A macroscopic web accessibility evaluation at different processing phases

Abstract

This paper details a comparative experimental study to understand the difference of the Web's accessibility properties regarding two different evaluation approaches: using WCAG 2.0 techniques and targeting the pages as they reach the browser; using WCAG 2.0, but evaluating the pages after the browser processing, thus as they will be delivered to the end-user. For that, we evaluated over 20000 Web pages using already established accessibility metrics. We then compared the results obtained from the WCAG 2.0 evaluation of the two processing phases. We observed some changes in the macroscopic properties of the evaluation. Regarding the comparison between the two phases, we observed a narrower distribution of quality, i.e., the worst pages are in fact not that bad, and the best ones not that good.