Interlaboratory consensus building

Determining the degree of equivalence of participating laboratories from the results of measurement comparisons still prompts discussions among metrologists, especially where measurement uncertainties have been underestimated. This paper expands on a solution to a problem issued in 2020 by the Journal of Analytical and Bioanalytical Chemistry. The example illustrates an approach to consensus-building based on Bayesian selection among statistical models that attempt to explain the excess of data variation. The probability of any further model being correct can be similarly calculated.