Bayesian inference for the pairwise probability of agreement using data from several measurement systems

Abstract

This article deals with Bayesian inference in the comparison of measurement systems. Agreement between two systems can be evaluated using data from several measurement systems and using only data from the two systems being compared. With a measurement error model for replicated observations and the probability of agreement to compare measurement systems, we develop methods to compare measurement systems with either homoscedastic or heteroscedastic measurement errors under the Bayesian paradigm via Markov chain Monte Carlo methods. A graphical tool is described to check model adequacy. The methodology developed in the article is illustrated using a real dataset and through simulations.