Predicting student academic performance in Machine elements course

Abstract

The students frequently regard the fundamental mechanical engineering courses as demanding, and they often have high drop-out rates. Due to the width of prerequisite knowledge which needs to be integrated, advanced, and applied, Machine elements is one of such courses. In this article, the authors have used mathematical methods to determine the predictors of student performance in Machine elements course held at the Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb. The secondary education data and grades of preceding courses were collected for 729 students enrolled in Machine elements course. The obtained data were described using basic statistical methods and further used to develop models for predicting the students’ performance on the Machine elements course. Building on the results, the authors have answered three research questions: The preceding courses are better predictors when compared to secondary education (1). The Strength of Materials and Mathematics II were the best predictors; generally, the course’s complexity, rather than its scope, was an indicator of its importance for the prediction of student’s future success (2). Lastly, it was possible to group the students based on predicted future academic performance which, consequently, enables early segmentation and detection of students at risk (3).