Improving the outcomes of students with cognitive and learning disabilities: phase I development for a web accessibility tool
Aaron Andersen, Cyndi Rowland · 2007 · Proceedings of the 9th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS '07) · doi:10.1145/1296843.1296882
Summary
This poster paper from ASSETS 2007 reports on Phase I of a project to extend the WAVE (Web Accessibility Versatile Evaluator) open-source tool with a new suite of evaluators specifically targeting the cognitive load of web pages. The work, led by WebAIM and the National Center on Disability and Access to Education (NCDAE) at Utah State University in partnership with Adobe, was motivated by the observation that while considerable progress had been made on web access for blind, deaf, and motor-impaired users, students with cognitive and learning disabilities — by far the largest disability group in U.S. schools, with an estimated 3.5 million affected students — had been largely overlooked by automated evaluation tools. The authors used an iterative, stakeholder-driven research methodology: an initial pool of potential cognitive-load barriers was assembled from the literature and from expert opinion, then ranked and critiqued by panels of web accessibility experts, cognitive and learning disability specialists, parent advocates, and industry representatives. The ten top-ranked barriers were then field-tested in elementary, middle, and high school students using paired web pages — one with the barrier present and one without — to verify that mitigating each barrier produced measurable improvements in learning, comprehension, or task efficiency. The output of this process was a Content Requirements Document and a draft Technical Requirements Document, the latter of which forms the basis for the algorithms to be embedded in WAVE.
Key findings
The authors identify cognitive load as a multi-dimensional construct in web design that includes both obvious factors — reading level, syntactic complexity of text, consistent navigation, and the ability to suppress pop-ups or moving images — and less intuitive ones, such as the proportion of white space on a page, the use of headings to chunk information, and font selection for readability. Through their stakeholder ranking and field-testing process, the team narrowed the broad cognitive-load problem space to ten priority barriers that empirical testing with students confirmed actually affected learning, comprehension, or efficiency. The paper also flags an important practical constraint that emerged during specification: not every barrier identified as important by experts can necessarily be evaluated by an automated tool. Readability is offered as an example — if no free or open-source readability evaluator is available, proxy heuristics such as detecting words or sentences over a certain length may have to substitute. This honest acknowledgement of the gap between what experts know matters and what algorithms can detect is one of the more useful takeaways for the automated-testing community.
Relevance
Although nearly two decades old and presented as a poster rather than a full paper, this work remains historically important as one of the first concerted efforts to bring cognitive accessibility into the mainstream of automated web evaluation. The methodology — combining literature review, multi-stakeholder ranking, and student field-testing of paired barrier/no-barrier pages — is a model worth emulating for any team building accessibility tooling for under-served disability groups. For practitioners, the paper is a reminder that cognitive accessibility is still poorly served by automated checkers compared with WCAG conformance testing, and that meaningful evaluation of cognitive load often requires heuristics, manual review, and user testing rather than purely deterministic rules. The clear limitation is that the paper reports only on Phase I and does not publish the final ten barriers or the Technical Requirements Document, so readers must look to follow-up work and to current WAVE releases to see what was ultimately implemented.
Tags: cognitive accessibility · cognitive disabilities · learning disabilities · cognitive load · evaluation tools · WAVE · WebAIM · web accessibility · students · education · automated testing