Reconciling internal and external performance in a holistic approach: A Bayesian network model in higher education Laura Di Pietro a , Roberta Guglielmetti Mugion a , Flaminia Musella b,⇑ , Maria Francesca Renzi a , Paola Vicard c a Department of Business Studies – University Roma Tre, Via Silvio d’Amico 77, 00145 Rome, Italy b Department of Research – Link Campus University, Via Nomentana 335, 00162 Rome, Italy c Department of Economics – University Roma Tre, Via Silvio d’Amico 77, 00145 Rome, Italy article info Article history: Available online 15 November 2014 Keywords: Bayesian networks Internal and external performance Holistic approach Quality in higher education abstract University education is crucial for cultural and economic growth. Thus, the academic mission recognizes the achievement of both institutional and social objectives, and research provides the basis for the systematic creation of knowledge and the development of human capital. Universities attempt to manage a global system with a holistic vision based on data and facts and oriented to the continuous improve- ment of its effectiveness and efficiency. The goal is achieved by implementing a monitoring system based both on internal and external performances. As a consequence, it is necessary to consider both students perspective regarding needs, expectations, level of satisfaction and loyalty and internal key performance indicators. This paper proposes the use of Bayesian networks for jointly monitoring internal and external perfor- mance of a Master’s programme of an Italian University in a holistic approach. A Bayesian network is estimated using a learning algorithm able to analyze the association structure among mixed ordinal and nominal variables. Various scenarios are evaluated thanks to efficient computational algorithms of Bayesian networks. Ó 2014 Elsevier Ltd. All rights reserved. 1. Introduction Quality in the higher education system is essential for cultural and economic growth, considering that the mission of universities recognizes the achievement of both institutional and social objec- tives, and research provides the basis for the systematic creation of knowledge and the development of human capital (Di Pietro, Guglielmetti Mugion, & Renzi, 2012). Since 1980, total quality management (TQM) has become an essential of the higher education system and was considered an interesting philosophy for governance (Srikanthan & Dalrymple, 2002). To increase the quality of input, processes, and output, uni- versity systems must monitor their internal and external perfor- mance and emphasize the need for continuous improvements to achieve efficiency and effectiveness. The principles of TQM have been adopted in university systems to address external factors. Hwarng and Teo (2001) note that higher education has so far been sustainable because education has traditionally been funded by the government and the competition for quality students and grants has been less intense. The authors emphasize that this scenario is currently changing because of the increased competition that has been created by rapidly changing technology and growing interna- tional competition for students, faculty, and a demand for the delivery of excellent programmes from research institutions. Furthermore, due to declining public funds and rising tuition fees, higher education establishments worldwide have adopted a customer perspective (Eagle & Brennan, 2007). Certain universities, as suggested by Paswan and Ganesh (2009), are realizing that higher educational institutions are business entities, and as such, they too must compete for resources and customers, or students, in local and international markets. With respect to higher educa- tion, students are the main consumers of the didactic offer (Gremler & McCollough, 2002; Hill, 1995; Moosmayer & Siems, 2012; Sander, Stevenson, King, & Coates, 2000); therefore, univer- sity management must respond to stakeholders’ desires and needs (students, parents, etc.). Thus, the opinion of students with respect to teaching, organization and related auxiliary services, is relevant for the identification of inefficiencies and ineffectiveness within a http://dx.doi.org/10.1016/j.eswa.2014.11.019 0957-4174/Ó 2014 Elsevier Ltd. All rights reserved. ⇑ Corresponding author. E-mail addresses: laura.dipietro@uniroma3.it (L. Di Pietro), roberta. guglielmettimugion@uniroma3.it (R. Guglielmetti Mugion), f.musella@unilink.it (F. Musella), mariafrancesca.renzi@uniroma3.it (M.F. Renzi), paola.vicard@ uniroma3.it (P. Vicard). Expert Systems with Applications 42 (2015) 2691–2702 Contents lists available at ScienceDirect Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa