Bivariate logistic models for the analysis of the students' University "success"
- Authors: Enea, M; Attanasio, M
- Publication year: 2012
- Type: eedings
- Key words: bivariate ordered logistic models, penalized likelihood
- OA Link: http://hdl.handle.net/10447/63523
Abstract
We analyze the students’ success at University by considering their performance in terms of both “qualitative performance”, measured by their grade average, and “quantitative performance”, measured by University Credits accumulated. To jointly model both marginal and association relationships with covariates, the analysis has been carried out by fitting a bivariate ordered logistic model (BOLM), in a nonparametric fashion, by penalized maximum likelihood estimation. The advantages of such model are in terms of parsimony and parameters interpretation, while preserving goodness-of-fit. The application regards an engineering student (ES) cohort from the University of Palermo.