A measure of confidence in precision of a set of sample data. For a given numerical value of fiducial interval, the number of samples required from a given deposit to give an accurate measure of its value can be determined.