A Qualitative Study of ChatGPT's Feedback for Psychological Case Conceptualization

This project explores the use of ChatGPT-4 as a tool for psychological case conceptualization based on the LIBET model. Using semi-structured interviews with non-clinical participants, the study investigates the model’s ability to identify life themes and semi-adaptive plans through AI-generated feedback. Thematic analysis was employed to evaluate interpretative coherence, reasoning quality, and linguistic strategies. Supplementary materials include prompt design documentation (for both the MyGPT custom phase and the prompting structure used for transcript elaboration), and inter-rater agreement data. This study contributes to the growing field of responsible AI integration in mental health research and practice.