Correct evaluation of type-B standard uncertainty

The evaluation of uncertainty components “by means other than the statistical analysis of series of observations” (type-B evaluation) is considered. The problem comes down to selection of an appropriate a priori probability distribution of a random quantity within specified limits on the basis of all available information. Typical situations are indicated that lead to simple model distributions: rectangular (uniform), triangular, normal, and arcsine. Attention is drawn to the unjustified use of the iniform distribution model in examples of measurement-uncertainty evaluation. Translated from Izmeritel'naya Tekhnika, No. 5, pp. 33–34, May, 2000.