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

Table 3

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	2.148	2.953	1.995	2.85	2.059	3.111	0.965	1.054	0.916
	10	2.131	4.054	1.977	3.726	2.043	4.014	0.919	0.99	0.928
	20	2.156	6.656	1.998	5.937	2.063	6.379	0.892	0.958	0.931
	40	2.137	11.59	1.978	10.036	2.041	10.85	0.866	0.936	0.925
10	5	2.23	2.162	2.097	2.003	2.136	2.12	0.927	0.981	0.945
	10	2.254	2.85	2.113	2.579	2.147	2.69	0.905	0.944	0.959
	20	2.264	4.181	2.124	3.758	2.161	3.89	0.899	0.93	0.966
	40	2.27	7.027	2.127	6.268	2.163	6.483	0.892	0.922	0.967
20	5	2.313	1.824	2.18	1.673	2.207	1.725	0.917	0.946	0.97
	10	2.315	2.211	2.183	2.014	2.203	2.054	0.911	0.929	0.98
	20	2.309	2.892	2.181	2.612	2.199	2.654	0.903	0.918	0.984
	40	2.304	4.302	2.176	3.873	2.193	3.939	0.9	0.916	0.983
40	5	2.316	1.642	2.202	1.488	2.209	1.523	0.907	0.927	0.978
	10	2.325	1.835	2.201	1.678	2.213	1.693	0.914	0.922	0.991
	20	2.289	2.1	2.175	1.896	2.181	1.916	0.903	0.913	0.989
	40	2.325	2.907	2.206	2.632	2.214	2.657	0.905	0.914	0.991