Predicting the number of faults is not always necessary to guide quali
ty development; it may be enough to identify the most troublesome modu
les. Predicting the quality of modules lets developers focus on potent
ial problems and make improvements earlier in development, when it is
more cost-effective. In such cases, classification models rather than
regression models work very well. As a case study, this article applie
s discriminant analysis to identify fault-prone modules in a sample re
presenting about 1.3 million lines of code from a very large telecommu
nications system. We developed two models using design product metrics
based on call graphs and control-flow graphs. One model used only the
se metrics; the other included reuse information as well. Both models
had excellent fit. However, the model that included reuse data had sub
stantially better predictive accuracy. We thus learned that informatio
n about reuse can be a significant input to software quality models fo
r improving the accuracy of predictions.