Table 1.

Estimated coverage (1 − FCR) from 2000 replicate simulation studies for each of five scenarios

	FDR CI selection method
Scenario	UCB	M.2	BY
A	0.942	0.954	0.973
B	0.965	0.969	0.979
C	0.971	0.972	0.976
D	0.973	0.978	0.976
E${}^{*}$	0.780	0.935	0.969

A: 5 non-null blocks, weak effects; B: 10 non-null blocks, weak effects; C: 5 non-null blocks, strong effects; D: 10 non-null blocks, strong effects; E${}^{*}$: 0 non-null blocks (global null hypothesis).