Comparison of Methods for Teaching Accessibility in University Computing Courses
Qiwen Zhao, Vaishnavi Mande, Paula Conn, Sedeeq Al-khazraji, Kristen Shinohara, Stephanie Ludi, Matt Huenerfauth · 2020 · Proceedings of the 22nd International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS) · doi:10.1145/3373625.3417013
Summary
This paper presents the first controlled, longitudinal comparison of methods for teaching accessibility in university computing programs. Conducted over four years (2016-2020) at Rochester Institute of Technology, the study involved 29 sections of a required Human-Computer Interaction course taught by 10 different professors, with over 400 students. Students were assigned to one of five conditions: a control group (no accessibility content), a lectures condition (one week of accessibility lectures), a projects condition (team design project with an accessibility component), an interaction condition (projects plus direct interaction with someone who uses assistive technologies), and a team member condition (projects with a fellow student who has a disability on the team). Students completed identical surveys at three time points: before the course, immediately after, and 18-24 months later as seniors. The surveys measured three dimensions: likelihood of considering people with disabilities in a design scenario (Voting), awareness of accessibility barriers and assistive technologies (Awareness), and technical knowledge of accessible programming techniques (Knowledge). The study addresses a significant gap in accessibility education research, which has previously relied on anecdotal faculty reports and short-term evaluations without controlled comparisons.
Key findings
In the short term (pre vs. post-course), the Lectures, Projects, and Interaction conditions all produced significant improvements across the Voting, Awareness, and Knowledge measures, while the Control condition showed no significant gains on Voting or Awareness. Surprisingly, the Team Member condition only showed significant improvement on the Knowledge measure in the short term, despite students having daily contact with a disabled teammate. Follow-up interviews revealed this was partly because RIT has a large Deaf and Hard of Hearing population (nearly 5% of students), so many students had already interacted with DHH peers before the course, reducing the novelty. Additionally, team dynamics sometimes led DHH members to handle accessibility testing themselves, shielding hearing teammates from direct learning experiences. The most striking finding was the long-term analysis: when surveyed 18-24 months later, almost all short-term gains had disappeared. Only the Projects condition retained a significant difference — specifically on the Knowledge measure (p<0.001, n=29). The authors speculate that hands-on project work helps convert declarative knowledge into procedural knowledge, which persists longer in memory. This finding strongly suggests that single-course interventions are insufficient for lasting accessibility learning, and that accessibility must be reinforced throughout a degree program.
Relevance
This research is essential reading for anyone involved in computing education or organizational accessibility training. The finding that accessibility knowledge gained in a single course fades within two years is a sobering reality check for universities that treat accessibility as a one-time topic. It mirrors patterns seen in professional development: a single workshop or training session rarely produces lasting behavioral change. The practical implication is clear — accessibility needs to be woven throughout computing curricula, not confined to one module. For organizations, the parallel is that accessibility training should be ongoing and embedded in regular workflows rather than delivered as isolated events. The study also validates that lectures, projects, and interactions with disabled users are all effective short-term teaching methods, giving educators flexibility in choosing approaches based on available resources. The finding that hands-on project work produces the most durable learning is particularly actionable.
Tags: accessibility education · computing education · pedagogy · curriculum design · longitudinal study · human-computer interaction
Standards referenced: WCAG · Section 508