B0664
Title: A competing risk latent variable model for the analysis of university students' careers
Authors: Michela Battauz - University of Udine (Italy) [presenting]
Giuseppe Alfonzetti - University of Udine (Italy)
Ruggero Bellio - University of Udine (Italy)
Abstract: The performance on the exams plays an essential role in the possible outcomes of the university students' careers, which are degree attainment, change of course or dropout. A competing risk model is proposed to analyze such events in time where the explanatory variables include observed covariates, for example, age and grade obtained from high school, and latent variables, which are the student's ability and speed in giving the exams. These latent variables are measured through the grades obtained in the exams and the time needed to pass them using item response theory modelling. In the Italian university system, the students have great flexibility in the choice of the order in which to take their exams resulting in a large variability of the times. Furthermore, both the grades and the times can be censored, since the students might have dropped out or changed course before even attempting some exams. The censoring process is taken into account in the model, which is illustrated through an application to data from the University of Udine (Italy).
