Binary data in the presence of covariates and misclassifications: a Bayesian approach

摘要: In this paper we introduce a Bayesian analysis for binary data in the presence of covariates and misclassifications. As special situation diagnostic medical testing, obtain inferences sensitivity specificity covariates. We consider where individuals can be verified or unverified about their real disease status after test. When part even all are not verified, usually have great difficulties to get classical inference resultsfor parameters interest. For situation, introduction latent variables gives good alternative deal with missing under approach, specially using Markov chain monte Carlo (MCMC) methods posterior summaries illustrate proposed methodology on three sets.