DEFT – A Test Framework to Aid Decision Making

Abstract

Testing can benefit greatly if there are ways to predict the defects that will be detected during a phase of the software development lifecycle and none more so than during the System testing phase itself. There are several examples of defect prediction models in the literature [1]. These have been derived, used and also critically evaluated. However, they are not the easiest to use due to the dependencies on the data and metrics required for these models; as they take into account several variables. This paper describes a Defect Effort Fix & Test (DEFT) framework that was developed to provide an end date for testing for a challenging programme.