Effects of Virtual Reality Properties on User Experience of Individuals with Autism
Lal "Lila" Bozgeyikli, Evren Bozgeyikli, Andrew Raij, Redwan Alqasemi, Srinivas Katkoori, Rajiv Dubey · 2018 · ACM Transactions on Accessible Computing · doi:10.1145/3267340
Summary
This study systematically examines how five virtual reality user interface attributes affect the experience of high-functioning individuals with autism spectrum disorder (HFASD). The research addresses a critical gap in VR design guidelines for autistic users, particularly in vocational training contexts. Researchers conducted three controlled experiments with 15 HFASD participants and 15 neurotypical controls, all using a virtual warehouse environment designed to simulate real-world job tasks like locating boxes on shelves. The five attributes tested were: instruction delivery methods (animated, written, pictograph, verbal), visual fidelity (high vs. low detail), view zoom (normal vs. magnified), environmental clutter (present vs. absent), and motion (moving objects vs. static). Each experiment used a within-subjects design with counterbalancing, measuring both objective performance metrics (accuracy, completion time, estimation accuracy) and subjective user experience through surveys and interviews. The virtual warehouse setting was deliberately chosen for ecological validity, as warehouse and stocking jobs are common employment targets for individuals with ASD. The research team drew on prior studies of VR for autism intervention while recognizing that most existing work focused on social skills training rather than user interface optimization.
Key findings
Animated instructions delivered by a virtual character produced significantly better performance for HFASD participants compared to verbal instructions (p=0.004), with success rates of 96% vs. 80%. Written instructions (92%) served as an acceptable alternative, while pictograph instructions performed poorly (86%). For visual fidelity, low-fidelity graphics combined with normal view zoom led to significantly better task performance (p=0.044). Critically, 92% of HFASD participants stated they preferred high visual fidelity—a clear contradiction between stated preference and actual performance. Researchers interpret this as high-fidelity environments being more visually complex and harder to scan for targets. Environmental clutter significantly degraded performance (p=0.017), with 79% of participants preferring clutter-free environments. While clutter made environments feel more familiar and comfortable to participants, it demonstrably impaired their ability to complete tasks. Motion showed no significant effect on scores, though 86% of HFASD participants preferred static environments and found moving objects distracting. Neurotypical participants showed opposite preferences in several areas, preferring low visual fidelity over high and normal view over magnified—highlighting the importance of ASD-specific design guidelines rather than universal assumptions.
Relevance
This research provides concrete, evidence-based design guidelines for VR applications targeting autistic users, particularly in training and educational contexts. The key takeaway for practitioners is the documented gap between user preference and performance: asking users what they want may not reveal what actually works best. For vocational training VR applications, designers should use animated character demonstrations for instructions, low-fidelity graphics, normal (not magnified) views, and minimal environmental clutter. For entertainment or rehabilitation apps where engagement matters more than task efficiency, higher fidelity and some clutter may increase user interest. The study has limitations worth noting: participants were high-functioning adults (mean age 22.73), tasks were warehouse-specific, and results may not generalize to children, lower-functioning individuals, or non-vocational applications. Future research should explore these boundaries. Nevertheless, this remains one of the most systematic empirical studies of VR interface attributes for autistic users, offering practitioners a solid foundation for design decisions.
Tags: autism spectrum disorder · virtual reality · user interface design · vocational training · user experience · cognitive accessibility
Standards referenced: DSM-5