Table 2.

Simulation 2. Contamination of datasets. Average estimates (standard deviation, standard error; β_{95% CI coverage}) for generalized estimating equations (GEE), Poisson mixed effect (PME), negative binomial mixed effect (NBME), and marginalized zero-altered Poisson (MZAP) models. 500 replications.

y*	Model	βexp = 0.29	σ2 = 0.10
y+0	GEE	0.13 (0.15, 0.19; 0.92)
	PME	0.14 (0.15, 0.19; 0.93)	0.18 (0.08, 0.07)
	NBME	0.13 (0.15, 0.18; 0.93)	0.03 (0.05, 0.09)
	MZAP	0.15 (0.13, 0.16; 0.98)	0.07 (0.04, 0.05)
y+20	GEE	0.22 (0.14, 0.21; 0.99)
	PME	0.21 (0.15, 0.21; 0.99)	0.23 (0.08, 0.09)
	NBME	0.24 (0.15, 0.21; 0.99)	0.07 (0.07, 0.08)
	MZAP	0.24 (0.12, 0.18; 0.98)	0.13 (0.05, 0.06)
y+40	GEE	0.31 (0.14, 0.25; 1.00)
	PME	0.25 (0.15, 0.22; 1.00)	0.29 (0.08, 0.12)
	NBME	0.29 (0.15, 0.23; 1.00)	0.14 (0.07, 0.09)
	MZAP	0.29 (0.12, 0.21; 0.98)	0.20 (0.06, 0.08)
y+60	GEE	0.40 (0.13, 0.29; 1.00)
	PME	0.27 (0.15, 0.24; 1.00)	0.34 (0.08, 0.15)
	NBME	0.32 (0.15, 0.25; 1.00)	0.20 (0.07, 0.11)
	MZAP	0.30 (0.13, 0.23; 0.98)	0.26 (0.06, 0.09)
y+80	GEE	0.48 (0.13, 0.34; 1.00)
	PME	0.29 (0.15, 0.25; 1.00)	0.39 (0.09, 0.18)
	NBME	0.34 (0.15, 0.26; 1.00)	0.24 (0.07, 0.13)
	MZAP	0.29 (0.13, 0.24; 0.99)	0.31 (0.07, 0.11)
y+100	GEE	0.55 (0.13, 0.38; 1.00)
	PME	0.30 (0.15, 0.26; 1.00)	0.42 (0.09, 0.20)
	NBME	0.35 (0.15, 0.27; 1.00)	0.28 (0.07, 0.15)
	MZAP	0.26 (0.15, 0.25; 0.99)	0.35 (0.07, 0.11)