Construction of Large‑Scale Digital Accessibility Evaluation Indicator System for the Elderly and People with Disabilities

Digital services have been deeply integrated into public life, and the problem of digital exclusion among the elderly and people with disabilities has become increasingly prominent. Existing accessibility evaluations mostly rely on WCAG specifications, focusing on technical compliance detection for sensory‑related barriers, while failing to sufficiently cover local scenarios involving age‑related cognitive decline and multiple‑disability intersections. Moreover, high costs of manual assessment hinder large‑scale screening. Based on the capability‑demand matching theory and universal design theory, this paper constructs an initial indicator pool through literature coding and qualitative user analysis. The Delphi‑CRITIC‑grey relational combined analysis is adopted for indicator screening and subjective‑objective weight assignment, establishing a digital accessibility evaluation indicator system consisting of 5 primary‑level, 18 secondary‑level, and 62 tertiary‑level quantitative observation indicators (Martins, B., & Duarte, C., 2024). Reliability, validity and discrimination tests are completed. Empirical evaluation is conducted on 127 high‑frequency domestic public digital services, yielding an average composite sample score of 58.72 (W3C Web Accessibility Initiative, 2012). Cognitive comprehensibility, fault tolerance and security guarantee constitute major weak points. Traditional automated WCAG tools overestimate the real‑world accessibility level for composite groups of the elderly and people with disabilities. Supporting a hybrid evaluation mode combining automated parsing and sampling‑based manual review, this system can provide quantitative evidence for optimization of digitally accessible products and the formulation of industrial evaluation specifications.