Mutation testing-an evolving technique

The paper provides a brief review of the program testing technique known as mutation testing, an error-based testing technique. A large number of simple changes (mutations) are made to program one at a time. Test data then has to be found which distinguishes the mutated versions from the original version. An underlying assumption of mutation testing is that experienced programmers write nearly correct programs. Hence small perturbations of the original program are expected to cause an observable difference to the program´s performance. Mutation testing as originally developed was considered by many as somewhat bizarre and costly to perform. Subsequently a less demanding form known as weak mutation testing was proposed and the original technique was renamed strong mutation testing. More recently firm mutation has been introduced as an intermediate approach. The author discusses these types of mutation testing. As an example of the wide applicability of the mutation approach, its use in testing algebraic specifications is considered