Making Methods Work in Software Engineering : Method Deployment - as a Social Achievement

Sammanfattning: The software engineering community is concerned with improvements in existing methods and development of new and better methods. The research approaches applied to take on this challenge have hitherto focused heavily on the formal and specifying aspect of the method. This has been done for good reasons, because formalizations are the means in software projects to predict, plan, and regulate the development efforts. As formalizations have been successfully developed new challenges have been recognized. The human and social role in software development has been identified as the next area that needs to be addressed. Organizational problems need to be solved if continued progress is to be made in the field. The social element is today a little explored area in software engineering. Following with the increased interest in the social element it has been identified a need of new research approaches suitable for the study of human behaviour. The one sided focus on formalizations has had the consequence that concepts and explanation models available in the community are one sided related in method discourses. Definition of method is little explored in the software engineering community. In relation to identified definitions of method the social appears to blurring. Today the software engineering community lacks powerful concepts and explanation models explaining the social element. This thesis approaches the understanding of the social element in software engineering by applying ethnomethodologically informed ethnography and ethnography. It is demonstrated how the ethnographic inquiry contributes to software engineering. Ethnography is also combined with an industrial cooperative method development approach. The results presented demonstrate how industrial external and internal socio political contingencies both hindered a method implementation, as well as solved what the method was targeted to do. It is also presented how project members’ method deployment - as a social achievement is played out in practice. In relation to this latter contribution it is provided a conceptual apparatus and explanation model borrowed from social science, The Documentary method of interpretation. This model addresses core features in the social element from a natural language point of view that is of importance in method engineering. This model provides a coherent complement to an existing method definition emphasizing formalizations. This explanation model has also constituted the underpinning in research methodology that made possible the concrete study results.