Table 2.

Comparison of yearly income between SRH groups without and after imputation

Imputation method	SRH				P value
	Good or better		Fair or poor
	N	Yearly incomea Mean ± SD	N	Yearly income Mean ± SD
Training set (N = 3,121)
LD	1,115	26.39 ± 17.29	706	21.27 ± 14.77	<0.0001
MS	1,770	25.65 ± 13.76	1,351	22.77 ± 10.79	<0.0001
RB	1,770	25.56 ± 14.20	1,351	21.17 ± 11.30	<0.0001
RBE	1,770	25.62 ± 16.75	1,351	21.61 ± 15.52	<0.0001
MI	1,770	25.99 ± 13.74	1,351	22.15 ± 10.72	<0.0001
Testing set (N = 1,041)
LD	391	26.27 ± 16.94	219	21.83 ± 13.98	<0.0001
MS	606	25.70 ± 13.62	435	23.24 ± 10.01	0.001
RB	606	25.40 ± 14.09	435	21.37 ± 10.63	<0.0001
RBE	606	25.16 ± 16.33	435	20.90 ± 14.98	<0.0001
MI	606	25.88 ± 13.61	435	22.87 ± 9.97	0.0001

LD listwise deletion (no imputation), MI multiple imputation, MS mean substitution, RB regression-based single imputation, RBE regression-based single imputation with error term

^aYearly income per $1000