Purpose This research aimed to implement artificial intelligence (AI)-supported self-regulated learning (SRL) to improve learning outcomes (LOs). Design/methodology/approach A mixed-methods approach was used. Quantitatively, a quasi-experimental pretest–posttest control group (CG) design was employed. Qualitatively, semi-structured interviews were conducted before and after the intervention. The study involved first-year engineering students in the Department of Automotive Engineering Technology for 2024–2025. Five classes participated, with two classes randomly assigned to the experimental group (EG) using AI supported SRL and the remaining classes assigned to the CG without AI tools. An independent samples t-test was used to analyze quantitative data, while thematic analysis was applied to qualitative data. Findings The results showed that students in the EG using AI-supported SRL achieved higher mean scores than those in the CG without AI tools, and the difference was statistically significant. Thematic analysis revealed that AI-supported SRL functions as a pedagogical agent that scaffolds students across the phases of SRL. Research limitations/implications The study was limited to first-year engineering students, which may affect generalizability; therefore, future research could examine other contexts. From a research perspective, this study contributes to the SRL–AI literature by providing an empirical investigation into the role of AI tools as pedagogical agents that scaffold the SRL process across its distinct phases. Practical implications From a practical perspective, the findings suggest that AI-supported SRL can be effectively integrated into engineering physics courses to support personalized learning (PL), timely feedback, and improved learning outcomes. Social implications AI-supported SRL may reshape teacher–student dynamics by positioning teachers as instructional designers, while promoting collaborative learning and enhancing equity and accessibility through scalable, PL support. Originality/value This study provides empirical evidence that AI-supported SRL can be enacted as a virtual pedagogical agent guiding students through the phases of SRL, rather than merely serving as a content-delivery or automation tool, thereby enhancing LOs.
Thien Van Ngo (Wed,) studied this question.