Cognitive and behavioral determinants of ChatGPT adoption in education: an integrated theoretical approach

As technology rapidly advances, the integration of Artificial Intelligence (AI) in education is becoming increasingly widespread. This study explores the factors that influence university and graduate students' intentions to use ChatGPT, an AI-powered system, by integrating Cognitive Load Theory and the Theory of Planned Behavior. This dual-theoretical approach is relatively novel, offering a comprehensive understanding of how different cognitive load types—intrinsic, extraneous, and germane—affect students' perceived behavioral control, behavioral attitudes, and behavioral intention toward ChatGPT. We surveyed 573 participants, obtaining 518 valid responses. The findings indicate that higher intrinsic cognitive load reduces students' behavioral intentions to use ChatGPT, while perceived behavioral control significantly moderates this relationship, suggesting that students' self-efficacy can mitigate the negative impact of high cognitive load. Additionally, germane cognitive load positively influences students' behavioral attitudes toward ChatGPT, which, in turn, enhances their behavioral intention. However, extraneous cognitive load does not have a significant effect on behavioral intention. Given the growing integration of AI in education, this study offers timely insights for educators and developers on optimizing cognitive load to improve the acceptance and effectiveness of AI-based educational systems.