What drives people to use automated vehicles? A meta-analytic review

Lack of consumer acceptance is a prominent barrier to the large-scale adoption of automated vehicles (AVs). This study investigated the underlying mechanisms for AV acceptance and how the mechanisms differed across subgroups by reviewing and synthesizing existing literature. We proposed AV acceptance models by extending the basic Technology Acceptance Model (TAM) with trust and perceived risk factors. Data from 36 studies were extracted to fit the models using meta-analytic structural equation modeling technique. The results suggested that trust contributed most in determining AV acceptance, followed by perceived usefulness and perceived risk, and perceived ease of use makes the least contribution. The subgroup analyses showed that the model parameters differed across the levels of three variables, i.e., sample origin (Europe/Asia/America), automation level (full/partial), and age (young/middle-aged). Specifically, trust was unanimously identified as the most important determinant of AV acceptance across all subgroups. Perceived risk only remained significant in America, fully AVs, and middle-aged subgroups. Perceived ease of use was insignificant in the above-mentioned three subgroups while remained significant in the rest subgroups. Building trust could be the most useful and universal way to improve AV acceptance, and policy makers should consider the characteristics of consumers when making AV promotion strategies.