Evaluating accessibility the TechDis way

Yesterday I received a word document via email from a colleague which served as my introduction to the http://www.techdis.ac.uk project. This is a JISC funded resource that seeks to a resource to assist in implementing accessibility and usability in a range of organisations. Read their site for a more detailed explanation.

I’ve had a quick look at the document and it piqued my curiousity. Over the past three or so years we’ve aimed to integrate accessibility and usability into all the new work that we create, but we’ve been remiss in formalising what we’ve done and documenting the methods we’ve used in aiming to make our sites usable and accessible.

Over that period we’ve had many discussions within the team about the pros and cons of particular methods but we’ve no record of our thought processes. I reasoned that an evaluation exercise would shine a light on our decision making and help us to do things better. An immediate attraction of the word document is its brevity.

Once one has selected the URL’s you wish to evaluate it’s away you go with the technical stuff. I’ve decided to look at Glamlife , one of our sites that should be pretty accessible, so that I don’t have too daunting a list of things to fix, and the evaluate the evaluation! The URLs I’ve chosen to test are

The home page, a section home page and a content page.

Technical HTML Conformance

“Each HTML page should be put through at least one HTML validator and each CSS should be validated using a CSS validation service.”

I used The W3C Mark-up Validation Service

The one remaining error on the home page is an element that we use in our rollover javascript. An explanation of how this works can be found on http://jehiah.cz/archive/simple-swap I currently don’t know how to change this so that it validates, but it shouldn’t cause the page to choke on any accessibility tests so we can leave it in. It’s only on this page so my efforts are better employed elsewhere.

The section homepages are a different layout and based on a different template, and all validate as XHTML 1.0 strict. Being based on templates the overwhelming majority of pages validate, unless there are things entered in the content that break the validation. This is an unavoidable hazard of a CMS. However, it’s worth remembering that valid does not equal accessible. Validation is just one of the tools available to us that assists in assertaining that our code is of a certain standard.

CSS Validation

A quick run through of the CSS we use for glamlife came up a with a few errors, which were easily fixed. I think the CSS for the site could do with some tidying up, to make development easier

Screen Size/Resolution

The TechDis site recommend http://www.anybrowser.com/ScreenSizeTest.html as a resource to test ‘common’ browser resolutions, though they seem pretty tiny to me. The copyright info on the page says 1995-2001, so presumably this is when the info was relevant. Instead of that I’ve chosen to check our general site stats, and as at Feb 2007 we’ve around 10% of users using 800×600 resoultions. We’ve been pretty conservative with our design to accommodate this size.

Enlarging the font size

The font size scales up well in Safari and goes up and down through the text size options on IE6 on XP. (the smallest size is almost illegible), which I think is a common problem.

Is the site usable without images?

Yes

Does the site work without JavaScript?

Yes

Can you use the site without using the mouse?

I can tab through the site, but to thoroughly test this you would need to have input from someone for whom this method of navigating a site is important. How one lets the user know about the keyboard shortcuts that we use is an important question that needs more work. We have however, chosen our access keys with regard to the UK government’s advice

Navigation without a mouse (or other pointing device) is something that we need to get out and learn more about. And the Tab order of the page is something that we can defininitely develop some University standards on. The access keys for Glamlife can be found on the Accessibility Statement page.

Automatic Checkers

The guidance from TechDis recommends some automatic checkers

http://www.cynthiasays.com/ Used this to check, using the WCAG priority 1,2 and 3 options. One error we had was on link text. The checkpoint is 13.1 Clearly identify the target of each link. The fix for this will be a small change to the content.

Webxact

We used webxact throughout the development phase of Glamlife and so it does not flag up many errors and/or warnings. The one error that comes up is something that we have looked into and may look into again. It certainly not a showstopper.

Accessibility Heuristic Evaluation

This sounded quite daunting to me until I realised that it’s a method to include some judgement based criteria for any site that you care to evaluate. Essentially one answers the following questions and assigns scores for how fully you feel the site satisfies the questions. There is a fuller explanation on the TechDis site.

  • Does the website have a clear, intuitive and logical navigation structure?
  • Does the website have a clean, consistent and clear visual design?
  • Does the site provide appropriate alternative descriptions for all visual elements?
  • Are all the website interactions, forms, navigation scripts etc accessible and usable?
  • Does the website use clear and simple language, appropriate for its audience?

We can answer yes to all these questions, gaining 3-4 points for each answer. Which tells us that we are doing ok. They are very useful for providing a mechanism to evaluate non-technical issues, matters of personal preference or other qualitative issues.

Usability Heuristic Evaluation

Similarly we score well on the usability Heuristics too. We’ve kept the site pretty free of inappropriate Frames, Java, Flash and animation, and the content is clear and well written.

Assistive Technology Testing

We have tested the site with the screen reader that is installed as standard in the IT labs here at Glamorgan, however, This is of limited value because we are not regular users of the software, and consequently this skews the tests. Useful feedback would come from a regular screen reader user.

Just prior to launch I posted a request for users opinions on the accessify forums, and was rewarded with some useful feedback.Specifically, a user explained what was being read out when he viewed the site suggested some changes to our access keys and a few other things about what was actually being read by his screen reader. Which we then implemented. The whole area of assitive technology testing is an area that the University needs to get to grips with. We have done what we can.

Browser Compatibility Testing

Have tested on IE6, Firefox,Safari which are the most popular on our site with approx 98% of users using them. We also feel that the web standards based approach that we’ve taken is likely to be helpful to users with browsers that we have not directly tested.

Conclusion

The evaluation exercise has been good to go through, but it does take some time. If one was doing it for sites with bigger issues about accessibility then I can imagine it would be an arduous task, but a completely necessary one. Accessibility and Usability was central to Glamlife from the start and it still threw up some issues, and continues to do so. They are not features that once provided can be ticked, it is a continual process of evaluation and development.

If you’ve found this article interesting, please add some comments. As a team we are always keen to get feedback on what we do.