Social media has become an important channel for contemporary travelers to obtain destination information, seek travel inspiration, and make travel decisions. However, the idealized presentation, information overload, and authenticity confusion in soc...
Social media has become an important channel for contemporary travelers to obtain destination information, seek travel inspiration, and make travel decisions. However, the idealized presentation, information overload, and authenticity confusion in social media travel content may induce users’ travel anxiety and inhibit their travel intention. As a digital service tool that is always available, interactive, and capable of providing immediate responses, chatbots offer new possibilities for alleviating users’ negative emotions during the travel decision-making process. However, chatbot communication style is not a single mode, and different communication styles may exert different effects on users’ psychological perceptions and behavioral tendencies. From this theoretical perspective, the present study develops a research framework linking chatbot communication style, an anxiety-based emotional response to social media travel content, self-efficacy, and travel intention. Specifically, chatbot communication style, including task-oriented and social-oriented styles, serves as the predictor. Travel intention is treated as the outcome, this anxiety response as the emotional pathway, and self-efficacy as the boundary condition. Through three progressive experiments, this study examines how chatbot communication style affects users’ travel intention and further explores how and when this effect occurs. Study 1 employed a text-based scenario experiment involving 140 participants to provide an initial examination of the effect of chatbot communication style on users’ travel intention. Participants were exposed to a chatbot response that varied in communication style, namely task-oriented or social-oriented communication style condition. Users in the social-oriented chatbot condition showed stronger travel intention than those in the task-oriented condition. Study 2 (N = 142) further improved the stimulus materials by using screenshots of chatbot conversation interfaces, making the experimental context more realistic. Study 2 further examined whether an anxiety- based emotional mechanism explains the link between chatbot communication style and travel intention. The findings suggest that the social-oriented chatbot enhanced travel intention through reduced anxiety, and the mediation was supported. Study 3 (N = 232) adopted a 2 × 2 between-subjects design and manipulated self-efficacy through an autobiographical recall task. The results showed that self-efficacy changed both the anxiety-related path and the indirect pathway to travel intention. Specifically, among users with low self-efficacy, the social-oriented chatbot had a stronger effect on increasing travel intention by reducing travel anxiety; among users with high self- efficacy, no clear difference emerged between the two communication styles. Overall, this study extends tourism chatbot research by showing that chatbot communication style can shape travel intention through an anxiety-related emotional pathway and that this process depends on users’ self-efficacy. The findings also provide guidance for designing more adaptive chatbot services on tourism platforms.