Towards a "personal cost" model for end-user development

Abstract

Software cost estimation techniques are used to provide a useful measure of the complexities, efforts, and costs involved in system development. Despite three decades of research on software cost estimation, the research community has yet to provide a viable model for End-User Development (EUD) environments. This is both surprising and significant as EUD has been estimated to account for somewhere between 80--95% of all personal software development activities worldwide. As EUD is predominantly idiosyncratic, self-determined and opportunistic, existing cost estimation techniques are either inappropriate, overly complex or insufficient. In response, we present a novel "personal cost" estimation model for EUD. This model reflects key features of non-professional programming, and provides a qualitative evaluation of personal cost as motivation, and hence, likely project success.