Exploring XP's efficacy in a distributed software development team

Abstract

Since the first edition of Beck's book [1], the Extreme Programming (XP) has attracted attention from academia and industry, and its values, principles and practices are becoming increasingly popular. Strong interest in the software engineering community has generated substantial literature and debate over Extreme Programming. However, current research on the applicability and effectiveness of Extreme Programming is still very scarce and researchers and practitioners need to assess concretely XP's advantages and drawbacks. One disadvantage, which has been noted, is that Extreme Programming is more effective for small to medium size projects with co-located team. Despite such observation, Beck asserts that XP can work with teams of any size and also multi-site [2]. The main goal of this research is to evaluate the effectiveness of Extreme Programming, when the size of development team is large and distributed. There are two different approaches to investigate the applicability and effectiveness of a software method: empirical studies and simulation process modeling. These approaches are usually applied separately, but there are many interdependencies between simulation and empirical research. On the one hand simulation model generalizes empirical studies and provides a framework for the evaluation of empirical models. On the other hand, empirical studies provide the necessary fundament for simulation models because through empirical studies it is possible to collect real data to validate the simulation model. In the present research these two approaches are combined. © Springer-Verlag Berlin Heidelberg 2005.