A critique of the methods for measuring the impact of CASE software

There is a common belief that CASE software enhanced developer productivity and the quality of applications software. However, though the few empirical studies of the impact of CASE have produced inconclusive findings, they do indicate an absence of appreciable productivity gains. There is a need to determine the extent to which CASE products are worth their cost, and this paper focuses on the key methodological issues involved in assessing the cost-effectiveness of CASE products. Existing studies of the impact of CASE software have considered developers’ perceptions, but have not considered system users’ perceptions nor used software metrics to assess applications software. It is also rare for the characteristics of the development environments and other factors that have an important bearing on the productivity and quality of software to be investigated. In this paper, perceptual and objective measurement, and the different ways of designing the research and of accessing the population of CASE users, are outlined and evaluated. While the aim is to establish the most appropriate research design and measurement approaches for determining the value of CASE tools, the conclusions are also relevant to the assessment of the impact of other new software technologies.