Using and evaluating CASE tools : from software engineering to phenomenology

CASE (Computer-Aided Systems Engineering) is a recent addition to the long line of "silver bullets" that promise to transform information systems development, delivering new levels of quality and productivity. CASE is particularly intriguing because information systems (IS) practitioners spend their working lives applying information technology (IT) to other people's work, and now they are applying it to themselves. CASE research to date has been dominated by accounts of tool development, normative writings (for example practitioner success stories) and surveys recording IT specialists' perceptions. There have been very few in-depth studies of tool use, and very few attempts to quantify benefits, therefore the essence of the CASE process remains largely unexplored, and the views of stakeholders other than the IT specialists have yet to be heard. The research presented here addresses these concerns by adopting a hybrid research approach combining action research, grounded theory and phenoinenology and using both qualitative and quantitative data in order to tell the story of a system developer's experience in using CASE tools in three information systems projects for a major UK car manufacturer over a four year period. The author was the lead developer on all three projects. Action research is a learning process, the researcher is an explorer. At the start of this project it was assumed that the tools would be the focus of the work. As the research progressed it became evident that the tools were but part of a richer organisational context in which culture, politics, history, external initiatives and cognitive limitations played important roles. The author continued to record experiences and impressions of tool use in the project diary together with quality and productivity metrics. But the diary also became home to a story of organisational developments that had not originally been foreseen. The principal contribution made by the work is to identity the narrow positivistic nature of CASE knowledge, and to show via the research stories the overwhelming importance of organisational context to systems development success and how the exploration of context is poorly supported by the tools. Sixteen further contributions are listed in the Conclusions to the thesis, including a major extension to Wynekoop and Conger's CASE research taxonomy, an identification of the potentially misleading nature of quantitative IS assessment and further evidence of the limitations of the "scientific" approach to systems development. The thesis is completed by two proposals for further work. The first seeks to advance IS theory by developing further a number of emerging process models of IS development. The second seeks to advance IS practice by asking the question "How can CASE tools be used to stimulate awareness and debate about the effects of organisational context?", and outlines a programme of research in this area