Abstract:

Early student risk detection and amelioration continue to challenge the management education community. Traditionally, student performance via examinations and characteristics/demographics are two factors that have been used in detecting students at risk. However, waiting until midterm exam results to intervene can often prove problematic. With the advent of learning management systems (LMS), these traditional factors can now be complemented by a variety quantitative and qualitative metrics, including practice quiz performance and on-task engagement. The goal of identifying students at academic risk early in the matriculation process is not new. Specifically, Seidman proposed in the late 1990s the following relationship, which links student retention to both early detection and continuous intervention: Student Retention = Early Detection + Continuous Intervention. This construct underscores that early detection of students at risk as well as continuous intervention can be key to student retention. Recent data suggest that machine learning-based predictive models can provide automatic and timely student assessments, which allow for both the detection of students at risk and the implementation of appropriate intervention initiatives. The empirical results from this study, which included both student demographic and performance data, support these findings and further suggest that machine learning can be used to also design customized intervention strategies. The primary purpose of this presentation is to highlight how machine learning can reduce student dropout rates and improve overall learning outcomes throughout the management education community of practice.

Keywords:

Machine learning, management education, student risk detection, intervention strategies.