The evaluation of University educational processes: a quantile regression approach

Cristina Davino, Domenico Vistocco
(2007) Statistica, Vol. 3, 281-292, ISSN 0390-590X

DOI  pdf Online

Nowadays, the evaluation of educational processes receives great attention by Universities because of its crucial role in the quality certification. In order to be- come worthy of a certification, Universities must be able to monitor each phase of the educational process and to measure the achieved results (CRUI, 2003). This evaluation can deal both with the efficiency and with the effectiveness of the process (Lockheed e Hanushek, 1994), (Vittadini, 2004), (Aitkin, Longford, 1986), (Hanushek, 1986). The first concept is related to the analysis of costs and benefits deriving from the resources allocated to the educational process while the second concept is based on the comparison between the expected and the ob- tained results deriving from the educational process. Many statistical contributions have been developed in literature to construct and analyse efficiency and effectiveness indicators (Chiandotto, 2002), (Piccolo 2004), (Gori, Montagni, 1997), (Cozzucoli, Ingrassia, 2005), (Davino, Vistocco, 2007).

This paper aims to investigate on the internal effectiveness, namely the effect of an educational process on the students learning capability. In particular, the effect of personal data and of the student career features on the final degree mark is analysed through quantile regression (Koenker and Basset, 1978). The pro- posed approach allows to focus on the effects that the explanatory variables have on the entire conditional distribution of the dependent variable. The paper is or- ganized as follows: in Section 2 quantile regression is introduced highlighting its main added value with respect to classical regression; in Section 3 an analysis of the final degree marks of University of Macerata (Italy) students is carried out; finally some concluding remarks and further developments are provided.