Experiences with incorporating experimental practice into Software Engineering research
David Budgen, Mitchell Thomson, St Bg
While research activity in software engineering often results in the development of software tools and solutions that are intended to demonstrate the feasibility of an idea or concept, any resulting claims of success are rarely substantiated to the extent of producing supporting experimental evidence. In this paper, we describe our experiences with conducting an experimental `evaluation' at the end of an investigation that developed a system intended to support opportunistic and non-prescriptive software design practices. This involved designing and performing a set of experimental studies in such a manner as to obtain the maximum amount of useful data. The paper concludes by reviewing our experiences and considering how readily such practices could be applied to other software engineering activities. 1 Introduction The disciplines of computer science and software engineering…