University departments evaluation: a multivariate approach

Aim of the paper is to present a new model, based on multivariate statistic analyses, allowing to express a synthetic judgement on Departments activities by taking into consideration the whole set of indicators describing them both as aggregations of researchers and as University autonomous organs. The model, based on Principal Component Analysis and Cluster Analysis, allows both to explain the determinants of Departments performances, and to classify them into homogeneous groups. The paper shows the results obtained by testing the proposed model on University of Naples “L’Orientale” Departments, using data extracted by the 2007 assessment report to the Ministry of University and Research.