The Impact of Student Behavior Characteristics on Academic Performance in Blended Learning Environments at Chinese Private Universities: A Structural Equation Analysis

The continuous optimization and development of online learning, along with the ongoing reflection on traditional teaching methods, have contributed to the popularity of blended learning (BL) as a favored approach to teaching and learning in higher education. However, systematic research on how students’ learning behavior characteristics influence academic performance (AP) in BL contexts remains limited, particularly regarding the mediating role of learning commitment (LC) and moderating effects of demographic factors. This study addresses this gap by investigating the direct and indirect pathways through which behavioral engagement (BE), self-directed learning (SL), and technical application ability (TAA) affect AP. Drawing on data from 1,870 undergraduates across five Chinese private universities（collected from September to October 2024, we employed structural equation modeling (SEM) to analyze the interplay between these variables. Results reveal that SL (β = 0.544, p < 0.01) and TAA (β = 0.228, p < 0.01) directly enhance AP, whereas BE exerts no significant direct effect. Crucially, LC fully mediates the relationships of BE and TAA with AP (100% mediation) and partially mediates the SL-AP link (38.4%). Contrary to expectations, demographic variables (gender, grade, discipline) showed no moderating effects. These findings enrich BL theory by elucidating LC’s central role as a mediator and challenging assumptions about demographic influences. Practically, the study advocates for curriculum designs that prioritize SL and TAA development while leveraging LC through personalized feedback and adaptive learning technologies. This research provides actionable insights for educators and policymakers to optimize BL effectiveness in diverse institutional settings.