The arcticle "The Impact of Chatbot Familiarity and Frequency of Use on Human-Likeness" by Johannes Göbel, Hans Betke, Johannes Boldt, Mary Linh Tran, Prof. Dr. Guido Schryen in: Proceedings of the Thirty-first Americas Conference on Information Systems (AMCIS 2025), Montréal, Canada, 14-16 August 2025.
Abstract
This study investigates the relationship between user familiarity, frequency of use, and perceptions of human-like qualities in chatbots. Using survey data from 357 participants interacting with three simple chatbots, we examined how familiarity and frequency of use correlate with perceptions of human-likeness. Results revealed modest but statistically significant positive correlations, supporting our hypothesis that increased familiarity and usage frequency lead to stronger anthropomorphic perceptions. Regression analyses showed that familiarity and usage frequency accounted for 6.2% and 5.4% of the variance in perceived human-likeness, respectively. These findings align with Epley's Three-Factor Theory of Anthropomorphism, particularly the roles of elicited agent knowledge and effectance motivation. The study highlights the importance of users’ past experience in shaping anthropomorphic perceptions. Future research should explore additional factors influencing perceived human-likeness in chatbot interactions.
