Table II.

Power, Type I Error Rates and Sample Size Characteristics for Method C Applied to Parallel Bioequivalence Designs

CV	N 1	Power	Type I error rate	AvgN	F5	F50	F95
0.1	48	1.000	0.0504	48.0	48	48	48
0.2	48	0.908	0.0499	48.2	48	48	48
0.3	48	0.827	0.0479	74.4	48	72	122
0.4	48	0.806	0.0410	148.5	48	152	212
0.5	48	0.790	0.0311	231.6	160	228	316
0.6	48	0.787	0.0298	318.5	218	314	434
0.7	48	0.785	0.0295	412.4	282	406	562
0.8	48	0.784	0.0297	511.0	350	504	696
0.9	48	0.784	0.0297	612.2	420	604	836
1.0	48	0.784	0.0297	714.7	490	704	976
0.1	120	1.000	0.0502	120.0	120	120	120
0.2	120	0.999	0.0502	120.0	120	120	120
0.3	120	0.938	0.0506	120.0	120	120	120
0.4	120	0.830	0.0454	131.2	120	120	178
0.5	120	0.824	0.0478	185.8	120	200	278
0.6	120	0.815	0.0467	280.9	120	308	388
0.7	120	0.804	0.0368	401.0	280	410	504
0.8	120	0.797	0.0303	509.9	406	508	624
0.9	120	0.796	0.0298	611.9	488	608	748
1.0	120	0.795	0.0294	714.6	570	710	874

The evaluation for bioequivalence is based on an assumption of unequal variances. Both test and reference were simulated with the CVs indicated

N ₁ sample size at stage 1, AvgN average total sample size, F5 5th percentile of total sample sizes, F50 50th percentile of total sample sizes, F95 95th percentile of total sample sizes