A Formal Evaluation of Mutation and Data Flow

Based Test Adequacy Criteria?

Aditya P. Mathur and W. Eric Wong
Department of Computer Sciences
Purdue University
West Lafayette, IN 47907

Revised on December 28, 1993

Abstract
Evaluation of the adequacy of a test set is a frequently encountered problem in software testing. Data flow and mutation-based testing methods provide several criteria to evaluate test sets. This work considers the c-use, p-use, and all-uses based data flow criteria and compares these with a mutation-based criterion. We show that for the most general class of programs, mutation and the above data flow criteria are incomparable. We also identify classes of programs for which a mutation-based criterion subsumes the c-use, p-use, and alluses based criteria. Our results are significant in the light of the fact that the determination of the subsumption relationship between data flow and mutation continues to be an open problem in the field of software testing.

?This research was supported in part by NSF award CCR-9102331 and Purdue Research Foundation Research Grant No. 6901999.