The Roles of Social Perception and AI Anxiety in Individuals' Attitudes Toward ChatGPT in Education

Chengcheng Wang,Xing Li,Zheng Liang,Yingying Sheng,Qingbai Zhao,Shi Chen

DOI: https://doi.org/10.1080/10447318.2024.2365453

IF: 4.92

2024-06-23

International Journal of Human-Computer Interaction

Abstract:Public attitudes are essential for technology promotion and policy formulation. The present study investigated the Chinese public's knowledge of ChatGPT, as well as examined the roles played by Big-Five personality traits, social perception, and AI anxiety in shaping the public's attitudes toward ChatGPT using the questionnaire method. Results showed that: (1) Nearly, 1/3 of teachers surveyed did not know ChatGPT at all, and all of them were primary and secondary school teachers. (2) The level of knowledge about ChatGPT was significantly related to gender, educational level, teaching stage (in teacher samples), and major (in student samples). (3) The public's positive attitude is higher significantly than the negative attitude. (4) Social perception positively predicted positive attitudes and negatively predicted negative attitudes. Moreover, a notably higher predictive power for positive attitudes compared to negative attitudes was demonstrated by social perception. There is an equally predictive effect of competence perception and warmth perception on attitudes, without any domain effect observed. (5) AI anxiety only positively predicted negative attitudes but did not impact positive attitudes. In explaining negative attitudes, AI anxiety exhibited a higher explanatory power compared to the Big-Five personality traits, primarily correlating with neuroticism. The findings indicate that it's inappropriate to consider attitude evaluation toward AI as a single dimension. There are relatively independent components for positive and negative attitudes toward AI. The roles played by other predictive variables in attitudes are discussed.

computer science, cybernetics,ergonomics