The Interplay between Cognitive Load and Self-Regulated Learning in a Technology-Rich Learning Environment

Cognitive load can be induced by both learning tasks and self-regulated learning (SRL) activities, which compete for limited working memory capacity. However, there is little research on the relationship between cognitive load and SRL. This study explored how cognitive load interplayed with SRL behaviors and their joint effects on task performance (i.e., diagnostic efficiency) in the context of clinical reasoning. Specifically, twenty-seven (N = 27) medical students diagnosed three virtual patient cases in BioWorld, a simulation-based learning environment to improve medical students' clinical reasoning skills. Students' SRL behaviors were automatically recorded in BioWorld log files as they accomplished the tasks. We employed text mining techniques to extract four linguistic features from students' concurrent think-aloud, i.e., "cognitive discrepancy," "insight," "causation," and "positive emotions," which were further used to represent students' cognitive load. The latent profile analysis was then performed to cluster students into high- and low-load group. We also conducted a path analysis to investigate the mediation roles of SRL behaviors in the relationship between cognitive load and diagnostic efficiency (task performance). The results revealed that cognitive load negatively affected diagnostic efficiency, mediated by the ratio of SRL behaviors in the self-reflection phase. This study provides theoretical and methodological insights regarding the measurement of cognitive load and its interplay with SRL. This study informs the design of effective interventions for managing cognitive load in SRL within intelligent tutoring systems.