Changes in online engagement at the within-person level, profiles, dynamics and association with achievement

While the dominant population-based research is immensely useful in revealing the general trends, it falls short of capturing the intraindividual –or within-person – changes on the individual level. Capturing the individualized processes helps create personalized interventions and advance our understanding of the person in which the learning process takes place. This study aimed at modeling the within-person profiles of change in engagement and charting their temporal unfolding and transitions of 307 students, spanning 12 courses —3684 course enrollments in total—, across three years of education. We use a combination of mixture models, sequence and transition analysis, as well as multilevel repeated measures ANOVA to capture such changes, their longitudinal unfolding and how they relate to performance. Our results revealed three within-person profiles of change (up, down, and average). These profiles were plotted, their sequential unfolding was studied by sequence analysis and then we clustered the longitudinal trajectories of these profiles resulting in three distinct trajectories: an ascending, a descending, and a stable trajectory. We found that the baseline average profile represented an attractor state that was indicative of resilience, stability, and high achievement and students who did below their own average were at risk if they did not catch up. Therefore, our results suggest that tracking the within-person changes in engagement at the individual level opens the door for individualized tracking and intervention that require no constraints of samples or averaging across aggregate data