A review of methods for measurement systems capability analysis

We review methods for conducting and analyzing measurement systems capability studies, focusing on the analysis of variance approach. These studies are designed experiments involving crossed and possibly nested factors. The analysis of variance is an attractive method for analyzing the results of these experiments because it permits efficient point and interval estimation of the variance components associated with the sources of variability in the experiment. In this paper we demonstrate computations for the standard two-factor design, describe aspects of designing the experiment, and provide references for situations where the standard two-factor design is not applicable.