Modelos de regressão binomial correlacionada
Abstract
In this thesis, a class of correlated binomial regression models is proposed. The model is based on the generalized binomial distribution proposed by Luceño (1995) and Luceño & Ceballos (1995). The regression structure is modeled by using four different link functions and the dependence between the Bernoulli trials is modeled by using three different correlation structures. A data augmentation scheme is used in order to overcome the complexity of the mixture likelihood. Frequentist and Bayesian approaches are used in the model fitting process. A diagnostics analysis is provided in order to check the underlying model assumptions and to identify the presence of outliers and/or influential observations. Simulation studies are presented to illustrate the performance of the developed methodology. A real data set is analyzed by using the proposed models. Also the correlated binomial regression models is extended to include measurement error in a predictor. This new class of models is called additive normal structure correlated binomial regression models. The inference process also includes a data augmentation scheme to overcome the complexity of the mixture likelihood.