Human vs. machine-like representation in chatbot mental health counseling: the serial mediation of psychological distance and trust on compliance intention.

Abstract

This study examined a serial mediation mechanism to test the effect of chatbots' human representation on the intention to comply with health recommendations through psychological distance and trust towards the chatbot counselor. The sample of the study comprised 385 adults from the USA. Two artificial intelligence chatbots either with human or machine-like representation were developed. Participants had a short conversation with either of the chatbots to simulate an online mental health counseling session and reported their experience in an online survey. The results showed that participants in the human representation condition reported a higher intention to comply with chatbot-generated mental health recommendations than those in the machine-like representation condition. Furthermore, the results supported that both psychological distance and perceived trust towards the chatbot mediated the relationship between human representation and compliance intention, respectively. The serial mediation through psychological distance and trust in the relationship between human representation and compliance intention was also supported. These findings provide practical guidance for healthcare chatbot developers and theoretical implications for human-computer interaction research.