An asymmetric logit model to explain the likelihood of success in academic results

Abstract

The aim of this work is to explain the probability of passing first year Mathematics, subject that is included in the Degree of Business Administration and Management at the University of Las Palmas de Gran Canaria. In order to reach this goal, this paper identifies some of the factors that influence this probability measure by using firstly a classic logistic regression model (logit) and, secondly, an asymmetric Bayesian logit model. The obtained results show that the significant variables that affect academic achievement in terms of the relative probability of passing the subject are: regular attendance to lectures and tutorials, student's opinion about material provided by lecturers to prepare the subject, the type of high-school where students took secondary studies and attendance to extra supporting lectures. Future extension of this work consists of analyzing the relationships between some of the explanatory variables considered.