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

Table 5

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	2.079	2.457	0.225	0.527	2.083	2.717	1.072	0.97	1.106
	10	2.09	3.537	0.224	0.683	2.107	3.691	0.994	0.953	1.044
	20	2.098	5.796	0.24	1.153	2.096	6.005	0.981	0.947	1.036
	40	2.096	10.296	0.237	1.951	2.104	10.53	0.964	0.943	1.023
10	5	2.199	1.908	0.078	0.284	2.205	1.967	1.02	0.989	1.031
	10	2.193	2.551	0.094	0.369	2.185	2.6	1.001	0.982	1.019
	20	2.213	3.824	0.068	0.512	2.211	3.839	0.985	0.981	1.004
	40	2.214	6.349	0.112	0.857	2.208	6.376	0.984	0.98	1.004
20	5	2.276	1.674	0.033	0.231	2.263	1.676	0.997	0.996	1.001
	10	2.286	2.015	0.048	0.26	2.274	2.013	0.992	0.993	0.999
	20	2.256	2.628	0.052	0.344	2.245	2.605	0.987	0.995	0.991
	40	2.265	3.964	0.046	0.463	2.252	3.934	0.984	0.992	0.992
40	5	2.288	1.511	0.034	0.203	2.274	1.497	0.987	0.996	0.991
	10	2.281	1.685	0.038	0.193	2.268	1.658	0.98	0.996	0.984
	20	2.294	2.028	0.018	0.212	2.274	2.001	0.982	0.995	0.987
	40	2.302	2.737	0.011	0.305	2.284	2.703	0.984	0.996	0.988