Developmentally Aligned Design: A Three-Layer Framework for Neurodiversity Accommodation in AI Research Tools

Abstract

Artificial intelligence tools increasingly mediate research workflows, yet their design frequently excludes neurodivergent researchers through sensory overload, cognitive fragmentation, and structural inaccessibility. Current accessibility frameworks—WCAG 2.1, ISO 9241-210, Ability-Based Design—provide essential but incomplete guidance, emphasizing perceptual access while underserving cognitive and neurodevelopmental needs. This work presents Developmentally Aligned Design (DAD), a three-layer hierarchical framework organizing neurodivergent accessibility through Safety (bio-cognitive harm prevention), Fit (neurotype-specific optimization), and Advocacy (structural agency support). The framework operationalizes requirements through 21 quantitative metrics grounded in psychophysical, cognitive, and clinical research, enabling objective compliance assessment. A mixed-methods case study demonstrates proof-of-concept implementation through a patented wearable device [61] operationalizing the Sensory Sensitivity Mental Health Distractibility (S²MHD) model—a theoretical framework proposing that mental health mediates the relationship between sensory sensitivity and distractibility in autistic adults—deployed with 15 participants over 8 weeks. Results show clinically significant reductions in fatigue (d=0.62) and anxiety (d=0.58), high intervention acceptance (64%), and qualitative evidence of reduced masking and burnout prevention. The framework provides actionable guidance for developers, procurement officers, and policymakers, positioning accessibility as design quality rather than compliance obligation.

Ähnliche Arbeiten