BBC’s MyWeb MyWay, and video tutorials on how to make basic accessibility changes; there’s an ongoing debate on whether a site’s accessibility page is the best place to provide helpful accessibility information.

However, the paradox with all these efforts is that the people who are most likely to be affected are those who are also highly unlikely to find these solutions independently. Without assistance, these valuable tools may remain undiscovered by the people who need them most.

What’s important to realise is that this issue concerns not those with significant impairments that might require, for example, a full blown screen reader, and where an accessibility problem – and solution – is pretty easy to define.

Instead, the people who are most likely to be affected by an inability to make accessibility changes are people with relatively minor, but multiple impairments, which may fluctuate short-term but gradually decline long-term. This includes many older web users, experiencing age related capability decline – and what makes the problem difficult to solve is that the gradual nature of change, allied to an unfamiliarity with the nature of the Web, these users might be unaware they have accessibility requirements.

We’re used to assuming that in accessible design, the user at least knows they have accessibility needs, but what happens when they don’t?

There’s been a lot of work focusing on defining and storing user needs, and using them making adaptations to resources based on these accessibility needs, for example the Supple project. User profiling for accessibility is now standardised, in an e-learning context, in ISO 24751. The WAI-AGE project  is investigating the overlap and differences between ageing and web accessibility; and browser manufacturers like Opera do now seem to appreciate how difficult ‘simple’ accessibility tasks can be for many users.

This is all important work; but beyond technical problems we need to look at how people develop skills in IT use, in how they learn web browsing techniques – and what role, if any, accessibility plays. And we need to understand better how we can discover that a person’s visual, cognitive or mobility capabilities have declined to such a point that applying an accessibility solution (or changing an existing one) would help to sustain independent access.

Tracking a user’s capability changes, keeping this information as accurate and up to date as possible, and using it to make helpful changes – all in an ethical way – is something we plan to tackle in the SUS-IT project over the next three years. I’m really looking forward to working with a team drawn from sociology, psychology, information science as well as computing science to tackle this problem in a more holistic way.

And maybe at the end of it all, we can have a situation where, to use one of my favourite analogies, awareness of the text resizing functions of a web browser is as widespread as that of another accessibility aid – the TV volume control.