Table 2.

Power (%), Scenario 1: Uncensored data, 4 outcomes, θ_ϕ= (.03,.08,.16,.28). n, m = sample size per strata in each group, respectively. T_O = O'Brien original test; $T_O^U=\mathit{\text{Proposed O'Brien test}}$; $T_O^{U,\mathit{\text{Ad}}}=\mathit{\text{Proposed Adaptive O'Brien test}}$; $T_O^{U,\mathit{\text{Opt}}}=\mathit{\text{Optimal O'Brien Test}}$. w_opt = Optimal weight vector. ρ = correlation between outcomes.

ρ	No. Strata	n	m	TO	$T_O^U$	$T_O^{U,\mathit{\text{Ad}}}$	$T_O^{U,\mathit{\text{Opt}}}$	wopt
0	2	20	20	55.6	54.1	52.6	71.6	(.053, .136, .281, .530)
		40	40	84.5	84.2	84.8	95.6
	4	10	10	55.5	56.7	53.2	74.3
		20	20	85.5	85.7	83.9	95.7
0.2	2	30	30	54.5	53.7	59.4	80.1	(0, .024, .276, .700)
		60	60	84.1	83.8	90.4	98.2
	4	15	15	53.7	54.4	61.3	82.7
		30	30	84.2	84.4	90.7	98.3
0.5	2	30	30	38.7	37.7	52.6	77.0	(0, 0, .094, .906)
		60	60	66.4	66.1	84.2	96.8
	4	15	15	39.4	39.7	56.5	77.9
		30	30	66.6	66.5	86.5	96.9
0.8	2	30	30	30.8	30.2	50.1	75.8	(0, 0, 0, 1)
		60	60	53.2	52.8	79.6	97.4
	4	15	15	30.7	30.9	57.0	78.2
		30	30	52.6	52.9	84.2	97.0