Process designed to remove inconsistencies in measured or computed quantities by applying derived corrections to compensate for random or accidental errors.