Generalized Mixed Models - an application to longitudinal data of citrus canker

• Autor(a) principal: Serenini, João Pedro
• Data de Publicação: 2019
• Outros Autores: Guedes, Terezinha Aparecida, Roberto, Juliana Glória Franco, Nunes, William Mário de Carvalho
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Acta scientiarum. Technology (Online)
• Texto Completo: http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/41646
• topic: range; repeated measurements/measures; mixed models.; Probabilidade e Estatística

There are several techniques available for longitudinal data analysis. In the last decade, much emphasis has been placed on generalized mixed models. The present work is dedicated to give an overview of this technique, with emphasis on the formulation, interpretation and inference of the model. The guidelines are given for statistical practice in general. This form of modeling was applied to data from an experiment to evaluate the resistance of 12 varieties of sweet orange to citrus canker. The experiment consisted of provoking lesions on the leaves of orange trees and monitoring the diameter of the lesion over time. The adjustment of the observed data to the proposed model provided reliable results, since the assumptions necessary for the validity of the model were satisfied. Therefore, it can be said that this methodology is adequate to model the data, since it allowed the detection of the varieties more susceptible to citrus canker.