Using predictive analytics in a self-regulated learning university course to promote student success

Abstract

Prior research offers evidence that differing levels of student engagement are associated with different outcomes in terms of performance. In this study, we investigating the efficacy of a model of behavioural and agentic engagement to predict student performance (low, middle, high) at four timepoints in a semester. The model was significant at all four timepoints. Measures of behavioural and agentic engagement predicted membership across the three groups differently. With a few exceptions, these differences were consistent across timepoints. Looking at variations in student engagement across time can be used to target interventions to support student success at the undergraduate level.