Early Evaluation of Software Process Change Proposals

Sammanfattning: Software process improvement involves evaluation of process change proposals, and it is important to evaluate changes as early as possible. A process impact analysis method for interview based evaluation of change proposals is presented. It is based on the subjective opinions of people in the organisation, and it is carried out in three steps. In the first step, the impact on sub-processes is estimated through interviews, and in the second step the overall impact in terms of, for example, lead time and effort is estimated for the whole process. In the third step the result is presented as a basis for decision. It is presented how the process impact analysis method can be integrated and used in a framework for technology evaluation and introduction. Moreover, it is reported from experiments where aspects of the first step of the impact analysis method have been evaluated, and it is also reported from studies where evaluations based on the whole method have been carried out in industrial organisations. Process changes often include comparison and evaluation of competing technologies. Experimentation and case studies are hence good complements to the process impact analysis method. These types of empirical studies can be conducted in an educational or industrial environment. It is in particular described how experimentation can be conducted using the Personal Software Process (PSP) as a context for experimentation. Experience from conducting experiments both in the context of the PSP and in an industrial environment is presented. The results presented in the thesis show that it is feasible to carry out early evaluations of process change proposals. This can, for example, be carried out through interviews and experimentation.