In a Key Comparison experiment, the Mutual Recognition Arrangement (MRA) requires the calculation of the Degrees of Equivalence (DoE) and the accompanying measures of uncertainty. Unilateral DoEs have been defined as a difference between a laboratory?s measurement and a Key Comparison Reference Value (KCRV). The functional form of the KCRV has been a source of much debate. Much of the proposed methodology has been based on concepts from classical (or frequentist) statistics. The focus of this paper is on the case when the result of each laboratory?s measurement is interpreted as a mean and standard deviation of a probability distribution of belief about a measurand. Bayesian statistical methods are best suited to deal with this situation and lead to interesting insights about the form of the KCRV and the DoEs.
Pub Type: Journals
Degrees of Equivalence, Key Comparison Experiments, Belief Distributions, Synthesis of Probability Distributions