Assessing Productivity of UML-based Systems Analysis and Design: A DEA Approach

Due to the recent economic woes, information systems (IS) departments are under enormous pressure to cut costs while maintaining productivity in software development projects. Improving software project productivity has become a critical issue for every organization especially under the current economy uncertainty. Previous studies focus mainly on the efficiency of the traditional SAD process and there have been no attempts, to our best knowledge, to explore the efficiency of OOSAD process. This study uses UML complexity metrics to measure the productivity of OOSAD projects. DEA models are developed to explore the efficiency of these projects. A preliminary empirical study is conducted to apply the proposed methodology. Finally, sensitivity analysis of DEA is used in the study to identify factors and ways to improve the project efficiency and managerial implications are also discussed.