TABLE 2.

Simulation Designs A–C

	True model		MISSPECIFIED MODEL
Models	Class	Linear predictor and parameter values	Class	Linear predictor
Simulation 3
		$\beta =(-0.20,0.30,0.15,0.22,0.15,-0.15),{\alpha }_0=(9,-0.05,0.05,-0.05),{\alpha }_1=(10,-0.05,0.05,-0.05)$
		$1,M_1=\exp (0.10X_1),M_2=X_2(1+X_1)+10,M_3={(0.04X_3+0.60)}^2,M_4={(X_4+20)}^2$
Design A
PS	Binomial, logit	$1,X_1,X_2,X_3,X_4,X_5={X_1}^2$	Binomial, logit	$1,M_1,M_2,M_3,M_4$
OR	Poisson, log	$1,X_1,X_2,X_3,X_4$	Gaussian, identity	$1,M_1,M_2,M_3,M_4$
		$\beta =(-0.40,0.60,0.30,0.44,0.30,-0.30),{\alpha }_0=(9,-0.05,0.05,-0.05),{\alpha }_1=(10,-0.05,0.05,-0.05)$
		$1,M_1=\exp (0.10X_1),M_2=X_2(1+X_1)+10,M_3={(0.04X_3+0.60)}^2,M_4={(X_4+20)}^2$
Design B
PS	Binomial, logit	$1,X_1,X_2,X_3,X_4,X_5={X_1}^2$	Binomial, logit	$1,M_1,M_2,M_3,M_4$
OR	Poisson, log	$1,X_1,X_2,X_3,X_4$	Gaussian, identity	$1,M_1,M_2,M_3,M_4$
		$\beta =(-0.60,0.90,0.45,0.66,0.45,-0.45),{\alpha }_0=(9,-0.05,0.05,-0.05),{\alpha }_1=(10,-0.05,0.05,-0.05)$
		$1,M_1=\exp (0.10X_1),M_2=X_2(1+X_1)+10,M_3={(0.04X_3+0.60)}^2,M_4={(X_4+20)}^2$
Design C
PS	Binomial, logit	$1,X_1,X_2,X_3,X_4,X_5={X_1}^2$	Binomial, logit	$1,M_1,M_2,M_3,M_4$
OR	Poisson, log	$1,X_1,X_2,X_3,X_4$	Gaussian, identity	$1,M_1,M_2,M_3,M_4$