ABSTRACT
The Bureau International des Poids et Mesures (BIPM) terminology guide (International Vocabulary of Metrology, VIM) provides a concise definition of uncertainty as “non-negative parameter characterizing the dispersion of the quantity values being attributed to a measurand, based on information used.”1 In other words, the uncertainty is the range in which we expect a value to lie. Notice that we cannot say the range in which the true value lies, although analytical methods are designed to obtain a result that is as close to the true value as possible. Estimating of uncertainty is about estimating the expected spread of the measurement results. Uncertainty is not primarily about accuracy, nor does it describe the goodness of a result. All that uncertainty expresses is the range associated with a measured result. Any analytical method addresses accuracy and quantifies through comparison to what we previously described as “generally accepted as true”; we measure this as bias or trueness and often express it as % error. Bias/trueness/% error are single numbers; uncertainty is a range that is accompanied by a significant value or confidence level. It is worth noting here that uncertainty does encompass both random and systematic components. Accordingly, when appropriate,
the “single numbers” of bias/trueness/%error can be added to the range of random components to reflect an uncertainty composed of both random and systematic factors. Ideally, any measurable bias of a validated method is reduced such that a measured result is indistinguishable from a generally accepted true value. Thus, for now we will focus on the random components.
