Confidence Limits and Their Robustness

Abstract Confidence limits are common place in physics analysis. Great care must be taken in their calculation and use, especially in cases of limited statistics when often one-sided limits are quoted. In order to estimate the stability of the confidence levels to addition of more data and/or change of cuts, we argue that the variance of their sampling distributions be calculated in addition to the limit itself. The square root of the variance of their sampling distribution can be thought of as a statistical error on the limit. We thus introduce the concept of statistical errors of confidence limits and argue that not only should limits be calculated but also their errors in order to represent the results of the analysis to the fullest. We show that comparison of two different limits from two different experiments becomes easier when their errors are also quoted. Use of errors of confidence limits will lead to abatement of the debate on which method is best suited to calculate confidence limits.