Development of New Robust Optimal Score Function for the Weibull Distributed Error Term in Multilevel Models

Table 11

Bias and MSE of using Weibull, Wilcoxon, and REML methods when  = 3..

		REML		Weibull		Wilcoxon		Precision
N2	N1	Bias	MSE	Bias	MSE	Bias	MSE	Weib/REML	Weib/Wil	Wil/REML
5	5	2.207	2.243	0.212	1.237	2.342	2.535	1.13	0.989	1.143
	10	2.152	2.648	0.183	1.618	2.265	2.749	1.038	0.998	1.04
	20	2.163	3.975	0.245	2.737	2.284	4.055	1.02	1.029	0.991
	40	2.224	7.258	0.228	4.338	2.368	7.187	0.99	1.048	0.944
10	5	2.339	1.457	0.108	0.735	2.505	1.513	1.038	1.017	1.021
	10	2.325	1.792	0.076	0.812	2.49	1.829	1.021	1.045	0.977
	20	2.314	2.55	0.086	1.264	2.463	2.589	1.015	1.078	0.942
	40	2.307	4.145	0.093	1.96	2.456	4.18	1.008	1.086	0.929
20	5	2.32	1.123	0.033	0.561	2.498	1.144	1.019	1.049	0.972
	10	2.347	1.286	0.073	0.687	2.518	1.301	1.012	1.075	0.941
	20	2.332	1.693	0.059	0.846	2.512	1.698	1.003	1.082	0.927
	40	2.376	2.601	0.041	1.266	2.543	2.622	1.008	1.094	0.922
40	5	2.387	1.009	0.036	0.562	2.567	1	0.991	1.055	0.94
	10	2.396	1.126	0.018	0.492	2.587	1.119	0.994	1.076	0.924
	20	2.359	1.273	0.003	0.559	2.529	1.28	1.005	1.094	0.919
	40	2.372	1.728	-0.011	0.764	2.551	1.734	1.003	1.095	0.917