The student talent in a random effects Quantile Regression Model for university performance

Abstract

This paper is a development of a previous work on performance of Italian university students. A Quantile Regression was carried out on a proposed performance indicator, based on a transformation of the median of the marks weighted by credits. Results suggested to investigate the role of students peculiar features on their performance, measured by the transformation of the weighted marks. Therefore, a random intercept Quantile Regression model is fitted on real data concerning graduates over the legal university duration set in Italy, enrolled in 2002 in two Degree Courses of the University of Palermo. Results show that the student performance seems to be influenced by the student's aptitudes, motivation, interest or more in general the students talent, rather than socio-demographic characteristics.