Table 2.

Simulation Results when Predictor and Outcome are Sometimes Incorrect

p	nv	ρu,u* = −0.5				ρu,u* = 0				ρu,u* = 0.5
		%-Biasa	$\overline{\text{SE}}$b (×100)	Coverage	MSE (×107)	%-Bias	$\overline{\text{SE}}$ (×100)	Coverage	MSE (×107)	%-Bias	$\overline{\text{SE}}$ (×100)	Coverage	MSE (×107)
0.05	0	−2.29	0.042	0.907	2.4	−4.69	0.043	0.785	4.3	−7.08	0.043	0.619	7.2
	25	−2.62	0.073	0.877	19.6	−0.90	0.086	0.849	20.0	0.76	0.103	0.797	28.7
	50	−2.62	0.061	0.872	7.4	−1.36	0.071	0.869	9.0	0.11	0.084	0.867	11.9
	100	−1.55	0.053	0.916	3.6	−0.81	0.061	0.918	4.8	−0.08	0.070	0.914	6.0
	300	−0.42	0.046	0.948	2.2	−0.20	0.049	0.946	2.6	−0.01	0.053	0.952	2.8
	1000	−0.01	0.032	0.946	1.0	0.03	0.032	0.947	1.0	0.02	0.032	0.953	1.0
0.1	0	−4.54	0.042	0.796	3.9	−9.05	0.043	0.439	10.3	−13.62	0.044	0.147	21.0
	25	−5.51	0.097	0.793	28.4	−1.57	0.124	0.758	34.0	1.63	0.151	0.727	55.4
	50	−3.06	0.075	0.856	10.0	−1.26	0.095	0.853	13.6	0.19	0.115	0.840	19.9
	100	−1.42	0.062	0.917	4.6	−0.62	0.075	0.914	6.9	−0.09	0.089	0.907	9.5
	300	−0.49	0.050	0.948	2.6	−0.16	0.055	0.943	3.3	−0.12	0.061	0.948	3.8
	1000	−0.06	0.032	0.954	1.0	0.02	0.032	0.951	1.0	0.04	0.032	0.949	1.0
0.2	0	−8.23	0.041	0.473	8.5	−16.73	0.043	0.035	30.1	−24.92	0.044	0	64.6
	25	−5.54	0.132	0.787	37.7	−1.78	0.172	0.764	51.2	3.00	0.217	0.795	84.4
	50	−2.42	0.098	0.883	13.0	−0.98	0.128	0.868	20.9	0.84	0.158	0.866	32.1
	100	−1.05	0.076	0.933	6.5	−0.66	0.096	0.924	9.9	0.57	0.117	0.924	15.2
	300	−0.29	0.056	0.951	3.2	−0.23	0.064	0.946	4.2	0.17	0.073	0.944	5.6
	1000	0.08	0.032	0.954	1.0	−0.09	0.032	0.949	1.0	0.02	0.032	0.951	1.0
0.3	0	−11.60	0.040	0.179	15.2	−23.06	0.042	0	55.3	−34.60	0.044	0	122
	25	−4.90	0.162	0.817	46.5	−0.80	0.210	0.804	71.4	4.54	0.266	0.834	115
	50	−2.39	0.117	0.893	17.5	−0.47	0.151	0.886	28.3	2.49	0.191	0.893	45.8
	100	−1.29	0.088	0.925	9.0	−0.19	0.111	0.927	13.7	1.33	0.137	0.925	20.9
	300	−0.58	0.061	0.947	3.8	−0.11	0.071	0.952	5.2	0.33	0.082	0.946	6.8
	1000	−0.08	0.032	0.950	1.0	−0.01	0.032	0.946	1.0	−0.04	0.032	0.955	1.0

^a $\text{Percent}\text{bias}=\left(\frac{\widehat{\beta }-\beta }{\beta }\right)\times 100\%$.

^b $\overline{\text{SE}}=\text{mean}\text{standard}\text{error}\text{of}\text{estimates}$.