“AI is reducing options I had used”: exploring faculty perceptions of accessible course design in a ChatGPT world

Abstract

Demand for individual accommodations across Canadian universities has skyrocketed, leaving accessibility services to face an overwhelming workload in a model that requires students to formally prove a disability. Universal Design for Learning (UDL) is a pedagogical philosophy that, with sufficient uptake by professors, may serve to reduce the burden experienced by accessibility services and students alike. This study examines professors’ knowledge and perceptions of accessibility and UDL at an Ontario university, including barriers and facilitators to implementation, using a mixed-methods design. Key themes include professor’s’ frustrations about generative AI reversing much of the progress in accessible course design, limited capacity to learn how to design and run an accessible course, controversy over recorded lectures as a method to enhance accessibility, and potential solutions for promoting accessibility while acknowledging faculty wellbeing. Frustration over managing individual accommodations is evident. Of 20 commonly employed strategies in UDL, faculty demonstrated a lack of awareness of at least four strategies and demonstrated a lack of uptake of seven despite having awareness. Our findings suggest that increased support is required to improve the acceptability of UDL in the face of generative AI, and can come in the form of additional resources, forums to discuss common challenges, better support from accessibility services, and a higher degree of recognition for the initial work that goes into creating an accessible course.