Uncertainty and traceability in calibration by comparison

The objective of this article is to present a methodology for the evaluation of uncertainties in the results of calibration by comparison, based on the implementation of the concept of traceability and on the use of weighted least squares fitting. The general case, where the measuring equipment is tested against a standard of higher metrological quality at several points on a range, is considered. The proposed methodology deals with not only the experimental data in each calibration point but also their quality. The fact that both the outputs of the reference standard and the indications of the device under calibration are characterized by their own uncertainty is also taken into account. In this way calibration leads not only to determination of the coefficients of the calibration curve, but also to estimation of their uncertainties. This is particularly useful when a statement of uncertainty which concerns a range of values rather than a single result is required.