A proposed new definition of measurement uncertainty

Abstract

A new definition of the measurement uncertainty (MU) was proposed by the Working Group 1 of the Joint Committee for Guides in Metrology, JCGM-WG1. This definition avoids any quantitative aspect of the measurement uncertainty and focuses on its subjective nature. The proposed new definition reads as follows: doubt about the true value of the measurand that remains after making a measurement. Here, MU is no longer quantitative; it is rather a (subjective) state of mind. MU is the concept, and its quantitative measures, such as the standard measurement uncertainty, are different in nature. This separation greatly contributes to clarity. Being defined as a state of mind, MU is subjective and reflects the belief of the experimenter in the result. There is no “true uncertainty” in nature to be estimated. There exists the measurand, and the uncertainty about its true value is a personal matter. Of course, the state of belief is based on objective data, and a good experiment is conceived in such a way as to minimise subjectivity. Yet, the hope to eliminate subjectivity from a measurement or from science at large is just a hope. The proposed new definition explicitly uses the term “true value”. Perhaps, at the philosophical level, the concept can be questioned, whereas in the context of parameter estimation, the mathematics behind calculations needs a unique true value, which is ideally represented by a unique real number. The doubt is about the unknown value of the measurand, not about the estimate. The estimate is viewed as a realization of a random variable describing the state of knowledge about the measurand. As such, the estimate is fixed and has no uncertainty. Randomness is in the variable, not in its realizations.