Towards human-centric software testing

Abstract

Software testing is widely perceived to be the main activity in the software development process that provides confidence in the quality of a product prior to release. However, the term software testing itself provokes a multitude of different definitions and opinions as to the nature of the profession, the role of software testers and the utility of different processes and tools that come with the territory [1][2]. We argue that in order for researchers to effectively study the field and contribute to its progress, a consensus first needs to be reached about the entity being studied. In this paper we present an empirical study based on the modified Delphi card sort method involving four cohorts of testers in Malta and London. The result of this study is a consolidated consensus-based mental model outlining how software testers perceive their profession. This mental model can be used to align any future research efforts and tool development with testers' own perception of their context.