Introducing Legacy System Migration Technologies in an Academic Context: a Controlled Experiment Massimo Colosimo, Andrea De Lucia, and Genoveffa Tortora Dipartimento di Matematica e Informatica, University of Salerno, Via Ponte Don Melillo, 84084, Fisciano (SA), ITALY cillo_81@yahoo.it, {adelucia, tortora }@unisa.it Giuseppe Scanniello Dipartimento di Matematica e Informatica, University of Basilicata, Viale Dell'Ateneo, Macchia Romana, 85100, Potenza, ITALY giuseppe.scanniello@unibas.it Abstract To verify whether or not migration technologies can be adopted, systematic and quantitative evaluations should be performed. In this paper we present the results of a controlled experiment aimed at assessing the usefulness of an Eclipse plug-in, named MELIS (Migration Environment for Legacy Information Systems). This plug-in has been developed to support the migration of legacy information systems to a web-enabled multi-tier target architecture according with an incremental migration strategy. The context of the experiment was constituted of master students in Computer Science at the University of Salerno. The subjects without COBOL programming experience, while half of them had J2EE programming experience. The results of the experiment confirmed that the use of MELIS increases the productivity with respect to the use of traditional development tools. 1. Introduction Research in the empirical software engineering field aims at acquiring general knowledge about which process, method, technique, language, or tool is useful for whom to conduct which tasks in which environments [5][18][19][31]. In particular, empirical studies should be devised to transfer software engineering technologies produced in research laboratories to practitioners and to evaluate whether they potentially fulfill the industry needs [6][22][28]. To verify whether or not software engineering technologies can be adopted, systematic and quantitative evaluations should be performed in terms of surveys, controlled experiments, and case studies [6][39]. Among these evaluation techniques controlled experiments are recognized to be important means since they represents the classical scientific method for identifying cause-effect relationships [31].