The Relative Importance of Cognitive and Behavioral Engagement to Task Performance in Self-regulated Learning with an Intelligent Tutoring System

Self-regulated learning (SRL) is essential in promoting students' learning performance, especially in technology-rich environments where learning can be disorienting. Student engagement is closely associated with SRL, although the regulation of engagement in SRL is still underexplored. In this study, we aimed to compare the relative importance of cognitive and behavioral engagement in the three SRL phases (i.e., forethought, performance, self-reflection) to learning performance in the context of clinical reasoning. Specifically, students were tasked to solve two virtual patients in BioWorld, an intelligent tutoring system. We measured student behavioral engagement as their time spent on diagnostic behaviors. Students' cognitive engagement was extracted from their think-aloud protocols as they verbalized their thinking and reasoning process during the tasks. We analyzed the relative importance of cognitive and behavioral engagement in the three SRL phases to diagnostic efficacy. Results suggested that the effects of engagement on student performance depend on task complexity. In the complex task, the six predictors (i.e., two types of engagement in the three SRL phases) explained 36.81% of the overall variances in learner performance. Cognitive engagement in SRL played a more significant role than behavioral engagement in predicting students' performance in clinical reasoning.