Evaluating the university educational process. A robust approach to the drop-out problem

The use of robust procedures in regression model estimation identifies outlier data that can inform on specific subpopulations. The aim of this study is to analyse the problem of first year dropouts at the University of Florence. A set of administrative data, collected at the moment of enrolment, combined with the information gathered through a specific survey of the students enrolled in the 2001-2002 academic year at the same athenaeum, was used for the purpose. In order to identify the most important variables affecting the students' dropout, the data were first fitted with generalized linear models estimated with classical methods. The same models were then estimated with robust methods that allowed the detection of groups of outliers. These in turn were analysed to determine the personal or contextual characteristics. These results may be relevant for the implementation of academic policy changes.