Bayesian Correction for Misclassification in Multilevel Count Data Models

Figure 1

Directed acyclic graph illustrating how the outcome model, exposure model, and measurement model are connected. and are the fallible assessments of the true exposure, . The covariates measured without error, , describe both and the outcome . Lastly, the random effects for both the exposure model and the outcome model are represented by and , respectively.