Comparison of Prior Setting Methods for Multilevel Model Effect Estimation Based on Small Sample Imbalanced Nested Data in Bayesian Framework

Table 2

Estimated accuracy results of treatment effects and nesting effects for the SSIND under all operating conditions (c = 12).

The number of subjects in each group	ICC	Prior setting	Convergence rate		Cover		Bias		RMSE
			Treatment	Nesting	Treatment	Nesting	Treatment	Nesting	Treatment	Nesting
			Effect	Effect	Effect	Effect	Effect	Effect	Effect	Effect
5	0.05	U	96.9	98.8	97.42	96.66	−148	385	90	47
		G	96.8	97.1	93.49	96.19	−132	−253	91	31
	0.10	U	97.3	99.3	97.64	95.87	−139	623	94	67
		G	98.2	97.9	93.58	97.96	−102	−137	94	28
	0.15	U	96.4	98.9	96.99	91.41	−153	899	98	93
		G	97.5	97.2	91.49	97.12	−125	−14	100	30
10	0.05	U	96.7	98.9	96.28	96.66	−72	−3	71	30
		G	98.1	97.5	92.97	84.72	−55	−326	71	36
	0.10	U	96.6	99.1	96.38	94.35	−80	315	77	46
		G	97.8	97.6	91.62	93.24	−59	−154	78	29
	0.15	U	97	99.2	95.77	91.03	−93	641	83	72
		G	97.1	98.3	92.79	94.51	−72	31	83	31
15	0.05	U	98.2	99.5	97.15	96.28	−44	−120	61	27
		G	97.7	98.2	92.73	76.48	−26	−354	61	37
	0.10	U	97.3	99	96.4	95.66	−39	204	70	38
		G	96.9	98.2	92.88	89.21	−32	−152	69	27
	0.15	U	97.3	98.9	96.61	91.61	−52	551	77	63
		G	97.6	96.7	91.39	94.31	−43	51	77	28
20	0.05	U	96.3	98.9	96.16	90.6	−44	−193	56	28
		G	97.2	97.4	92.7	69.82	−51	−375	57	38
	0.10	U	96.8	98.7	95.45	94.33	−62	145	66	34
		G	97.5	97.9	91.79	89.17	−60	−160	66	25
	0.15	U	97.6	99	95.7	92.63	−90	490	74	57
		G	97.4	98.1	91.27	93.07	−64	49	74	26