Ability Heuristics for Conducting Accessibility Inspections
Claire L. Mitchell, Junhan Kong, Jesse J. Martinez, Shaun K. Kane, Amy J. Ko, Alexis Hiniker, Jacob O. Wobbrock · 2026 · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI '26) · doi:10.1145/3772318.3791504
Summary
Mitchell and colleagues develop and evaluate a set of nine 'ability heuristics' intended to let designers and developers — not just accessibility specialists — inspect interactive technology for accessibility problems. The authors argue that existing approaches have significant gaps: WCAG's 87 success criteria are numerous, web-focused, and daunting for non-experts; Nielsen's usability heuristics do not centre accessibility; and empirical user testing is costly and slow. Drawing on ability-based design, the team ran a structured brainstorming and thematic coding process (informed by two years of classroom ability assumption elicitation assignments) to produce nine heuristics: Adaptability, Equitable Experience, Flexible Task Completion, Efficiency and Effectiveness of User Action, Multiple Modalities, Understandable Messages, Ease of Adoption, Ability Data Transparency, and Help, Support, and Community. The heuristics are deliberately abstract, platform-agnostic, and organised from broad interactive themes to narrower social/ethical concerns. To test them, the authors ran a between-subjects study with 37 HCI and Design master's students at the University of Washington, assigning each student to evaluate three products — Microsoft PowerPoint, Google Maps mobile, and Amazon.com — using one of three methods: ability heuristics (n=14), usability heuristics framed through an accessibility lens (n=12), or WCAG 2.2 guidelines with success criteria stripped (n=11). Students logged issues, severity ratings, heuristic mappings, and NASA-TLX workload.
Key findings
Across 1,103 logged issues, 657 (59.6%) were accessibility-relevant. WCAG produced the highest raw issue count per evaluation (23.83 ± 14.39), ability heuristics were second (18.50 ± 6.76), and usability heuristics lowest (10.18 ± 10.26); a Kruskal-Wallis test detected an overall difference (p=.019) but pairwise comparisons were not significant after Holm correction. The most important qualitative finding was that ability heuristics surfaced far more issues about the *quality* of accessibility features — incomplete or inaccurate alt text generated by AI, paywalled accessibility features, inequitable experiences, missing modality alternatives, burdens shifted to users — whereas WCAG and usability heuristics tended to flag presence/absence of features. Ability heuristics also spread findings more evenly across disability groups rather than defaulting to blind/low-vision concerns, and distributed use more evenly across the nine heuristics (usability heuristics showed a sharp drop after the top one). Severity distributions for ability heuristics and WCAG both peaked at 'Major accessibility problem'; usability heuristics peaked at 'Minor.' NASA-TLX workload was statistically indistinguishable across methods (mental demand ~5.57). Students rated Equitable Experience easiest to apply and Ease of Adoption hardest (due to unfamiliarity with assistive-technology internals). Inter-rater reliability on issue coding reached kappa=0.59.
Relevance
For practitioners, this work offers a lightweight, memorable inspection method that complements rather than replaces WCAG audits or user testing. The nine heuristics push reviewers past binary conformance checks toward questions about feature quality, adaptability, and equity across disability groups — exactly the gaps that cause real-world accessibility to fail despite formal compliance. Teams without in-house accessibility experts could adopt these heuristics for early-stage design reviews, and educators can teach them in a single assignment. Limitations worth noting: the study used design students rather than professional designers or disabled users as evaluators, tested only three mainstream products, and did not validate issues against end-user experience. The 'Ease of Adoption' difficulty also suggests practitioners will still need baseline assistive-technology literacy. With the European Accessibility Act making accessibility competence mandatory across product teams, heuristics like these are a plausible bridge between expert audits and everyday design decisions.
Tags: heuristic evaluation · ability-based design · accessibility inspection · accessibility evaluation · design methods · inspection methods · accessibility quality · design education
Standards referenced: WCAG 2.2 · WCAG · POUR principles · European Accessibility Act · GDPR