Auxiliary variables in multiple imputation in regression with missing X: a warning against including too many in small sample research

Hardt, Jochen; Herke, Max; Leonhart, Rainer

doi:10.1186/1471-2288-12-184

Table 3 Results of the simulations on linear regressions, MCAR

From: Auxiliary variables in multiple imputation in regression with missing X: a warning against including too many in small sample research

Percent missing		Auxiliary variable correlation	n = 50				n = 100				n = 200
			Bias	SE	Std. Bias	RMSE	Bias	SE	Std. Bias	RMSE	Bias	SE	Std. Bias	RMSE
20	CC	-	0.02	0.34	0.06	0.19	0.01	0.24	−0.05	0.12	−0.01	0.16	−0.05	0.08
	MI+0	-	0	0.30	0.01	0.12	0	0.20	0.01	0.09	0	0.14	−0.01	0.06
	MI+10	0.1	−0.06	0.31	−0.17	0.14	−0.04	0.21	−0.16	0.09	−0.01	0.14	−0.06	0.06
	MI+20	0.1	−0.16	0.32	−0.49	0.19	−0.06	0.21	−0.25	0.11	−0.02	0.14	−0.14	0.06
	MI+40	0.1	−0.13	0.33	−0.41	0.18	−0.17	0.22	−0.75	0.18	−0.06	0.15	−0.42	0.08
	MI+80	0.1	−0.15	0.32	−0.46	0.19	−0.19	0.25	−0.77	0.20	−0.17	0.15	−1.09	0.17
	MI+10	0.5	−0.03	0.40	−0.06	0.18	0	0.27	0.01	0.11	0	0.19	−0.01	0.08
	MI+20	0.5	−0.06	0.40	−0.13	0.19	−0.03	0.28	−0.09	0.12	−0.02	0.19	−0.08	0.08
	MI+40	0.5	−0.09	0.41	−0.23	0.21	−0.07	0.28	−0.24	0.13	−0.03	0.19	−0.15	0.09
	MI+80	0.5	−0.06	0.40	−0.14	0.20	−0.30	0.27	−1.13	0.30	−0.08	0.19	−0.42	0.11
50	CC	-	0.12	0.62	0.18	0.43	−0.02	0.38	−0.08	0.28	0.04	0.27	0.15	0.19
	MI+0	-	−0.04	0.40	−0.04	0.28	−0.01	0.26	0.04	0.18	−0.01	0.18	−0.01	0.12
	MI+10	0.1	−0.20	0.43	−0.44	0.31	−0.10	0.28	−0.34	0.20	−0.04	0.18	−0.19	0.13
	MI+20	0.1	−0.49	0.42	−1.25	0.50	−0.18	0.29	−0.64	0.22	−0.10	0.19	−0.52	0.14
	MI+40	0.1	−0.35	0.41	−0.94	0.40	−0.49	0.29	−1.77	0.49	−0.19	0.19	−1.02	0.20
	MI+80	0.1	−0.33	0.40	−0.89	0.38	−0.35	0.28	−1.38	0.38	−0.48	0.20	−2.55	0.48
	MI+10	0.5	−0.06	0.53	−0.09	0.31	0	0.35	0.06	0.23	−0.01	0.24	0	0.16
	MI+20	0.5	−0.34	0.51	−0.72	0.37	−0.03	0.36	−0.07	0.21	−0.03	0.24	−0.10	0.15
	MI+40	0.5	−0.20	0.47	−0.48	0.35	−0.31	0.35	−0.91	0.32	−0.06	0.24	−0.26	0.15
	MI+80	0.5	−0.22	0.48	−0.51	0.36	−0.19	0.33	−0.68	0.31	−0.30	0.24	−1.31	0.30

Back to article page

ISSN: 1471-2288

Contact us

General enquiries: journalsubmissions@springernature.com

BMC Medical Research Methodology

Contact us