2025/02/27 by Plagge, Eileen, Jucks, Regina
#Cognition and Perception #Education #FOS: Psychology #Psychology #Social Psychology #Social and Behavioral Sciences #anthropomorphism #artificial intelligence #higher education #warmth
paper · doi:10.17605/osf.io/vsbwh
The integration of chatbots such as OpenAI's ChatGPT into everyday interactions has redefined how humans engage with technology, particularly in education. In our initial study, a 2x2 experimental design with 307 university students investigated the perceived warmth and benevolence of a chatbot described as a tool to assist with study-related questions. The experimental factors included descriptions of the chatbot that manipulated kindness (high vs. low) and anthropomorphism (technical vs. anthropomorphic). While kindness significantly increased perceived warmth and benevolence, no effects were found for anthropomorphism, nor were there any interaction effects. Building on these findings, the current study aims to replicate the warmth effect and further explore the role of anthropomorphism in shaping user perceptions. Specifically, the study focuses on qualitative responses from students to gain deeper insights into how anthropomorphic descriptions influence their interaction with the chatbot. By emphasizing qualitative analysis, this study seeks to refine our understanding of how AI design influences warmth perceptions and anthropomorphism, with implications for enhancing trust and engagement in educational contexts.