Adequacy check of models representing measurement objects as a necessary condition for the applicability of the Guide to Expression of Uncertainty in Measurement
摘要
The article considers a condition for the applicability of the 1999 Russian translation of the Guide to the Expression of Uncertainty in Measurement. This Guide presents an approach to the accuracy assessment of measurement problem solutions obtained via an indirect measurement method, which was new to domestic metrology at the time. In order for the Guide to be applicable, any uncertainties associated with the incomplete definition of target quantities should be negligible compared to the required accuracy of the obtained result. The translation, however, states that “this may not always be practical.” For this kind of uncertainty, the International Vocabulary of Metrology uses the term “definitional uncertainty,” stating that the Guide considers definitional uncertainty to be negligible compared to other components of measurement uncertainty. In practice, the Guide is often used without checking mathematical models for adequacy. The situation is exacerbated by the fact that the Guide does not define a quantitative indicator for the incompleteness of a mathematical model representing a measurement object, even though the need for a statistical check of models for adequacy is mentioned. This applicability issue can be resolved through the use of a cross-observation scheme.