Beyond the Manual: Mapping Peer-Generated Content about Wheelchair Care and Adaptation on YouTube
Wen Mo, Aneesha Singh, Lan Xiao, Catherine Holloway · 2026 · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI '26) · doi:10.1145/3772318.3791906
Summary
This CHI 2026 paper presents a content analysis of 290 YouTube videos (from 194 unique channels, 8,629 initial results narrowed via inclusion/exclusion) and 800 sampled comments from a total 10,781 harvested, examining how wheelchair users and their caregivers document, share, and exchange practical knowledge about wheelchair repair, maintenance, and customization. The authors applied hybrid top-down/bottom-up thematic coding (Cohen's Kappa 0.67-0.97) and fitted negative binomial regression models for viewcount, likecount, and commentcount, controlling for video duration, channel subscribers, and recency. The corpus is English-language only, 70% manual and 25% power wheelchairs, 88% presented by wheelchair users themselves, published primarily 2017-2025. Videos distributed across seven topics — Repair (29%), Build & Modification (23%), Maintenance (17%), Costume & Everyday Styling (12%), Custom-Built Wheelchair (11%), Accessory Hacks & Add-Ons (11%), Adjustment & Setup (5%) — and five delivery styles (Tutorial 46%, Overview/Walkthrough 20%, Process Vlog 16%, Demo 13%, Showcase 4%). The authors derive a 'Ladder of DIY Adaptation' (Accessorizing → Bricolage → Light Fabrication → Major Modification → Full Custom Build) and a three-layered model of viewer comments (inner direct reactions, semi-periphery personal reflections, outer tangential social context).
Key findings
Nine user needs emerged: affordable products and service ('beat the system'), acquiring caring skills, custom-built chair solutions, daily-task support (carrying/storage n=26 solutions), safety and well-being (n=36), enhanced mobility and performance (maneuverability n=8, off-road conversions n=16, DIY power assist n=8, wheelchair-from-scratch n=8), accessible leisure (n=7), and self-expression (n=38). Engagement analysis showed a sharp split: Costume & Everyday Styling was a 'blockbuster' topic with ~10x baseline views (IRR 9.92, p<.01), while technical topics were severely disengaged — Maintenance IRR 0.10 views, Repair IRR 0.22, Build IRR 0.05 for likes. Custom-Built Wheelchair videos became community hubs with ~2.77x baseline comment counts (p<.01), concentrating deep Q&A and emotional support despite modest view counts. Comment analysis revealed both solidarity (peer validation, shared struggle, memorial threads for deceased creators) and visible ableist hostility ('if you can walk, why you need wheelchairs'), alongside 27 videos still using derogatory terms like 'cripple' or 'handicap' — with no decline over time 2011-2025. Repair dominance (29%) signals official manuals failing daily needs; power wheelchair repair/build videos were over-represented relative to baseline, indicating larger service gaps for power mobility.
Relevance
For AT designers and accessibility researchers, this paper is a living archive of what wheelchair users actually need but cannot get from formal channels, and a method template for mining YouTube as a 'design material' for any AT domain (hearing aids, prosthetics, white canes, AAC). Concrete takeaways: (1) design for repairability and modular customization, not just feature completeness — users adapt every bolt; (2) treat bricolage (duct tape, pool noodles, PVC) as a legitimate rung of making, not a failure mode, and build platforms that support community-vetted safe modifications; (3) recognize styling and self-expression as first-class AT functions with the broadest audience reach; (4) design platforms for solidarity — credits, multilingual captioning, moderation against ableist hostility, interdependence-oriented features. The Ladder of DIY Adaptation and three-layer comment model are transferable analytical tools. Limitations: English-only videos, United States and United Kingdom dominant, participation bias toward digitally-literate users, YouTube only (no TikTok, Reddit, or Instagram), and performative self-presentation effects. Future work should apply the method cross-platform and cross-language.
Tags: wheelchair · DIY assistive technology · YouTube · content analysis · peer support · bricolage · accessibility · thematic analysis