Estimation of the common mean and the comparison reference value

Rok, strany: 2008, 53 - 60

We discuss the problem of evaluating measurement results from interlaboratory comparisons in metrology. Here we permit the laboratories to have systematic errors that are assumed to be uniformly distributed. Two approaches for modeling the measurement results are introduced and compared, and the associated problems connected with estimation of the quantity in question are analyzed. The first is a classical (frequentist) statistical approach resulting in a heteroscedastic one-way random effects model. The second approach is based on metrological methodology. In both cases the comparison reference value (an estimate of the unknown measured quantity) with the approximate interval estimators is proposed.

ISO 690:

Witkovkský, V., Wimmer, G. 2008. Estimation of the common mean and the comparison reference value. In Tatra Mountains Mathematical Publications, vol. 39, no.1, pp. 53-60. 1210-3195.

APA:

Witkovkský, V., Wimmer, G. (2008). Estimation of the common mean and the comparison reference value. Tatra Mountains Mathematical Publications, 39(1), 53-60. 1210-3195.