Finding an optimal method for conducting accessibility evaluations of the Norwegian Tax Administration Website

Abstract

This research was done in collaboration with the Norwegian Tax Administration who had concluded that they were in need of an accessibility-testing manual that they could use for ensuring accessible contents in their website. During this research an accessibility-testing manual has been developed, the intention for this manual is that it can be utilized by usability experts, testers and developers so that they can be assured that their web contents are accessible. This was done by studying literature about issues regarding accessibility testing, web accessibility and suggestions for solutions. All solutions were included in the final document. Literature about screen readers and automated accessibility evaluation tools was also investigated. No automated accessibility evaluation tool could be recommended solely based on the literature review. Therefore, an evaluation of five potential evaluation tools was conducted; two of these were suggested as supplementary evaluation tools, which testers might utilise, in the final testing manual. The purpose of this research was to find out common challenges and problems faced by users of screen readers when interacting with the website that was investigated. Based on the results some suggestions could be provided, as to how accessibility testing may be conducted in the future. A limitation of the scope was that not all accessibility evaluation methods were included in the analysis of the page. Three of the most common methods for accessibility testing were included. A qualitative research method was selected for the research. This consisted of user testing with users of screen readers followed by a short interview, evaluation of automated accessibility evaluation tools, and a heuristic evaluation. The participants were given the same tasks that required interaction with several pages and forms. All three methods were used to test one common webpage; the results were compared in a table. The results indicated that the methods could be combined. No method found all accessibility issues; the accessibility-testing manual includes common issues that were easy to find manually that testers might check for themselves, suggestion for use of automated accessibility evaluation tools, and planning in addition to conducting a user test with people with disabilities. The accessibility-testing manual presented to the Norwegian Tax Administration is this research’s most important contribution to the field of web accessibility research; it indicates that several accessibility evaluation methods should be mixed for optimizing accessibility testing, there are rooms room for improvements in the field of web accessibility testing.