AccessQuest: A Game-Based Approach to Digital Accessibility Education
P D Parthasarathy, Swaroop Joshi · 2026 · ACM Transactions on Computing Education · doi:10.1145/3802826
Summary
Parthasarathy and Joshi present AccessQuest, a single-player, role-playing platformer that teaches five WCAG 2.1 Level A success criteria — text alternatives (1.1.1), captions for pre-recorded content (1.2.2), transcripts for audio-only content (1.2.1), focus order (2.4.3), and on input (3.2.2) — to computing students. Unlike prior serious games for accessibility that either focus on awareness-building or designer-oriented training, AccessQuest positions learners as a software engineer protagonist whose weekly deliverable has been flagged as non-compliant, then uses nostalgic pixel-art scenarios (helping Grandma identify people in a photo, helping Grandpa activate captions, fixing a vending machine's focus order) to teach reasoning about accessibility rather than rote standards recall. The game was built in Unity with a WebGL frontend and a Firebase backend, incorporates lock-and-key progression and a count-up timer for self-monitoring, and was itself audited against WCAG 2.1. This paper reports the first iteration of a Design-Based Research (DBR) cycle and includes a conjecture map linking specific game mechanics (scenario-based challenges, explanatory feedback, gamified progression) to mediating processes (activation of prior misconceptions, analytical reasoning, motivational engagement) and learning outcomes. Evaluation used pre/post tests (10 MCQs), 10 open-ended post-test questions, Likert post-survey, telemetry, and six focus group discussions with 67 sophomore/junior CS and IS students at an Indian public university who had completed at least one web programming course. The researchers deliberately conducted the study off-campus to avoid bias from their own students. Game quality was also assessed externally using the Serious Game Design Assessment (SGDA) framework.
Key findings
MCQ scores rose from a mean of 2.82 (pre) to 7.49 (post) out of 10 (Wilcoxon p < 0.001, rank-biserial r = 0.999) — a very large effect. McNemar's test on each item showed 9 of 10 questions with large effect sizes (Cohen's g > 0.25) and the remaining one still statistically significant. On 10 open-ended application questions graded by an accessibility expert and a computing researcher (inter-rater Cohen's kappa = 0.84), students averaged 6.72/10, significantly above the 5.0 neutral benchmark (p < .001). Visible, perceptible features were mastered most readily — captions questions scored near 0.90 — while abstract, interaction-dependent concepts like focus order (0.515) and on input (0.522) were hardest, consistent with prior findings that workflow-dependent accessibility is harder to internalize than media-level alternatives. Confidence in identifying WCAG violations shifted sharply: Very/Extremely Confident rose from 5/67 to 50/67 post-gameplay; none remained Not Confident. Gameplay telemetry showed a textbook learning curve — Level 1 averaged 6 min 58 sec, dropping to 38 sec in Level 5; 10 participants voluntarily replayed. Time spent had only a weak correlation with learning outcomes (Spearman rho mostly 0.00-0.25), reinforcing that quality of engagement matters more than time on task. Six qualitative themes emerged from FGDs: accessibility as inclusive design, ethical/professional framing, revelations about WCAG principles, value of visuals and nostalgia, usability suggestions (navigation, pause, rewards), and intrinsic replay motivation.
Relevance
For accessibility educators, AccessQuest is a concrete, open-source (accessquest.learndigitalaccessibility.com) intervention that produces measurable, statistically significant knowledge gains in a single 65-minute session — making it viable as a bolt-on module for web development, HCI, or software engineering courses that cannot spare a full unit on accessibility. The conjecture map is a useful artefact: it names the specific mechanisms (misconception activation, analytical reasoning, motivational engagement) that the game targets, which lets instructors and tool builders reason about what to preserve when adapting the design. The finding that students readily learn captions/alt text but struggle with focus order and on input is important for curriculum design — perceivable failures are a gateway, but operable/understandable failures need deeper scaffolding, possibly paired with real code. Reframing accessibility as "doing the right thing" and inclusive design (rather than compliance) through a narrative lens is a practical template for practitioners building internal training that needs to shift culture, not just impart standards. Limitations matter: the study uses computing students (not practicing engineers), the post-test is immediate with no retention follow-up, and the game currently covers only 5 of 30 Level A criteria. The authors plan a follow-up game (Accessibility Bug Hunter) to address transfer to real code, which would complete the pathway from awareness to implementation.
Tags: accessibility education · game-based learning · serious games · WCAG · computing education · gamification · design-based research · role-playing game · higher education
Standards referenced: WCAG 2.1 · WCAG 2.2 · ARIA · POUR · WCAG 1.1.1 · WCAG 1.2.1 · WCAG 1.2.2 · WCAG 2.4.3 · WCAG 3.2.2 · GIGW · IS 17802