A Comparison Between the Probabilistic and Possibilistic Approaches: The Importance of a Correct Metrological Information

When dealing with measurement uncertainty, a metrologist is required to take into consideration many different points. The most important ones are: the uncertainty contributions affecting the measurement process; the random or systematic nature of these contributions; the available metrological information about these contributions; and last, but not less important, the particular measurement process itself. This paper shows, with two simple examples, the importance of correctly considering all the aforementioned points in evaluating uncertainty. The guide to the expression of uncertainty in measurement (GUM) approach, a Monte Carlo approach, sometimes different from the one recommended by the GUM Supplement 1, and the random-fuzzy variable (RFV) approach are compared, proving that the RFV approach leads to the simplest and most straightforward way to consider all these points.