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

Table 15

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	2.157	2.376	0.23	0.287	2.167	2.613	1.1	0.979	1.124
	10	2.143	3.579	0.242	0.432	2.138	3.753	1.049	0.955	1.098
	20	2.148	5.87	0.22	0.675	2.151	5.995	1.021	0.944	1.082
	40	2.14	10.554	0.238	1.231	2.135	10.641	1.008	0.936	1.077
10	5	2.267	2	0.095	0.151	2.28	2.113	1.057	1.001	1.056
	10	2.268	2.621	0.104	0.19	2.291	2.723	1.039	0.991	1.048
	20	2.25	3.839	0.111	0.259	2.267	3.973	1.035	0.988	1.048
	40	2.264	6.512	0.096	0.422	2.284	6.686	1.027	0.985	1.042
20	5	2.322	1.707	0.064	0.106	2.328	1.767	1.035	1.006	1.029
	10	2.326	2.063	0.049	0.113	2.341	2.104	1.02	0.994	1.026
	20	2.325	2.747	0.045	0.146	2.34	2.801	1.02	0.994	1.025
	40	2.333	4.14	0.047	0.213	2.351	4.206	1.016	0.993	1.023
40	5	2.342	1.566	0.007	0.088	2.359	1.584	1.011	0.996	1.015
	10	2.339	1.728	0.031	0.1	2.35	1.752	1.014	0.999	1.015
	20	2.333	2.07	0.022	0.099	2.341	2.102	1.015	1	1.015
	40	2.353	2.803	0.034	0.139	2.363	2.834	1.011	0.997	1.014