Application of the Predictive Posterior PDF to MU and Conformity Assessment in EMC

In Electromagnetic Compatibility (EMC) testing and calibration the measurand cannot always be characterized by an essentially unique value. Sometimes it can be represented only as a distribution of values. This is not consequence of the incomplete (read “inadequate”) definition of the measurand but, on the contrary, it is intrinsic to the definition, in order to make it significant and relevant to applications (i.e., conformity assessment and calculation of measurement uncertainty). The Guide to the Expression of Uncertainty in Measurement (the GUM [1]) does not provide guidance when the measurand is represented by a distribution of values. The aim of this work is to justify the use of predictive posterior density functions for representing measurands characterized by distributions of values. A novel predictive posterior density function in the case of rectangular model is here derived. Examples and applications to conformity assessment and calculation of EMC measurement uncertainty are provided.