Depending upon the degree of measurement and the type of measurement being performed, we can offer several alternatives. If the radiation measurements are done with dose projection, we use MIDAS as the dose assessment software to mirror the nuclear power plant facility. We then use RASCAL, an NRC-provided software, to replicate this resultant knowing that a degree of 10 may play a factor. If the measurements are physical in nature, we simply ask someone on-scene to duplicate the measure. If there is no one else on scene or in the vicinity, we will check all of the calibration readings and background levels (which should have been done prior to scene arrival), and if the readings are still discrepant, then we take as many readings as possible and document such to determine later on if the readings were due to instrument error or human error. An average may be taken if the readings are low enough. If they are significantly higher with larger discrepancies, one would take the highest reading that may prompt an action to protect the health and safety of the public.
First, we must make sure our QA equipment is within a calibration period. Then I would repeat measurements using similar ion chambers and electrometers, by similar, I mean same or close physical characteristics and measurement specifications.