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

Table 9

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	2.738	11.782	2.365	5.657	2.453	6.057	0.48	0.934	0.514
	10	2.716	11.193	2.203	6.258	2.283	6.527	0.559	0.959	0.583
	20	2.684	16.996	2.234	11.033	2.319	11.498	0.649	0.96	0.676
	40	2.669	26.463	2.184	17.785	2.272	18.639	0.672	0.954	0.704
10	5	2.905	5.869	2.413	3.397	2.451	3.439	0.579	0.988	0.586
	10	2.994	9.838	2.514	7.659	2.554	7.703	0.779	0.994	0.783
	20	2.774	7.637	2.339	5.097	2.373	5.189	0.667	0.982	0.68
	40	2.933	17.461	2.473	13.642	2.512	13.83	0.781	0.986	0.792
20	5	3.142	6.441	2.678	4.829	2.699	4.857	0.75	0.994	0.754
	10	2.907	4.578	2.431	3.145	2.451	3.151	0.687	0.998	0.688
	20	2.934	5.54	2.415	3.89	2.431	3.921	0.702	0.992	0.708
	40	3.009	9.454	2.52	7.155	2.537	7.204	0.757	0.993	0.762
40	5	3.142	6.441	2.678	4.829	2.699	4.857	0.75	0.994	0.754
	10	2.907	4.578	2.431	3.145	2.451	3.151	0.687	0.998	0.688
	20	2.934	5.54	2.415	3.89	2.431	3.921	0.702	0.992	0.708
	40	3.009	9.454	2.52	7.155	2.537	7.204	0.757	0.993	0.762