Efficient Accessibility Testing
Version 1.0 of the W3C's Web Content Accessibility Guidelines (WCAG) contain 65 checkpoints, so significant resources would be required if every page of a site was to be tested against each checkpoint. Fortunately this is not necessary, and there are several efficiencies to be made:
- Only test the appropriate checkpoints. The 65 checkpoints are divided into three Priority levels. The Priority 1 checkpoints are the most important, so if you only need to achieve a basic level of accessibility it is only necessary to test these 16 checkpoints. Most websites should aim to meet the Priority 2 checkpoints, which requires a further 30 checkpoints to be tested, but this still represents a 30% saving. Many of the Priority 3 checkpoints require coding techniques that are not part of mainstream development practice. It is rarely worth testing them unless the website has been specifically designed to meet this level.
- Don't test every page. Most sites are based on templates, and it is usually sufficient to test one example of each template. Additional pages may be required depending on the content e.g. data tables, forms, lists, images maps etc. We will gladly perform an audit of your site to determine the minimum number of pages to be tested; on a typical 100-page website we often find that it is sufficient to test just 6 or 7 pages.
- Lightweight test tools. Lightweight tools such as the AIS Accessibility Toolbar and Colour Contrast Analyser greatly improve the efficiency and accuracy of code inspection.
- Expertise and experience. It takes time to learn all the web technologies such as HTML, XHTML, CSS, JavaScript, Java, Flash, DHTML, AJAX etc, to understand the rationale behind the WCAG checkpoints and develop the techniques for testing them. We are already at the top of this learning curve, so our efficiency will lower your costs and shorten your timescales.
Note: WCAG 2.0 was released in December 2008. The number of checkpoints is slightly different, but all the same principles still apply.
