Introduction

An estimate of the measurand Y, noted y, is obtained from the equation called mathematical model of measurement , by using the estimates x1, x2, ..., xj of the input quantities X1, X2, ..., Xj. The standard deviation associated with the output estimate or the measurement result y, called combined standard uncertainty and noted uc(y), determined from the estimated standard deviation associated with each input estimate xi called standard uncertainty and noted u(xi). Each input estimate xi and its associated standard uncertainty u(xi) are obtained from a distribution law of the possible values of the input quantity Xi. This probability law can be based on a series of replicate observations Xi,k of the several Xi,in which case we will refer to a évaluation de Type A of the components of the standard uncertainty. It can also be an a priori law, and will therefore correspond to a Type B evaluation. In both cases, the laws that are used depend on our level of knowledge of the measurement mean.