Reducing mutation costs through uncovered mutants

Mutation testing is a very effective testing technique that creates mutants in order to design test cases that will kill the mutants. One problem of mutation testing is high costs: creating mutants, executing mutants and calculating the mutation score. This paper empirically analyzes the effect of covering mutants through the mutant schema improved with extra code (MUSIC) technique. This technique annotates the statements covered by the tests in the original system in order to filter the mutant executions, because tests are only executed against the mutants whose mutated statement is covered by the tests. Therefore, MUSIC is meant to reduce the number of required executions and identify infinite loops at a reduced cost. Besides, an experiment was performed to evaluate the advantages and disadvantages of analyzing the covered mutants. As a result, we conclude that analyzing the covered mutants reduces the execution cost of mutation testing and its application is therefore recommended. Copyright (C) 2014 John Wiley & Sons, Ltd.