LessFuss Design

WebAIM (Web Accessibility In Mind) have published the results of a screen reader user survey they did at the end of 2008. Some of the results confirmed what many of us assume (or are led to believe) about screen reader user behaviour, but others were a bit of a surprise. The full results and some analysis is published on the WebAIM website [external link], here’s my list of the main points of interest.

1. Headings – 76% of screenreader users prefer to navigate web pages by headings. Therefore properly marked up pages that use a range of headings (in the right order) is important.
2. Access keys – Only 10% of survey respondents say they never use access keys (22% say they’ll use them ‘when available’). Access keys had gone down my list of priority accessibility features following advice I’d received from a testing company. I’ll definitely be giving this some thought.
3. Site maps – 51% ‘seldom or never’ use site maps, despite claims from some accessibility consultants that these are essential. Site maps have always been touted as one of the things you should do for SEO (so the search bots can easily index all your pages), but clearly they’re not that much use to screenreader users.
4. ‘Pop up’ windows (when you follow a link and opens a new, often fixed size browser window) – 53% find them somewhat or very difficult to deal with (because they take the focus away from the web page they were on, often without any warning). That’s no surprise, but pop-ups are still very common on the Web.
5. Flash – no real surprise that 71% found Flash content ‘somewhat’ or ‘very’ difficult to use. The importance of providing alternative routes to content alongside Flash movies, and of making the best effort to make the movies themselves as accessible as possible (or, if you can help it, avoiding Flash), is clearly as important as ever.

Other survey findings show that, as with sighted users and their knowledge and experience of computers, there’s different levels of skill and experience amongst screen reader users. Although the results aren’t broken down to show how, say, experienced users answered compared to novice users, it’s evident that an experienced user will probably have modified their software settings and will probably find some tasks much easier than a newbie or less frequent user would.

All of which goes to show that while user surveys are useful for showing trends in behaviour – and this is one of the most comprehensive I’ve seen – there’s no substitute for talking to people or, better still, watching them using websites with assistive technologies, to get an idea of the real issues they come up against.