Recommendation 4.2.22:
Follow WCAG guidelines when applicable to chatbot content
Benefits Users
Visual | Auditory | Cognitive | Speech | Physical
Phase 1: Gather & Organize
Additional Information
- The WCAG document does not specifically address the wide range of cognitive abilities in target users. Latest recommendations from W3C’s Cognitive and Learning Disabilities Accessibility Task Force (COGA TF) have information explicitly directed at cognitive disabilities that can fill gaps in WCAG guidance. Ultimately, user tests will reveal how accessible the chatbot is.
Read More
- The table below matches recommendations from this playbook to the W3C WCAG and COGA recommendations to which they correspond. In most of these cases the W3C source gives explicit information such as user stories and techniques that directly address the recommendation from the playbook.
Phase 2: Design & Implement
Design Question
-
How will your chatbot handle gaps in existing guidance?
- Of special interest are the recommendations included in this playbook that do not correspond with existing WCAG or COGA standards. These recommendations are ripe for future research and definition. An example is “Clearly identify the chatbot as a bot and don’t pretend to be a human” (4.2.1). Future research can clarify how lack of messaging or different kinds of messaging of the bot’s nature affect users with different kinds of disabilities.
Tip
- When WCAG isn’t clear about chatbot features, go to users instead.
Examples
- When keeping the content simple (4.2.5), refer to WCAG Criterion 3.1.5 for assessing the content’s reading level as well as COGA Design Pattern 4.4.1 for making sure words are clear.
Phase 3: Test & Evaluate
Self Check
-
Did you review WCAG as you completed the Gather and Organize phase?
-
Did you revisit WCAG during the Design and Implement phase?
-
Did you test and evaluate the chatbot against WCAG standards for content?