A new data analysis procedure is presented for the evaluation of comparisons of reference materials individually prepared and value-assigned by different organizations to measurements made in a single laboratory under repeatability measurements. The objective of such comparisons is to assess the value assignment capabilities of participating organizations when they are not constrained by severely limited resources. The proposed statistical analysis uses several novel concepts, most importantly a leave-one-out strategy for the estimation of the consensus value of the measurand, model fitting via Bayesian posterior probabilities, and posterior coverage probability calculation for the assigned 95% uncertainty intervals. The benefit of this analysis method is illustrated using data from the CCQM-K54 comparison of eight cylinders of hexane in methane, each prepared by a participating national metrology institute.
Citation: Analytical and Bioanalytical Chemistry
Pub Type: Journals
Bayesian Analysis, Degrees of Equivalence, Generalized Distance Regression (GDR), Leave-One-Out analysis (LOO), Posterior Coverage Probability