Repeatability: some aspects concerning the evaluation of the measurement uncertainty

Various publications stress the importance of the repeatability (i.e. precision) of the calculation of the measurement of uncertainty. We reveal by detailing an example from production control in the pharmaceutical industry that the effect of other influence quantities should not be neglected, because their magnitude is even larger than the contribution of repeatability. We review the role of repeatability within the calculation of measurement uncertainty for several common validation and day-to-day measurement scenarios. They show that measurement models need to consider the measurement sequences of the various scenarios. Otherwise the size and effect of the repeatability might be overestimated. At the end Monte Carlo simulations were used to investigate the determination of the repeatability under certain restrictions. The simulation uncovered a significant bias toward the common formula for calculating the standard deviation when it is based on a duplicated measurement of a sample.