2014/04/24 by Valentina Raponi, Raponi, Valentina, Francesca Martella +3
Mathematics · Psychology · Social Sciences · #Applications (stat.AP) #Artificial intelligence #Biclustering #Cluster analysis #Computer science #Data science #Descriptive statistics #Econometrics #Economics #Evaluation of Teaching Practices #FOS: Computer and information sciences #Higher Education Governance and Development #Homogeneity (statistics) #Machine learning #Management science #Mathematics #Psychology #Regional science #Social psychology #Sociology #Statistics #Strengths and weaknesses #stat.AP
paper · pdf · doi:10.48550/arxiv.1404.6193
published in arXiv (Cornell University) (Cornell University)
arxiv created 2014/04/24 · openalex publication_date 2014/04/24 · arxiv updated 2014/04/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
University evaluation is a topic of increasing concern in Italy as well as in\nother countries. In empirical analysis, university activities and performances\nare generally measured by means of indicator variables, summarizing the\navailable information under different perspectives. In this paper, we argue\nthat the evaluation process is a complex issue that can not be addressed by a\nsimple descriptive approach and thus association between indicators and\nsimilarities among the observed universities should be accounted for.\nParticularly, we examine faculty-level data collected from different sources,\ncovering 55 Italian Economics faculties in the academic year 2009/2010. Making\nuse of a clustering framework, we introduce a biclustering model that accounts\nfor both homogeneity/heterogeneity among faculties and correlations between\nindicators. Our results show that there are two substantial different\nperformances between universities which can be strictly related to the nature\nof the institutions, namely the Private and Public profiles . Each of the two\ngroups has its own peculiar features and its own group-specific list of\npriorities, strengths and weaknesses. Thus, we suggest that caution should be\nused in interpreting standard university rankings as they generally do not\naccount for the complex structure of the data.\n