Joe Clark: @media 2005: Day 1

Accessibility: Simple Facts About a Tricky Subject

Notes from a presentation given 2005.06.09 at the @media conference, London

Topics

Today we’re going to talk about:

The true audience for the World Wide Web, and who’s been forgotten so far
What accessibility is and is not
Some statistical facts about people with disabilities
And, in broad terms, the needs of different disabled groups in using the Web

The old audience

Our mind’s-eye image of a Web user used to be a young single guy using el-cheapo equipment.
Later on, Internet marketers tried to pressure us into believing that everybody had so many interests that we could all go to the same sites. Remember portals?
By now we’ve grown up a lot and we understand that there are many kinds of Web users. We now know there’s a distinct Web site for everybody. We also know that people want to create their own Web sites, as with the millions of Weblogs. We’re much more sophisticated these days as developers.
But even sophisticated developers have their own biases and limitations. One method people have developed to test Web sites is the use of personas or use cases. We come up with imaginary Internet users and try to think how they would hypothetically use our sites.
These personas may be slightly older or younger people, or women, or people with specific interests or tasks or goals.
But in nearly all those personas, the imaginary people we’re talking about would never know the difference between accessible and inaccessible sites. Disabled persons, however, will definitely know the difference.
And the difference we’re talking about is the difference between conveniently and easily using a Web site within the abilities you have or sitting there in frustration. In extreme cases, it’s the difference between using and not using a site at all.
People with disabilities have always been part of the Web audience. And that includes your corner of the Web. The next stage in cultivating the Web is to accept reality and accommodate this group.

Which kind of accessibility?

If you’re at this seminar, you know we’re not making Web sites for nondisabled people anymore.
We’re also not making Web sites just for blind people.
Our goal in these sessions today and tomorrow is to learn how to make Web sites that almost everyone can use, within extremely broad but still reasonable limits.

What is accessibility?

The definition I use is “Making allowances for characteristics a person cannot change or cannot change easily.”
Hence, accessibility is about people first, and only incidentally about the technologies they use. If you make a site that’s accessible to people with disabilities, it may work better on a PalmPilot, perhaps, but that’s a bonus, not a goal.

Why be accessible?

Many of you will have read about the reasons to make Web sites accessible many times before. So I’m gonna start with a reason you probably will not have heard before.

Because you can: Most of the time, it’s technically simple and inexpensive to make Web sites reasonably accessible to most disabled people. And there are things we can still do even in the exceptional cases where it isn’t simple or inexpensive.; The Web is different from old media like books, which cannot carry accessibility along with them. Because the Web is an electronic medium, it can include accessibility features within itself. So let’s use them!
Because you have to: You’re really lucky here in Britain because your Disability Discrimination Act is all but unambiguous about applying to Web sites. It does apply to Web sites. Thus, many of you have a legal obligation to be accessible
Because it makes you money: Not all people with disabilities are poor or underemployed. One example is senior citizens, who often have considerable disposable income. People won’t buy your product or service if they can’t use your site
Because of good public service: The public sector has an obligation to serve the public. You aren’t doing that if the only people you’re serving are people without disabilities
Because you are mature: Accessibility is a component of the growing movement toward Web standards. It’s a more mature method of making genuine Web sites as opposed to Internet Explorer sites

What isn’t accessibility?

Here are a few things that are not what we’re talking about.

A text-only page. Text-only is not accessible or even desirable
Any separate “accessible” version, except in multimedia. In many cases you have no choice but to create a separate accessible version in multimedia, like a separate video file with captions or audio descriptions or both
Any page custom-crafted for only a single disabled audience. That means if you make a so-called blind-friendly page, you still have not achieved accessibility.
A page made for an unusual device, like PDAs or TV sets. People first, devices second.

Some facts about people with disabilities

I checked various publications from the Office for National Statistics. The most recent disability figures they published are from 2001. I always urge people to trust impartial statistics rather than statistics published by lobby groups, which tend to inflate the numbers.
Over 11% of the U.K. population has a disability relevant to usage of the Web. That’s 808,000 people.
3.6% of the U.K. population have a “sensory” disability. That means “difficulty in seeing” or “difficulty in hearing.”
Now, 3.6% doesn’t seem like a lot. Why are we going to all this trouble when we can serve 96.4% of the population by not going to all that trouble?
Because, according to National Statistics, 3.6% of the population amounts to over a quarter of a million people – actually, 255,000. I don’t think anybody in this room can justify excluding that many people when it’s technically possible to include them.
In addition, some of the people with a mobility impairment – 5.7% of the population, or 404,000 people – will be affected by Web accessibility. And here the numbers refer specifically to musculoskeletal problems of the arms or hands.
And finally, some people with a learning disability (about 2.1%, or 149,000 people) will be served by Web accessibility.
Figures from the U.S. show that people with some disabilities have different participation rates than nondisabled people – that is, disabled people use the Internet less frequently than nondisabled people do. It’s fair to assume this is true everywhere.
The problem here is that all these statistics are not specific enough. Until the census starts asking specific questions about computer and Internet usage among people with disabilities, our numbers are going to be approximate.
The point, though, is that we’re dealing with minorities. If we were using the fact that the majority of the population is not disabled as a reason not to provide accessibility, well, we wouldn’t be here in the first place. Whether the population you’re serving is 3.6% or 0.3% 36% or some other number is beside the point. We know already that we’re creating accessible Web sites for a minority group. Or actually, several groups.

Disabled groups

Now let’s run through the major disability groups and what needs to be done in Web development to accommodate them.

Visual impairment (1.9% of the population)

Best-known of the disabilities needing accessibility.
And, in all fairness, blind and visually-impaired people need the biggest single boost. The fact is that computers are mostly monitors and keyboards. And the fact is that the Web is mostly visual, even though it has an underlying structure that computers can manipulate. That’s why this group needs a bigger boost.
The needs of blind people are different from those of low-vision people.
In fact, we’re only now beginning to understand the needs of low-vision people. We finally have some research on the topic, for example. And tomorrow we’re going to learn all about how you can use stylesheets to make your site more usable to a person with low vision as opposed to total or near-total blindness.
Blind people typically use screen readers to surf the Web. Screen readers turn a Web page or your entire computer into speech output or Braille or both.
Low-vision people tend to use screen magnifiers.
These adaptive technologies have an influence on Web design. For example, a lot of people want to try to build a site that works with Jaws, the single most popular screen reader. But again that’s essentially making a blind-only site rather than an accessible site. And that’s building Web sites for devices, not people.
And a final point: Colourblindness is not a significant problem in accessibility if you follow some easy rules.

Hearing impairment (1.7% of the population)

Almost never considered, in part because there isn’t that much audio online.
There’s little research into the needs of deaf and hard-of-hearing persons.
What’s mostly needed are transcripts of audio. That’s sometimes difficult to do, especially on large sites.
What we also need is captioning of video, which is usually even more difficult. In fact, even after I spent a year on a research project into “best practices in online captioning,” the best advice I can give you is as follows: Don’t try this at home. If you need video captioned, send it out to qualified captioners.
There’s some evidence that easy-reader versions of Web sites are helpful, but that’s inconclusive. What I’m talking about here is versions of your Web site written in simpler language.

Mobility or dexterity impairment (5.7% of the population)

Also almost never considered.
Little research.
Here we need to consider a very large range of mobility restrictions. And we’re talking about the hands and arms, not the legs.
- → Explanation of slow adaptive technology
Old methods to accommodate this group can be reduced or eliminated by better page design. We have much less of a need now for skip-navigation links, for example.
Interestingly, sites with Flash-based interfaces show a large problem for this group because they cannot easily manipulate the controls inside the Flash content.

Cognitive impairment or learning disability (2.1% of the population)

Learning disabilities like dyslexia are included here.
Little research, but different subgroups have different needs.
Most Web sites may be impossible to adapt for people with learning disabilities, and that may not change.
Some sites can, however, be adapted for use and reuse in different methods of presentation. We have some pretty good techniques, all based on standards, that may be useful. We can show and hide different kinds of content quite easily, for example.
And some of the techniques I’ll be talking about tomorrow are applicable here.

Standards

So: If accessibility is a subset of Web standards, what are the standards we’re talking about?
At present, there’s only a single international nondenominational standard for accessibility: The Web Content Accessibility Guidelines or WCAG.
WCAG is one of the W3C standards – that is, it comes from the World Wide Web Consortium, who also wrote the HTML and CSS standards.
The current version of WCAG is version 1.0, which was released in 1999. That’s a very long time ago in Internet terms.
You’ve got three levels of conformance in WCAG 1.0. You can claim to meet all the guidelines in an increasing range of levels, and there certainly are a lot of guidelines.
- 15 Priority 1s → Level A: This is the minimum level and it is definitely better than nothing
- 29 Priority 2s → Level AA: Level AA is much more comprehensive and it’s been widely adopted by governments and companies. Level AA also requires valid HTML
- 18 Priority 3s → Level AAA: Somewhat extraneous and arguably impossible to meet. For example, the Disability Rights Commission surveyed over a thousand U.K. Web sites and found that none of them met Level AAA
If you’re looking for a determined and bloodthirsty defender of WCAG, you need to put somebody else up on stage here. WCAG 1.0 does a lot of basic things right, but WCAG also completely muffs it when it comes to anything advanced.
Just as an example, only now is the WCAG Working Group willing to admit that WCAG 1.0 essentially bans JavaScript, which it does.
I’ve spent the last three years publicly criticizing the WCAG Working Group for writing lousy standards. And the best part of it is that all those criticisms are online. If you’ve got nothing better to do, you can sit there and read them.
I haven’t had much effect, actually, and for a long time the WCAG Working Group pretty much considered me the bête noire of accessibility.
But now they’re working on WCAG 2.0, which might be done by the end of this year, but don’t count on it.
And the ironic news is that now I’m a so-called invited expert with the W3C. At best this makes me a courtesan rather than a king. And, in an Orwellian twist, I had to apply to be invited.
Nonetheless, we’re all working on WCAG 2 right now, and it will definitely be an improvement in many ways. For one thing, they’re trying to make it technology-neutral: It will apply to Flash and PDF just as much as it applies to HTML.
Still, there are a lot of things that aren’t going well in the development of WCAG 2. And there’s a good chance that conscientious developers like you might be stuck with totally asinine requirements that exceed anything we saw in WCAG 1. If you want to help out, we sure could use you.
Now, given the limitations of the Web Content Accessibility Guidelines, competing standards have sprung up.
Many of you have heard of the Section 508 requirements that apply to U.S. government Web sites and a few others. There are separate accessibility specs in Germany and Japan. And now there’s some work being done on a British standard for accessibility.
Even though I work on WCAG, I have to say I’m all in favour of this. We’re not exactly talking about reinventing the wheel here; all the guidelines agree on basic concepts like text equivalents for images. Where they differentiate is how well they handle real-world Web development. With multiple published standards in place, all of which cover the basics well enough, then market forces come into play and developers can adopt the standard that’s actually relevant to their kind of sites.
I think it’s possible that, in a few years, really good developers will stop claiming to conform to any particular accessibility standard and will simply claim that their sites are accessible to many people with disabilities. Internally and behind the scenes, they’ll still meet the requirements of one of the specs, but they won’t see any benefit to naming the spec they follow. They’ll design for accessibility, not to a standard. It will be based on performance or outcomes.

Summary

Accessibility is achievable most of the time
It’s a subset of Web standards
It’s a constraint within which Web developers may find it challenging and rewarding to work. I certainly do