Abstract:
The article ciricises conventional credibility characteristics of compact testing methods. An alternative approach is suggested whereby efficiency is compared in terms of distributions of probabilities that errors of specified multiplicity will remain undetected.
Fulltext:
PDF file (899 kB)