Formación de Recurso Humano para la Ctel: Proyecto ejecutado con investigadores en empresas, industrias y Estado
Estimadores de regresión beta para la media poblacional en muestreo probabilístico: teoría y aplicaciones
Fecha
2019-08Autor
Pacheco Lopez, Mario José
Tellez Piñerez, Cristian Fernando
Institución
Resumen
In probabilistic sampling it is common to have auxiliary information available for the elements of the population of interest. Regression estimators are a type of estimator that makes efficient use of this auxiliary information. In this work the estimation of the mean of distributed beta variables is considered. The construction of a beta regression estimator (BREG) is proposed for situations in which the variable of interest is restricted to an interval (0.1) and is related to auxiliary variables through a linear or nonlinear regression structure. Important cases of inflated and variable dispersion beta 0-1 regression models are also discussed. We will illustrate the methodology through an application with information from DANE from household surveys, particularly the quality of life survey.