Using Log Data to Analyze the Impact of Adaptive Support on Self-Regulated Learning in Adult Online Education

Successful self-regulated learning (SRL) is essential for online learning, where learners must independently manage their activities with minimal guidance. This is particularly important for adult learners, who often have to balance learning with professional and personal commitments. A pervasive SRL challenge in this context is effective time management, including issues such as procrastination and inconsistent engagement, which highlights the need for support. While adaptive SRL interventions can support learners in navigating self-paced learning, little is known about the mechanisms underlying their effectiveness. A hypothesis from recent work suggests that engaging in learning behaviors in consistent time intervals, as identified through log data, positively influences learning outcomes. We study this hypothesis in a novel context, where \textit{N} = 89 adult learners in an online course were experimentally assigned to one of two conditions, either with or without adaptive time management support. We generate insights into when and with what frequency learners engage in SRL activities inside the course's learning management system (LMS), and how these activities impact their learning outcomes. Results show that adaptive support significantly increased the regularity and frequency of engagement with the LMS. Regularity was marginally associated with improved learning outcomes, consistent with the spacing effect, which predicts better memory retrieval when practice is evenly spaced. Compared to past research, we contribute novel evidence that SRL support can increase regularity in study behavior derived from log data, and that regularity can help adult learners benefit from self-paced online learning.