Bayesian evaluation of comparison data

Abstract

This paper develops a Bayesian procedure to evaluate comparison data when more than two participants measure independently a stable travelling standard. The procedure yields the analytic probability density function (pdf) that encodes the collective state of knowledge. From this pdf, the estimated value of the measurand may be derived numerically, together with its associated standard uncertainty and coverage interval. The pdfs for the degrees of equivalence of the participating laboratories can also be calculated.