The effect of virtual peer-led learning and students’ academic performance in a STEM course at a research university

Abstract

A sense of belonging, engagement, self-confidence, and adequate college preparation are known predictors of student success, particularly in undergraduate retention and graduation. Yet, bottleneck courses often hinder academic progress. The use of a virtual peer-led study application has shown promise in supporting student achievement through collaborative learning. This pilot study examined whether participation in a virtual peer-led study application reduced DFW (grades of D, F, or course withdrawal) rates in a bottleneck STEM (Sciences, Technology, Mathematics) course, Organic Chemistry. Logistic regression and analysis of variance (ANOVA) results indicated that participants using a virtual peer-led study application had 27% lower odds of receiving a DFW grade than non-participants. Prerequisite Chemistry GPA strongly predicted course outcomes: each one-point increase corresponded to a 67.5% reduction in DFW odds. Conversely, underrepresented minority (URM) and Pell Grant recipient statuses were linked to higher DFW odds—76% and 67% higher, respectively. Findings suggest that peer-led, technology-enhanced learning can strengthen academic performance and foster meaningful peer connections. The results also demonstrate how learning analytics can illuminate students’ capacity to absorb, process, and retain knowledge, reinforcing the theoretical foundation of collaborative learning in challenging STEM courses. Considering the contributions of virtual peer-led learning to improving students’ academic performance, particularly in STEM courses, this study’s most important finding is the confirmation that prior Chemistry GPA plays a fundamental role in predicting students’ DFW outcomes in Organic Chemistry.