A framework for evaluation of factors affecting software development productivity
Olga Stefanova Mihova-Petkova, JD Roode · 1999
The costs associated with software development have grown significantly. It is essential that managers should strive to decrease them through improvement of software development productivity. In theory, productivity is defined as the amount of output produced per unit of input. Due to the many intangibles associated with software, the relevant outputs and inputs are very difficult to measure and hence their elusive nature. The purpose of this research is to provide a way for gaining better understanding of the influence of the factors affecting software development productivity and their interaction within a particular project. This task is viewed as a complex and “messy” problem. An investigation into the current state of the art of software cost and effort estimation is reported. A review of the influence of social science and systems thinking on Information Systems development over the last 20 years is presented. It is asserted that besides systems thinking another field, multiple criteria decision making, also has the potential to be used for the complex problem that is the scope of this research. An investigation of its schools of thought in general, and on the Analytical Hierarchy Process in particular is carried out. On the basis of the findings in the research, a pluralist systemic framework for the evaluation of factors affecting software development productivity is suggested. It is justified along the lines of the recently formulated meta-theoretic concepts of Multimethodology. An emphasis is placed on its validation and legitimisation. This framework is applied to a case study related to a complex software development project. The research is indirectly linked to work in the area of software cost and effort estimation, software process management, and software quality assurance. It does not aim, however, to substitute for any of the existing approaches related to the investigation of relationships between factors affecting software development productivity, but rather to complement them for facilitating deeper insight into the role of the various factors affecting software development productivity.