Bayesian Statistics for Inference over Electoral Behavior

The results provided by the electoral entities do not allow to know the support to the parties by social classes, age groups or races. In this work, the electoral population was divided by age classes and inferences were made about the proportions of support by age for the Alianza País and CREO part...

Descripción completa

Detalles Bibliográficos
Autores: Vivas Paspuel, Atal Kumar, Vivas Paspuel, David Alfredo, Santillán Tituaña, Alberto Benjamín
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2022
País:Ecuador
Institución:Universidad Central del Ecuador
Repositorio:Revista Ingenio
Idioma:español
OAI Identifier:oai:revistadigital.uce.edu.ec:article/3712
Acceso en línea:https://revistadigital.uce.edu.ec/index.php/INGENIO/article/view/3712
Access Level:acceso abierto
Palabra clave:Cadenas de Markov
Monte Carlo
modelos bayesianos
inferencia ecológica
Markov chain
bayesian models
hierarchical models
ecological inference
Descripción
Sumario:The results provided by the electoral entities do not allow to know the support to the parties by social classes, age groups or races. In this work, the electoral population was divided by age classes and inferences were made about the proportions of support by age for the Alianza País and CREO parties, the most important of the presidential contest for Ecuador in 2013. The results of the political contest were taken in contingency tables of RxC size at the district level and through the ecological inference the proportions of support for the candidates by these classes were estimated. Inferences were made through Bayesian techniques by a Dirichlet-Multinomial hierarchical model and Markov Chain Monte Carlo computational methods executed by the RStan package were used.