Measurement Uncertainties and Efficiencies Using Correlated and Un-correlated Cycles of Comparison Measurements

摘要

Comparison measurements between a standard and a test object may be performed and evaluated by various sequences of measurement cycles in repetition. It is understood that from each cycle a single measurement result and from the total number of them a mean and an uncertainty are evaluated. Consecutive cycles may have independent data or a certain number of common data. When evaluating of the measurement uncertainty of the mean, cycles with common data have to be considered as correlated and corresponding procedures have to be observed, which differ from the statistical procedure for a number of independent observations. Here, the uncertainty calculation for the often-used correlated RTR cycle is presented followed by a calculation for any type of correlated or uncorrelated cycle. Finally, an efficiency is defined and discussed for some selected cycles.