Calibration by an Enhanced Full Least Squares Method

This paper deals with the problem of the regression of measured quantities when measurement uncertainty affects both the regressed quantity and the independent variables. A new criterion is given, based on an enhanced full least-squares (FLS) method. The approach shares with the FLS similar metrological foundations, but it differs in that the propagation of the measurement uncertainty is performed numerically and not, as in a conventional FLS, on the basis of the classical law for the propagation of uncertainty based on a first order Taylor expansion. An example of application of the method is given together with a comparison with the regression obtained by the conventional FLS method