Chapter 36. Qualitative Studies of XP in a Medium-Sized Business

Robert Gittins, Sian Hope, and Ifor Williams

This chapter introduces novel methods of gathering qualitative data, data collation, and data analysis and is designed to both improve research techniques and provide useful information to companies considering a move to Extreme Programming (XP). The work has been conducted to understand some of the problems encountered during early XP implementation. In particular, the chapter looks at the XP practices adopted by one software development company and reveals how some developers feel about and respond to change; discloses what their managers have to say about adopting XP practices; and considers both advantages and problems of introducing some early, selected XP practices.

Previous qualitative research into software development concentrated on nonjudgmental reporting, with the intent of provoking discussion within the culture being studied [Seaman1999; Sharp+1999; Cockburn+2000]. Such work provided observations and evidence, with collaborators deciding for themselves whether changes were required. The fieldwork study undertaken here followed the format of [Gittins+2000], whereby the researcher is immersed for a period in the software developer team. In this way, the active researcher becomes instrumental in the development and improvement of XP. Seaman described an empirical study that addressed the issue of communication among members of a software development organization [Seaman1999]. Sharp used combined ethnography and discourse analysis to discover implicit assumptions, values, and beliefs in a software management system [Sharp+1999]. Cockburn and Williams investigated the cost benefits of pair programming [Cockburn+2000]. Sharp, Robinson, and Woodman described a "cross-pollination" approach to inform and improve the development of software engineering through a deeper understanding of the software engineering community's implicit values and beliefs [Sharp+2000]. What has been reported has been done through the successful use of qualitative methods in the evaluation of software development approaches.

Now we look at XP and its evaluation by using qualitative procedures. What are the effects of applying XP in a software development business environment? To answer this question, qualitative research methods [Gittins2000] were explored and used to discover the problems dominating staff and management development. Productivity and efficiency are parts of a complex human dimension uncovered by this approach. The interpretation and development of XP's "rules and practices" in a medium-sized software development company, as well as the intricate communication and human issues affecting the implementation of XP, are discovered.