Supporting self-regulated learning through generative AI feedback in online higher education: the importance of student perceptions of the source of feedback

Abstract Given the increased autonomy and limited direct instructor presence in online learning settings necessitating advanced self-regulated learning skills (SRLs), many online students face significant learning difficulties, such as low course completion rates. External support, particularly in the form of feedback, is crucial in addressing these difficulties. However, providing timely and personalised feedback at scale remains a key challenge. This study examines the effects of generative artificial intelligence (GenAI) feedback on students' SRLs in an online higher education context. The study compares GenAI feedback, informed by student trace data and learning analytics, with tutor-generated feedback, assessing both student perceptions of feedback and their SRLs development. Employing a mixed-methods research design, we collected survey data, trace-based SRL indicators, and open-ended responses from 46 university students. Quantitative findings revealed that students rated GenAI feedback more positively than tutor-generated feedback, with a statistically significant difference in Genuineness. Moreover, the treatment group showed significant improvement in the Task Strategies dimension of SRLs, suggesting that personalised GenAI feedback may enhance specific regulatory behaviours. Qualitative insights revealed varying student awareness of the feedback source, with some prioritising content over the source of feedback. In contrast, others indicated their attitudes might have differed had they known the provider, highlighting the socio-emotional aspects of feedback adoption and impact. This study concludes that GenAI has the potential to scale SRLs feedback in online learning environments, but the student perceptions of feedback should be carefully managed to observe the expected impact.