Delegators or Self-regulators? Exploring University Students’ Self-Regulated Learning with AI

Abstract

This study examines how university students regulate their learning when employing AI: self-regulated learning with AI (SRL-AI). Specifically, it addresses the following research questions: to what extent do students adopt SRL-AI behaviours?; are there distinct student profiles based on their SRL-AI behaviours?; is students’ SRL-AI associated with the frequency of AI use?; is students’ SRL-AI associated with the timing of seeking AI assistance when encountering study difficulties? To address the research questions, a cross-sectional survey was conducted with university students who reported using AI for academic activities (N = 134). The instruments employed measured the frequency of AI use, the timing of recourse to AI in case of difficulties, and SRL-AI, assessed through a five-item Likert-type scale assessing the overall intensity of five SRL-AI behaviours (goal setting, monitoring, adaptation, evaluation, reflection). On average, the SRL-AI score was close the scale’s midpoint, indicating a moderate frequency SRL-AI behaviours. Cluster analysis on the five scale items identified two balanced profiles: “Higher SRL with AI” or “Self-regulators” (n = 69) and “Lower SRL with AI” or “Delegators” (n = 65), clearly differentiated by their overall level of SRL-AI, with large and consistent differences across all behaviours rather than by specific configurations of behaviours. SRL-AI was weakly but significantly and positively associated with students’ frequency of AI use, whereas the timing of AI recourse showed no significant associations with SRL-AI.