作者: N.F. Schneidewind
DOI: 10.1109/32.135774
关键词: Software metric 、 Data mining 、 Software quality 、 Quality assurance 、 Software 、 Quality (business) 、 Project management 、 Metric (unit) 、 Repeatability 、 Computer science 、 Software measurement
摘要: A comprehensive metrics validation methodology is proposed that has six validity criteria, which support the quality functions assessment, control, and prediction, where are activities conducted by software organizations for purpose of achieving project goals. Six criteria defined illustrated: association, consistency, discriminative power, tracking, predictability, repeatability. The author shows nonparametric statistical methods such as contingency tables play an important role in evaluating against criteria. Examples emphasizing power criterion presented. process integrates factors, metrics, functions. >