Accounting for Measurement Uncertainty
3.1 Measurement Errors
The de facto standard for the evaluation of measurement uncertainty is the IEC-ISO Guide to the Expression of Uncertainty in Measurement (GUM) (Joint Committee for Guides in Metrology, 2008). The Guide defines measurement uncertainty as “a parameter, associated with the result of a measurement, which characterizes the dispersion of values that could reasonably be attributed to the measurand”. It is recognized that the uncertainty results from two basic types of errors – random errors (Type A errors) and systematic errors (Type B errors) – and assumed that the measurement can be corrected for recognized systematic errors. Standard uncertainty is defined and the use of confidence intervals and levels of confidence to relate an expanded or overall uncertainty to the standard uncertainty is advocated. The guide goes on to specify how the expanded uncertainty can be calculated when a series of observations of the measurand are available (e.g. 95% confidence level for a coverage factor of two standard deviations) and when other kinds of information (e.g. manufacturers’ specifications, calibration information) about the uncertainty must be utilized.
Until recently, the main mathematical tool for dealing with measurement uncertainty has been probability theory. The limitations of this approach have now been recognized (Ferrero and Salicone, 2005) and some of the assumptions underlying the proposed techniques for describing ...