Empirical investigation of a GPT-4o mini-based tutor for the learning of R programming

This study investigated the impact of an AI-based tutor using GPT-4o mini on R programming task outcomes and subjective evaluations among psychology students at the University of Tübingen. Students were divided into three groups: AI tutor, video tutorial, or both. Confirmatory analyses showed no significant differences in performance points and subjective evaluations across the three groups. Descriptive results and exploratory analyses suggest that our AI tutor improved subjective evaluations of the learning environment without affecting time on task or performance. We discuss the implications of our results for future research on the use of AI in higher education.