Academic performance in discrete mathematics: a predictive study

Abstract

This article aims to identify the predictable variables joined the academic performance in Discrete Mathematics, in the university students of the Computer Sciences degree from Havana. For that reason, the multiple linear regression model with cross-sectional data is firstly used to determine the predictor variables of the academic performance. Then, the variables that affect the probability of improving the student's academic performance are estimated through the logistic regression model. The outcomes show that the model satisfies the requirements of sturdiness and meaningfulness in the estimated parameters. It is concluded that the ability to motivate of the study materials and the grades obtained in the mathematics entrance test for Cuban higher education are the two variables that best predict the academic performance in Discrete Mathematics.