Assessing how socioemotional and social communication shapes chatbot perceived competencies
(A2025-124339)

AUTHORS

Álvaro Saavedra, Public University of Navarre; Monica Cortinas, Public University of Navarre; Natalia Rubio, Autónoma University of Madrid Q2818013A; Sandra Loureiro, ISCTE - Instituto universitario de lisboa

ABSTRACT

This study investigates how chatbot language styles— particularly social vs. socioemotional communication—affect their cognitive, social, and emotional competencies, ultimately shaping the global competence of the chatbot. Utilizing the Stereotype Content Model, we examine how each language style forms a users’ stereotype of the chatbot's competencies. Our research employs a detailed assessment of linguistic styles to understand how chatbots can enhance user interactions by incorporating social and emotional features into their responses. Different preliminary studies have been conducted to facilitate an experimental one-factor between-subjects design in which chatbot language was manipulated. Results suggest that while both social and socioemotional language styles significantly enhance perceived competencies, they do so in distinct ways that inform the development of more effective chatbot communication strategies. This study highlights the value of customizing chatbot language to improve user interactions, offering marketers practical strategies for using chatbots to build better customer relationships and increase engagement.