Quantitative Quality Assurance Approach

There is profound confusion at the final stages of software development as to which defect discovered belongs to which phase of software development lifecycle. Techniques like root cause analysis and orthogonal defect classification are commonly used practices that can be applied jointly with the software process to determine each defect attribute in terms of its type, trigger, source etc. However, there is a need to find a mechanism to determine the origin of those defects with respect to the verification technique applied to it in order to determine its efficiency in terms of its defects found and execution time. Defect containment matrix has been widely used to determine the efficiency of defect detection and removal processes at early life cycle phases.Nevertheless, having one vague value in terms of percentage of defects found for each phase proved to be inaccurate way due to the variations of defects severities and impact to the software schedule, effort and quality. This paper proposes a model that classifies each phase artifact to different work product according to its severity level.