Identification of test process improvements by combining fault trigger classification and faults-slip-through measurement

Successful software process improvement depends on the ability to analyze past projects and determine which parts of the process that could become more efficient. One source of such an analysis is the faults that are reported during development. This paper proposes how a combination of two existing techniques for fault analysis can be used to identify where in the test process improvements are needed, i.e. to pinpoint which activities in which phases that should be improved. This was achieved by classifying faults after which test activities that triggered them and which phase each fault should have been found in, i.e. through a combination of orthogonal defect classification (ODC) and faults-slip-through measurement. As a part of the method, the paper proposes a refined classification scheme due to identified problems when trying to apply ODC classification schemes in practice. The feasibility of the proposed method was demonstrated by applying it on an industrial software development project at Ericsson AB. The obtained measures resulted in a set of quantified and prioritized improvement areas to address in consecutive projects.