. Author manuscript; available in PMC: 2021 Feb 12.

Published in final edited form as: Bayesian Anal. 2019 Jun 11;14(3):829–855. doi: 10.1214/18-ba1132

Table 2:

Parameter estimates and 95% intervals for random effects ZINB model in simulation study 2.

Model Component	Parameter	Simulated Value	Estimate (95% Interval)
Model Component	Parameter	Simulated Value	Proposed Model^†	MLE^‡
Binary	β₁₀	0.25	0.06 (−0.17, 0.30)	0.05 (−0.19, 0.28)
	β₁₁	−0.25	−0.27 (−0.47, −0.08)	−0.27 (−0.47, −0.08)
	β₁₂	0.25	0.29 ( 0.21, 0.38)	0.29 ( 0.20, 0.38)
Count	β₂₀	0.50	0.39 ( 0.20, 0.58)	0.39 ( 0.20, 0.58)
	β₂₁	−0.25	−0.11 (−0.27, 0.03)	−0.12 (−0.27, 0.03)
	β₂₂	0.25	0.26 ( 0.21, 0.31)	0.26 ( 0.21, 0.30)
	r	1.25	1.32 ( 1.17, 1.54)	1.35 ( 1.19, 1.53)
Random Effects	Γ₁₁ = Var(ϕ_1i)	0.50	0.54 ( 0.33, 0.79)	0.52 ( 0.28, 0.76)
	Γ₂₂ = Var(ϕ_2i)	0.75	0.77 ( 0.64, 0.90)	0.76 ( 0.63, 0.89)
	Γ₁₂ = Cov(ϕ_1i, ϕ_2i)	0.25	0.30 ( 0.18, 0.41)	0.31 ( 0.19, 0.42)

^†

Posterior means and 95% credible intervals for proposed Bayesian ZINB model.

^‡

MLEs and 95% confidence intervals obtained using SAS Proc NLMIXED.