GenAI Use Behavior and Post‐Failure Perceptions Among People With Functional Disabilities: A Multimethod Study

GenAI Use Behavior and Post‐Failure Perceptions Among People With Functional Disabilities: A Multimethod Study

Summary

This multimethod paper investigates how people with functional disabilities (hearing, visual and mobility impairments) adopt and respond to generative AI (GenAI). The authors combine a two-part survey study (Study 1) and a controlled experiment (Study 2). Study 1 uses a genetic algorithm to select predictors and Bayesian linear regression plus IPWRA to test them; Study 2 uses a between-subjects vignette (GenAI vs human agent) to measure post-failure inferential judgement.

Key findings: habit, promotional benefits, trust and behavioural intention positively predict GenAI use among people with functional disabilities. Perceived value moderates the intention→use relationship in a counterintuitive way: higher perceived value can weaken the effect of intention on actual use. In the experiment, failures attributed to GenAI reduce willingness to use services more than human errors, but positive attitudes and greater prior GenAI use buffer that effect. The paper draws practical recommendations for designers, marketers and policymakers to improve inclusive GenAI adoption, error-handling and onboarding.

Key Points

• Study sample focused on people with functional disabilities (hearing, visual, mobility) across multiple European nationalities.
• Study 1 (n=284) identified habit, promotional benefits, perceived trust and behavioural intention as positive predictors of GenAI use behaviour using a genetic algorithm and Bayesian regression.
• Perceived value moderates the intention→use link: unexpectedly, higher perceived value can reduce the translation of intention into actual use for this group.
• Study 2 (n=231) experimental vignette: participants reported lower inferential judgement (less willingness to reuse) after GenAI-attributed failures than after human errors.
• Attitudes toward GenAI and prior GenAI use moderate post-failure reactions — positive attitudes and frequent use increase tolerance of GenAI failures.
• Findings emphasise PwDs as active agents (not passive recipients) whose prior experience and expectations shape adoption and post-failure perceptions.
• Managerial implications: accessible onboarding to build habit, targeted promotional support, transparent governance and inclusive error messaging are recommended.
• Limitations: focus on three disability types, cross-sectional survey and controlled vignettes — longitudinal and more diverse disability samples are suggested for future work.

Why should I read this?

Short version: if you care about making AI actually useful and fair for people with disabilities, read it. This study flips a few easy assumptions — e.g. higher perceived value doesn’t always equal more use — and shows why inclusive design, plain-error messaging and real onboarding matter. We’ve done the slog so you don’t have to: the paper pinpoints what to fix in GenAI products to stop losing users who most need them.

Author style

Punchy and focused — the authors treat people with functional disabilities as active, experienced users and bring together acceptance theory, behavioural reasoning and attribution theory to make recommendations that matter to product teams and policymakers. If accessibility is part of your roadmap, their practical takeaways (on habit-building, promotion, trust and failure recovery) are worth acting on.

Source

Source: https://onlinelibrary.wiley.com/doi/10.1002/mar.70028?af=R