Comparison of logistic-regression based methods for simple mediation analysis with a dichotomous outcome variable

Table 2 Bias and efficiency yielded by the three compared methods for models with a continuous mediator

		Multiple regression and SEM								Potential outcomes
		Crude		y-standardization		Full-standardization		Standardized logistic solution
Y prev		bias	MSE	bias	MSE	bias	MSE	bias	MSE	bias	MSE
0.5	Indirect effect									−0.001	0.002
	ab	−0.001	0.002	−0.001	0.000	−0.000	0.001	−0.001	0.002
	c-c’	−0.068	0.006	−0.008	0.000	−0.036	0.002	−0.021	0.003
	Proportion mediated									0.003	0.003
	ab/(ab + c’)	0.003	0.003	0.003	0.003	0.065	0.009	0.003	0.003
	ab/c	0.032	0.005	0.009	0.003	0.073	0.010	0.011	0.003
	1-(c’/c)	−0.044	0.004	− 0.008	0.003	−0.008	0.003	−0.010	0.003
0.3	Indirect effect									0.000	0.003
	ab	0.000	0.003	−0.000	0.000	−0.000	0.001	0.000	0.003
	c-c’	−0.061	0.005	−0.005	0.001	−0.033	0.002	−0.012	0.003
	Proportion mediated									0.001	0.004
	ab/(ab + c’)	0.001	0.004	0.001	0.004	0.064	0.010	0.001	0.004
	ab/c	0.027	0.005	0.005	0.004	0.069	0.010	0.006	0.004
	1-(c’/c)	−0.041	0.005	−0.006	0.004	−0.006	0.004	−0.007	0.004
0.1	Indirect effect									0.003	0.006
	ab	0.003	0.006	0.001	0.001	0.001	0.001	0.003	0.006
	c-c’	−0.036	0.005	0.005	0.001	−0.023	0.002	0.016	0.003
	Proportion mediated									0.006	0.008
	ab/(ab + c’)	0.006	0.008	0.006	0.008	0.070	0.017	0.006	0.008
	ab/c	0.023	0.010	0.002	0.008	0.065	0.015	0.001	0.008
	1-(c’/c)	−0.020	0.008	0.011	0.009	0.011	0.009	0.013	0.009

Abbreviations: SEM structural equation modeling, Y prev outcome prevalence, MSE mean squared error

ISSN: 1471-2288