Beyond technology: Role of technology-organization-environment-human factors in autonomous driving adoption for new energy vehicles

While prior autonomous driving adoption research has predominantly applied traditional models (TAM, UTAUT, TPB) focusing on individual-level factors, this study pioneers the application of the Technology-Organization-Environment-Human (TOEH) framework to reveal previously unidentified asymmetric influence patterns and paradoxical relationships. Unlike existing studies that treat perceived benefits and risks as opposite ends of a continuum, we demonstrate they operate through distinct, parallel cognitive pathways with different antecedents and consequences. Analyzing data from 445 Chinese new energy vehicle users, we uncover three novel findings that challenge established adoption theories: First, we identify a “dual-effect phenomenon” where technology and company factors enhance perceived benefits but fail to mitigate perceived negatives—contradicting assumptions of symmetric influence. Second, we reveal an “infrastructure readiness paradox” where environmental factors counterintuitively increase perceived negatives rather than benefits. Third, only driver factors exhibit the expected bidirectional influence, enhancing benefits while reducing negatives, highlighting the unique importance of human-machine complementarity. Furthermore, we extend beyond intention-based studies by examining actual purchase behavior, discovering that perceived benefits influence purchase only through full mediation by intention, while perceived negatives exert both direct and indirect effects—revealing distinct evaluation pathways for positive versus negative perceptions. Our Importance-Performance Map Analysis provides actionable insight absent from prior research: enhancing perceived benefits yields significantly greater impact than mitigating negatives (importance: 0.152 vs. −0.040). These findings necessitate reconceptualizing adoption frameworks to accommodate asymmetric pathways, paradoxical environmental effects, and dual perception processing, offering strategic guidance for manufacturers and policymakers in accelerating autonomous driving adoption.