Modeling acceptance of shared and private autonomous vehicles with social-media-informed psychological variables

Autonomous vehicles (AVs) are emerging rapidly in the AI era, and many countries are actively advancing AV pilot tests and regulations in metropolitan cities. Despite numerous studies exploring factors influencing public acceptance of AVs, few have proposed a comprehensive framework that integrates social media insights into survey design. This study develops a modeling framework to examine the joint effects of psychological and observed factors on AV acceptance using social-media-informed psychological variables. Comments on Douyin videos are collected to capture public perceptions of AVs, and a survey is subsequently designed and deployed in Shenzhen, China. Structural equation modeling is used to identify five latent factors (e.g., task–technology fit, attitudes, technology adoption propensity), followed by a discrete choice model to examine acceptance of private and shared AVs. Results show that technology adoption propensity, trust in AV performance, and attitudes directly affect acceptance, while trust in AV institution and technology adoption propensity indirectly influence acceptance through attitudes and trust in AV performance. AV exposure emerges as a significant antecedent of both perceptions and acceptance. We further distinguish potential users of private and shared AVs based on their socio-demographic characteristics and travel behaviors. The findings offer theoretical contributions and practical implications for urban planners, policymakers, and AV operators, particularly in tech-forward urban environments with rapid AV deployment.