Second moment of the Benjamini-Hochberg false discovery proportion

ID: second-moment-of-the-benjamini-hochberg-false-discovery-proportion

For independent p-values, with true-null values independent and uniform on , let denote the Benjamini-Hochberg procedure rejection count after replacing one true-null value by zero. The Benjamini-Hochberg leave-one-out identity gives
For distinct true-null indices, the Benjamini-Hochberg leave-two-out identity gives
Indeed, conditional on the remaining values and , the two independent uniform values both meet their threshold with probability , cancelling the denominator . Expanding the square of the false rejection count into its diagonal terms and ordered distinct pairs, and using symmetry among the true-null values, proves
If no null is true, both sides are zero; if exactly one is true, the pair contribution is absent. This formula complements the exact false discovery rate by describing second-moment variability.

New to topics? Read the docs here!