Predictive and statistical analyses for academic advisory support

TL;DR

This paper explores how predictive and statistical analyses can assist academic advisors in identifying student weaknesses and supporting decision-making through data mining techniques, including classification methods like C4.5.

Contribution

It introduces a framework combining statistical tests and data mining for academic advisory support, highlighting the effectiveness of C4.5 classification in student analysis.

Findings

C4.5 achieved the highest accuracy among classifiers.

Statistical analysis revealed differences in course registration behavior.

Predictive models can support timely academic interventions.

Abstract

The ability to recognize weakness of students and solving any problem may confront them in timely fashion is always a target of all educational institutions. This study was designed to explore how can predictive and statistical analysis support the academic work of adviser mainly in analysis progress of students . The sample consisted of a total of 249 undergraduate students: 46 % of them were Female and 54% Male. A one-way analysis of variance and t-test were conducted to analysis if there was different behavior in registering courses. Predictive data mining is used for support adviser in decision making. Several classification techniques with 10-fold Cross validation were applied. Among of them, C 4.5 constitutes the best agreement among the finding results.