Section 5.1: Defining Accessibility Evaluation and Testing, Part 2
Automated Evaluation
Automated evaluation relies on the use of a tool to identify possible accessibility errors. Typically the tools rely on an algorithm to identify specific code elements or attributes. For example, a tool can check the code of a Web page, look for the images and check whether there is an ALT attribute associated with each one. This ALT attribute gives the text equivalent for the image.
These types of tools can be very useful for getting an overall sense of the accessibility of the site and a way to check some basic accessibility issues.
Commonly used free Accessibility tools for automated evaluation of a Web page include:
These tools also identify warnings or areas that cannot be evaluated with an algorithm.
Evaluation & Testing: Automated Tools
- Automated tools typically identify errors and warnings
- Errors involve incorrect or missing code: Missing alt attributes, form controls without labels, incorrect markup for complex tables, etc.
- Warnings are about items that require human judgment: Use of color to convey information, need for long description, whether scripting conveys information or functionality needed to understand and use the content
Manual Review
The automated tools cannot identify all accessibility issues. Some of the tools provide warnings or reminders about areas that require manual review-that is, items that require human judgment. In most cases, these areas involve using your judgment to determine if the area the tool flagged is an accessibility issue. For example, images without an ALT attribute can be automatically indicated, but the appropriateness of the contents of the ALT text can only be determined by a person reviewing it. In the manual review, tools such as the Accessibility Toolbar can still be helpful for highlighting specific areas of a site that you need to review.
User Testing
User testing gives you feedback about whether the implementation of accessibility features actually works for users. It provides "quantitative and qualitative data from real users performing real task with a product". [ 1 
] In best development practices, users give feedback throughout development so that you are creating a project that will be easy for your audience to learn, navigate and use.
User testing for accessibility resembles typical user testing, with one important difference: People with disabilities are included in the subject pool in order to determine if individuals with and without disabilities have comparable user experiences.
