Friday, December 24, 2004
Usability and Accessibility
Yesterday I made the mistake of going to the post office around lunchtime. I needed to get a few things shipped, including important stuff like a present for my god-daughter and a confirmation of my place in the London Marathon in April. The queue snaked around the office. There must have been 30 people in front of me. 10 of them were actually in the queue to buy envelopes or stamps, rather than the queue for the counter - but that wasn't easy to spot. Usability problem one - don't put the stamp counter in the middle of the office, right next to the place where everyone queues for other stuff. Borders in Oxford Street just learnt this and have remodelled their ground floor completely. But the real reason that the queue was so large wasn't down to it being lunchtime or it being near Xmas. It was the introduction of "chip and pin" readers. No-one ahead of me appeared to have a clue how they worked. And why should they? The Post Office, with infinite wisdom, requires that you insert the card in the slot upside down - i.e. with the magnetic stripe facing upwards. Every ATM in the country works the other way, so why are we doing it differently here? When I lived in Vienna, the ATMs worked with the mag stripe upwards and there were always queues of tourists trying to figure out why their cards wouldn't go in the machines. How would they know that Austria had a different standard from the rest of the world? Is it only the Post Office that has done it this time round? So much for consistency. What got me on to this was reading a piece on the Reg where a group called Knexus - apparently an exclusive club (in the Groucho Marx sense perhaps) for execs in Fortune 100 companies - was raising issues about the problems with accessibility design. The clubsters are worried that it's not obvious what accessible design really means. One went so far as to say that it's all a bit "over the top". Heavens above. Naturally the DRC laments this point of view and said that it isn't about "doing the minimum" but about doing it all. They're going to be at "considerable legal risk" apparently if they don't get it right. Tom Adams, at eGU, speaking at a recenty Parliamentary Internet Group (and quoted in this Reg article) took (what I think is) a better line - he said that there needed to be some reference to standards. The DRC doesn't agree. So companies and public sector bodies are left to do what they think is right, whether it's A, AA, AAA, LocalGov LAWS standards, RNIB, RNID, Bobby, DDA, DRC guidelines or whatever. Plainly that's daft. Meanwhile, there are greater than 200 disability rights groups watching to see what happens. I spoke at a Public Sector Forums conference on Accessibility back in July at my almer mata, City University. I too was worried about the lack of clarity on which standard to follow and that if we wanted real "Universal Accessibility" then a single widely-endorsed standard was required along with accredited test tools and a full awareness campaign. In my list, that doesn't include the threat of legal action unless someone persistently shows a lack of attention - but that could only be once the other pieces were in place. My own view is that accessibility follows usability - if you get the design right it will be accessible, but few start with a good design. Instead they layer complexity upon complexity making the web experience frustrating for their customers. Public or private, no different. Here's an example: Should you allow inline links (as I liberally scatter throughout this blog) or insist that all links are outside the main body of text (as directgov insists on for its authors)? Most folks want inline links because they think it makes things easier. I disagree. If you lose a link, you have to edit the text; if a link changes content, you have to edit text. So that scores -2 in ease of management. Screen readers hate inline links - and so do the users of screen readers. So that's another -2. That's -4 out of 4. Not much good. But there are no standards on this and, worse, no agreement. No agreement means standards could be a long way away. And that means a greater legal risk for some companies who might do what they think is right but still be in the wrong. And the folks that are hurt the most in this are those who want a good web experience, whether they are part of the 14.3% of the population with some disability that affects their ability to use the web, or the 85.7% that don't have such a claim. Let me tell you though - if you get the design right for the 85-odd%, it will be pretty close to right for everyone else too. When legal threats start flying it's worrying. Trying to sue a corporate for not having an accessible site may turn out to be as hard as suing a restaurant for not cooking a steak the way you like it. One man's rare is another's medium. There are no clear standards for what "rare" means for a steak. I tend to seek out restaurants that I know cook it the way I like it; and avoid others. I don't consult lawyers. In practice, sites will need to adhere to some basic standards - most likely the AA - and then common good sense will enhance that slightly (to include some of the additional checks that, say, RNIB want). But that's not the end of the problem - the wide range of disabilities and varying levels of them coupled with the disparate monitoring groups will mean that, at any one time, most sites are not fully accessible to one group or another. So, good design will matter most; and accessibility will follow that.
Posted by Alan at Friday, December 24, 2004