Open Conference Systems, CLADAG2023

Font Size: 
A competing risk analysis of academic careers with students' ability and speed as predictors
Michela Battauz

Last modified: 2023-07-01

Abstract


A competing risk model in discrete time is employed to analyze the outcomes of students' academic career, which are degree attainment, drop out or transfer to another course. As covariates, besides using the variables available from the administrative database, we consider also the performace of the students in terms of ability and speed, which are predicted from an IRT model applied to the grades obtained in the exams. An application shows that these variables are good predictors of the outcomes.