Empowering Agile Project Members with Accessibility Testing Tools: A Case Study

Abstract

There is a growing interest in making software more accessible for everyone, which is emphasized by the numerous suggestions passed into law in many countries. However, many software organizations that use agile methods postpone or neglect accessibility testing. We aimed to understand how accessibility testing can be better integrated into the daily routine of agile projects by conducting a case study in a Norwegian software company. We investigated three accessibility testing tools: automatic checker, simulation glasses, and a dyslexia simulator. We hosted sessions at which agile project members used the tools while thinking out loud, responded to questionnaires, and were interviewed at the end. Additionally, we observed the project members for 18 workdays. Our results show that all three tools are suitable for agile projects. Especially the automatic checker and simulation glasses worked well in finding accessibility issues and were described as easy to use by the project members. Software organizations should empower their agile project members with low-cost and efficient accessibility testing tools to make their products more accessible for all. Doing this early and often in the development cycle may save the project from potential high costs at a later stage. Keywords: Accessibility testing Á Usability Á Cambridge simulation glasses Á SiteImprove accessibility checker Á WCAG Á Agile software development Á Universal design

What Methods Software Teams Prefer When Testing Web Accessibility

“…Our motivation for this study is to compare how other, lesserknown methods compare to the de facto method (the WCAG walk-through). Earlier, we have reported the findings from seven of these participants evaluating three of the tools [7].…”

Categories of User Impairment

“…The main characteristic of SUI is that users with various functional impairments should be considered. The key motivation for this is that such testing is particularly efficient to indicate general usability and accessibility problems (Schmutz, Sonderegger og Sauer, 2016, 2017, 2018Bai et al, 2018;Stray et al, 2019). Another reason is the necessity to mirror user diversity in development and testing based on population and impairment statistics.…”