Identification of factors that affect the loss of student status using a logit survival model for discrete time data

Abstract

The phenomenon of loss of the student status due to low academic performance is a serious problem that affects the main universities of this country. Low performance is also related in a direct way to student dropout. This is why it is of fundamental importance to identify factors which can be associated with the student’s loss of his/her status due to low academic performance. Based on information about academic performance collected at the National University of Colombia at Medellín, a survival analysis model for discrete time (based on the logit transformation in which the response variable under consideration was the probability for an individual of losing his/her student status for the first time in a determined time) was implemented. This model allowed us to identify and quantify the effect of socio-demographic variables related with the loss of the student status for the first time.