Fig. 3
From: A Bayesian hierarchical logistic regression model of multiple informant family health histories
Posterior Predictive Distributions for Models 1 and 5 by averaging the set of each family member’s marginalized latent variables (θ) across all informants