On improving the statistical method used to access the completeness of tests for software systems and discrete devices

This paper describes a statistical method for test completeness verification. The method is based on analyzing programs and mathematical models of discrete devices with introduced defects. It is proposed in the paper that the method allows largely reducing the time it takes to evaluate the tests by introducing multiple defects on the one hand, and, by using the results of the analysis of tests performed against the object under test and its individual components, on the other hand. The paper provides experimental data confirming the effectiveness of the proposed approaches.