Traceability, uncertainty and calibration certificates

摘要

The uncertainty of the estimated value of a directly measured quantity, as calculated following the guidelines that appear in the ISO Guide to the Expression of Uncertainty in Measurement, is influenced by the uncertainties of both the uncorrected result and the correction for possible systematic effects. If the only correction that needs to be considered is that arising from the bias of the instrument, its uncertainty should appear in the calibration certificate. The uncertainty of the uncorrected result should be type A evaluated, but usually there is not enough statistical information to use this procedure, for in most cases only one direct reading is taken. However, if it is possible to assume that the behavior of the influence quantities that affect the measurement during service is the same as during calibration, the variance of the calibration results could be used to calculate the uncertainty of the estimated value of the measurand. The formula that is obtained following this approach is developed in the paper, and it shows that calibration certificates should include more information than that which is regularly given.