A Flexible Software Process Model

Many different kinds of process are used to develop software intensive products, but there is little agreement as to which processes give the best results under which circumstances. Practitioners and researchers believe that project outcomes would be improved if the development process was constructed according to project-specific factors. In order to achieve this goal, greater understanding of the factors that most affect outcomes is needed. To improve understanding, researchers build models of the process and carry out studies based on these models. However, current models contain many ambiguities and assumptions, and so it is not clear what the results of the studies mean. The statement of this thesis is that it is possible to create an abstraction of the software development process that will provide a mechanism for comparing software processes and software process models. The long term goal of the research is to provide planners with a means of tailoring the development process on a project by project basis, with the aim of reducing risk and improving outcomes.