. Author manuscript; available in PMC: 2018 Apr 23.

Published in final edited form as: Biometrics. 2015 Sep 22;72(1):39–45. doi: 10.1111/biom.12412

Table 1.

Empirical type I error rate and power of each method obtained from 1000 replicates with nominal significance level of 0.05.

Sample size	Test	Type I error rate	Power
Sample size	Test	Type I error rate	(a)	(b)	(c)	(d)
n₁ = n₂ = n₃ = 50	U	0.048	0.281	0.289	0.132	0.044
	V	0.043	0.036	0.079	0.845	0.721
	UV	0.047	0.187	0.244	0.832	0.628
	LR	0.048	0.273	0.310	0.185	0.066
	GW	0.052	0.235	0.268	0.449	0.215

n₁ = n₂ = n₃ = 100	U	0.047	0.545	0.495	0.301	0.072
	V	0.036	0.049	0.137	0.995	0.965
	UV	0.054	0.431	0.464	0.995	0.935
	LR	0.051	0.533	0.521	0.349	0.092
	GW	0.053	0.469	0.479	0.777	0.356