Understanding Clinician Experiences with Game-Based Interventions for Autistic Children to Inform a Future Game Platform Focused on Improving Motor Skills
Hunter M. Beach, Devin Jay D. San Nicolas, Carly Miller, Cathy Ly, Jared Duval · 2026 · Extended Abstracts of the 2026 CHI Conference on Human Factors in Computing Systems (CHI EA ’26) · doi:10.1145/3772363.3799014
Summary
This CHI 2026 Extended Abstract reports a two-phase qualitative study with pediatric physical therapists (PTs) and occupational therapists (OTs) to inform the design of AutMotion Studio, a speculative modular platform of motor-skill minigames for autistic children. Up to 87% of autistic individuals experience motor challenges, and while serious games for health (SG4H) can produce clinically meaningful results in motivating contexts, the authors argue that current solutions are too rigid for clinical practice. The work is positioned as Research through Design and centres play as integral to pediatric therapy, child development, and wellbeing. In Phase 1, the team conducted semi-structured Zoom interviews with nine clinicians (4 OTs, 5 PTs) about their use of game and play-based interventions, then applied inductive thematic analysis to surface eight themes. In Phase 2, four 60-minute participatory design workshops with five clinicians (two returning, three new) used warm-up activities, bodystorming, Crazy Eights ideation, embodied sketching, and Figma prototyping under a Situated Play Design structure. The brief was to prototype games that secretly use a game-controller input paradigm so a therapist can guide motor exercises at home and in clinic. The paper uses identity-first language for autism, person-first for broader disability populations, and the term "support needs" rather than "level of function," and explicitly positions further co-design with autistic children and families as essential future work.
Key findings
Eight interdependent themes emerged as foundational requirements for clinically-adoptable motor games: (1) stepwise task breakdown so therapists can decompose complex skills like running into balance, strength, and pattern components; (2) flexibility for diverse support needs, with rigidly graded games causing disengagement (P04 abandoned a handwriting game that withheld credit unless letters were "perfect"); (3) therapeutic specificity, since entertainment titles such as Wii boxing let children chase high scores without producing the trunk rotation, core activation, or elbow extension therapists target; (4) accessibility and user-friendliness, given uneven home internet, space, and console access post-COVID; (5) sensory customization to prevent over-stimulation; (6) lesson carryover beyond the clinic; (7) child autonomy through choice (even a forced two-option pick boosts engagement); and (8) social engagement that reframes therapy as a shared experience. Adoption barriers were rigid automated assessment, technological complexity, and misalignment between in-game and therapeutic success. The resulting AutMotion Studio speculative design uses therapist-built activity "maps," a three-button Wizard-of-Oz feedback system (approve / partial approve / retry), phone-to-TV casting, QR-linked exercise videos, sensory adjustment controls, and a caregiver-mediated at-home mode.
Relevance
This paper is directly useful to anyone designing health, rehabilitation, or educational technology for autistic users or children with motor differences. Its core argument inverts a common HCI assumption: rather than asking "how do we automate feedback?" designers should ask "how do we augment clinician judgment?" The authors reframe Wizard of Oz from a prototyping shortcut into a permanent interaction paradigm, treating evaluative ambiguity as a feature that lets therapists count imperfect motor output as progress. Practitioners building serious games, exergames, or telerehabilitation tools should treat the eight themes as a checklist and resist single-standard automated scoring. Limitations are significant: the sample is small (12 clinicians across both phases) and U.S.-based, no autistic children or families participated, and AutMotion Studio is speculative rather than implemented or evaluated. Still, the design particulars and the explicit critique of automation-first SG4H make this a useful reference for accessibility teams negotiating between flexibility and measurement in clinical contexts.
Tags: autism · motor skills · serious games for health · pediatric therapy · physical therapy · occupational therapy · participatory design · speculative design · wizard of oz · co-design