Decreasing the cost of mutation testing with second-order mutants

Although powerful, mutation is a computationally very expensive testing technique. In fact, its three main stages (mutant generation, mutant execution and result analysis) require many resources to be successfully accomplished. Thus, researchers have made important efforts to reduce its costs. This paper represents an additional effort in this sense. It describes the results of two experiments in which, by means of combining the original set of mutants and therefore obtaining a new set of mutants—each one with two faults—the number of mutants used is reduced to half. Results lead to believe that mutant combination does not decrease the quality of the test suite, whereas it supposes important savings in mutant execution and result analysis. Copyright © 2008 John Wiley & Sons, Ltd.