Developing a formative scale to measure consumers' trust toward interaction with artificially intelligent (AI) social robots in service delivery

This study develops and validates a scale of Social Service Robot Interaction Trust (SSRIT) that measures consumers' trust toward interaction with AI social robots in service delivery. Through a systematic literature review, semi-structured interviews, a focus group study, and rigorous quantitative studies, this study conceptualizes the interaction-based trust and proposes a third-order reflective-formative scale, which suggests that trust in interaction is measured by 3 s-order indicators: propensity to trust in robot, trustworthy robot function and design, and trustworthy service task and context. Propensity to trust in robot is predicted by familiarity, robot use self efficacy, social influence, technology attachment, and trust stance in technology. Trustworthy robot function and design is formed by anthropomorphism, robot performance, and effort expectancy. Trustworthy service task and context is determined by perceived service risk, robot-service fit, and facilitating robot-use condition. The convergent, discriminant, external, concurrent, and predictive validities of the scale are validated. Theoretical contributions and managerial implications are provided.