Anthropomorphism in Chatbot Systems Between Gender and Individual Differences

Zhida Sun, Manuele Reani, Yunzhong Luo, Zhuolan Bao

Abstract: Online advisory chatbots are becoming increasingly popular and are often used to support user decision-making in different sectors. The design of such systems has been the object of investigation since it can alter the user’s perception and cognition. This study investigates the effect of chatbot design, gender, and individual tendencies on perceived anthropomorphism in a financial investment scenario. Our online experiment with 116 participants revealed that manipulating chatbot design can alter the user perception of anthropomorphism contingent on both gender and anthropomorphism tendencies. Male users with a strong anthropomorphism tendency exhibited a higher degree of perceived anthropomorphism compared to female users with a similar tendency. However, this effect was only observed when the chatbots themselves lacked design elements that promote anthropomorphism. Our novel finding provides theoretical and practical implications to the domain of chatbot system design and applications.