I am developing uncertainty budgets for power sensor calibrations and the subject of power sensor linearity arised. For previous budgets, it was not included because the sensors were being tested at the same power level as they were standardized, 1mW. Upon review of Agilent's Application Note 1449-3, and the Agilent Power Measurement Uncertainty Calculator, power linearity is included no matter what level the sensor is tested and standardized at. This is resulting in very high uncertainties especially for the E-series sensors. I noticed the E9300x series sensors have typical linearity uncertainties listed which could be used instead of the specified values. Does the E441x series have any typical linearity values as well? Is there a way to justify omitting the linearity uncertainty?