Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
A concomitant variable mixture model for predicting freshmen gained credits
Leonardo Grilli, Carla Rampichini, Roberta Varriale

Last modified: 2013-06-16

Abstract


This work is motivated by the challenging task of modelling the number of credits gained by freshmen during the first year at the School of Economics of the University of Florence. The observed distribution of credits is multi-modal (with the highest frequency at zero credits, about 23%), hindering the use of fully parametric models. A viable alternative is represented by mixture models. Given that the number of gained credits is limited both below and above, we rely on a binomial mixture model. The available covariates (student background characteristics and scores from a pre-enrolment test) are entered as predictors of the mixture components.

Full Text: PDF