Bayesian latent class models for prevalence estimation and diagnostic test evaluation

This document was developped as part of a 2-day course on Bayesian latent class models developed with my colleague Juan Carlos Arango Sabogal. Bayesian statistical methods have been widely applied in veterinary science and epidemiological research. In particular, Bayesian latent class variable models (LCM) have been shown to be useful for estimating disease prevalence or diagnostic test sensitivity and specificity in the absence of a gold standard test. In this document we introduce these methods and illustrate their application using R and JAGS with concrete veterinary examples. There is a series of exercises (named Exercise 1 - Proportions to Exercise 5 - Accuracy varying with covariate) that can be done; they can be downloaded from the course repository on my Github account.