Analyzing Graduate Students ’ Open-Book Exam Experiences Before and After ChatGPT Based on Activity Theory

This study explored the impact of ChatG PT on the identity formation of graduate students through a qualitative case study of their open-book examination experiences. Using activity theory as a theoretical framework, semi-structured interviews were conducted with six graduate students who had taken op en-book exams before and after the introduction of ChatGPT. The interviews were analyzed to examine changes in the activity system, contradictions that emerged, and strategies for overcoming those contradictions. The findings reveal that the introduction of ChatGPT led to significant changes in the examination experience, such as reducing cognitive demand by delegating simple tasks and enhancing the quality of answers through collaboration with ChatGPT . However, students faced challenges related to fairness, time constraints, and uncertainty about their learning trajectory. Notably, the graduate students began to question their identity as researchers, particularly due to the inaccuracy and incomplete domain-specific knowledge provided by ChatGPT. To address these challenges, students developed new prompting skills and adopted the role of critical reviewers of ChatGPT ’ s output, though issues related to examination rules and their identity as researchers remained unresolved. This study discusses the implications of these findings for the direction of assessments in higher education as well as the evolving epistemological beliefs and researcher identities of graduate students in the age of AI.