Vicarious Interaction in Online Health Consultation Service: The Effects of Generative AI’s Anthropomorphism and Social Support on Intended Responses Through Social Presence and Source Credibility

Abstract

Drawing upon vicarious interaction and social support theory, we proposed that online users can acquire benefits from observing AI-human conversations. We examined how anthropomorphism influences behavioral intention through social presence and source credibility when delivering different types of social support from observation in online health consultation services. We recruited 306 individuals with insomnia to participate in a 2 (anthropomorphic GAI vicarious interaction vs. non-anthropomorphic GAI vicarious interaction) by 2 (emotional support vs. informational support) between-subjects factorial experiment. The results revealed that social presence and source credibility mediated the effects of anthropomorphism on behavioral intention. Compared with viewing a non-anthropomorphic GAI during vicarious interaction, viewing an anthropomorphic GAI lead to a higher perceived social presence and more source credibility, ultimately increasing intention to perform recommended behaviors against insomnia. When social support is informational support, vicarious interaction with anthropomorphic GAI significantly enhanced social presence. Theoretical and practical implication are discussed as well.