Table 9.

Simulation results: power estimates for 10 readers, 200 cases, and effect size (AUC difference) = .066. Notes: the “Actual power” estimate 0.781 is from the last line of Table 8; the “Reliable estimates” column contains power estimates based on the mean parameter estimates from Table 8; the “Sample estimates” column contains the mean of the sample power estimates across the 4000 simulated samples; P25 and P75 are the 25th and 75th percentiles of the sample power estimates; r =number of readers; c =number of cases.

			Power estimated from
r	c	Actual power	Reliable estimates	Sample estimates	P25	P75	P75−P25
3	50	0.781	0.729	0.773	0.639	0.955	.316
3	100	0.781	0.742	0.776	0.658	0.935	.277
3	200	0.781	0.745	0.772	0.653	0.918	.265
5	50	0.781	0.742	0.768	0.635	0.930	.295
5	100	0.781	0.744	0.766	0.655	0.906	.251
5	200	0.781	0.751	0.767	0.666	0.892	.226
10	50	0.781	0.749	0.765	0.649	0.903	.254
10	100	0.781	0.747	0.760	0.662	0.868	.206
10	200	0.781	0.749	0.760	0.675	0.856	.181
mean:			0.744	0.767			.252