Empirical studies of a prediction model for regression test selection

Regression testing is an important activity that can account for a large pr oportion of the cost of software maintenance. One approach to reducing the cost of regression testing is to employ a selective regression testing tech nique that 1) chooses a subset of a test suite that was used to test the so ftware before the modifications, then 2) uses this subset to test the modif ied software. Selective regression testing techniques reduce the cost of re gression testing if the cost of selecting the subset from the test suite to gether with the cost of running the selected subset of test cases is less t han the cost of rerunning the entire test suite. Rosenblum and Weyuker rece ntly proposed coverage-based predictors for use in predicting the effective ness of regression test selection strategies. Using the regression testing cost model of Leung and White, Rosenblum and Weyuker demonstrated the appli cability of these predictors by performing a case study involving 31 versio ns of the KornShell. To further investigate the applicability of the Rosenb lum-Weyuker (RW) predictor, additional empirical studies have been performe d. The RW predictor was applied to a number of subjects, using two differen t selective regression testing tools, DejaVu and TestTube. These studies su pport two conclusions. First, they show that there is some variability in t he success with which the predictors work and second, they suggest that the se results can be improved by incorporating information about the distribut ion of modifications. It is shown how the RW prediction model can be improv ed to provide such an accounting.