Abstract
Software development and maintenance involves organizational as well as technical issues. While software engineering has been offered as a way of resolving the intrinsic technical problems, the organizational problems need another approach. Still, the typical response to the persistent problems in software development has been to apply even more rigorously the principles of engineering. Organizational issues are often ignored or not properly addressed in much of the recent literature, with only a handful of studies being reported.To help fill this gap, this doctoral study was initiated to explore the relative importance of organizational issues in software process improvement. The position taken is strongly influenced by socio-technical theory with its central conception that organizations are both social and technical systems, and that the core of the software organization is represented through the interface between the technical and human (social) system. The fundamental assumption is that process improvement is a socially constructed learning process and that a commitment to learning rather than to ""best practice"" models is needed to accomplish improvements in software development.The research presented in this dissertation is based on several empirical studies performed within the context of the European Systems and Software Initiative and the Norwegian software process improvement projects SPIQ and PROFIT. The overall research design for the study was a sequential, mixed method design consisting of two major phases: first, a qualitative model-building phase, then a quantitative model-testing phase. The model-building phase was grounded in prior consulting experience in software development, in observations done in a pilot case study as well as in a multiple case study of twelve organizations, and in an extensive literature review. The model-testing phase focused on testing the associations between the key factors for success and the outcome variable by a quan