Predicting student burnout in blended environments: a complementary PLS-SEM and machine learning approach

Abstract

ABSTRACT Despite student burnout has attracted the attention of researchers and practitioners in education and psychology, little is known about the factors that contribute to it in blended contexts. This research aims to formulate and verify a predictive model elucidating how social support and self-regulated learning predict student burnout within blended learning contexts. Utilizing data collected from a sample of 303 students, a complementary method that combines partial least squares structural equation modeling (PLS-SEM) with machine learning (ML) algorithms was implemented. The favorable impact of perceived social support and self-regulated learning in mitigating burnout was demonstrated by the PLS results. Additionally, it was discovered that self-regulated learning fully mediates the correlation between offline social support students received and learning burnout. Furthermore, the employed ML algorithms achieved a prediction accuracy rate exceeding 70% in the majority of cases. Employing a complementary analytical method is believed to offer a substantial contribution to the current body of research on learning burnout generally and blended learning in particular.