This study aims to examine how anthropomorphic design cues in healthcare chatbots, specifically appearance and communication style, influence users' intention to adopt health information through perceived competence and perceived warmth in low-to moderate-risk health contexts.
Drawing on the persuasion knowledge model, elaboration likelihood model, stereotype content model and computers are social actors, we develop a research model and conduct two online experiments. Study 1 adopts a 2 × 2 between-subjects design with 352 participants to examine the effects of anthropomorphic appearance (with vs without human names and avatars) and communication style (task-oriented vs social-oriented) on users' adoption intention through perceived competence and warmth. Study 2 employs a 2 × 2 × 2 factorial design with 473 participants to validate the findings and examine the moderating roles of persuasion knowledge and health literacy.
The results indicate that anthropomorphic appearance enhances perceived competence and warmth. A task-oriented communication style increases perceived competence, whereas a social-oriented style enhances perceived warmth. Both perceived competence and warmth significantly increase users' adoption intention and mediate the effects of anthropomorphic appearance and communication style on adoption intention. Additionally, persuasion knowledge weakens the effects of anthropomorphic appearance but does not moderate the relationship between communication style and perceived warmth. Health literacy negatively moderates the relationship between perceived warmth and adoption intention, while positively moderating that between perceived competence and adoption intention.
This study extends existing research by simultaneously examining anthropomorphic appearance and communication style in healthcare chatbots and uncovering the underlying psychological mechanisms through which these cues influence users' adoption intention.
The peer review history for this article is available at: Link to the website.
