Estimation in a Generalized Correlated Binomial Model

Autor(es) e Instituição: 
Rubiane Maria Pires - UFSCar
Carlos Alberto Ribeiro Diniz - UFSCar
José Galvão Leite - UFSCar
Rubiane Maria Pires

A generalized correlated binomial model, which considers different correlations for the Bernoulli trails, is proposed. The maximum likelihood estimators for the parameters are obtained by direct maximization of the likelihood function using the Monte Carlo approach. To illustrate the proposed model we use a transgenic soybean dataset. Nonparametric bootstrap method is used to find confidence interval for the model parameters.