TABLE 2.

Estimates of size and power for three log rank tests under simulation design 2. LR = log rank, SLR = stratified log rank, CWLR = cluster size-weighted log rank, GWLR = group size-weighted log rank, and Frailty = frailty model

$\mathit{N}$	Censoring	Test	Size	Power
				$\delta =0.05$	$\delta =0.10$	$\delta =0.15$
30	Light	LR	0.318	0.470	0.716	0.865
		SLR	0.307	0.435	0.684	0.854
		CWLR	0.076	0.148	0.337	0.567
		GWLR	0.058	0.110	0.281	0.509
		Frailty	0.349	0.494	0.743	0.892
	Heavy	LR	0.370	0.534	0.745	0.900
		SLR	0.337	0.495	0.720	0.887
		CWLR	0.082	0.167	0.354	0.576
		GWLR	0.057	0.140	0.297	0.532
		Frailty	0.381	0.542	0.774	0.922
50	Light	LR	0.390	0.653	0.879	0.974
		SLR	0.349	0.602	0.853	0.973
		CWLR	0.064	0.209	0.514	0.794
		GWLR	0.053	0.151	0.436	0.741
		Frailty	0.396	0.656	0.899	0.983
	Heavy	LR	0.339	0.609	0.846	0.961
		SLR	0.326	0.540	0.818	0.951
		CWLR	0.066	0.206	0.499	0.768
		GWLR	0.049	0.148	0.428	0.711
		Frailty	0.369	0.614	0.865	0.969
100	Light	LR	0.251	0.962	1.000	1.000
		SLR	0.044	0.819	1.000	1.000
		CWLR	1	1	1	1
		GWLR	1	1	1	1
		Frailty	0.076	0.898	1.000	1.000
	Heavy	LR	0.237	0.919	1.000	1.000
		SLR	0.052	0.702	0.999	1.000
		CWLR	1	1	1	1
		GWLR	1	1	1	1
		Frailty	0.074	0.821	1.000	1.000