Sounds difficult to test for. Thinking about a Pilot I think there would be data points you could measure on the job that predict likelihood of making a bad judgement. For example heart rate variability, cortisol levels, skin moisture.
I think the better solution is to abstract out the judgement into a formula that the pilot must follow, so not courting "gethereitis"
I think the better solution is to abstract out the judgement into a formula that the pilot must follow, so not courting "gethereitis"