Técnica de conteo de ítems en dos ocasiones para poblaciones finitas
Cargando...
Autores
Hernández Romero, Willie Alexander
Director
Tipo de contenido
Document language:
Español
Fecha
Título de la revista
ISSN de la revista
Título del volumen
Documentos PDF
Resumen
Las preguntas sensibles han sido uno de los mayores inconvenientes para los departamentos de estadística dado que la jurisdicción protege fuertemente esta información personal. Por lo tanto, la literatura estadística ha venido desarrollando la técnica de conteo de ítems y de respuesta aleatorizada para resolver este inconveniente. Sin embargo aún no se ha desarrollado la teoría de estas dos metodologías bajo muestreo en varios tiempos. En este documento se propone un estimador para la metodología de respuesta aleatorizada en dos ocasiones con su respectiva varianza y su propuesta maestral. Igualmente, se propone un estimador para la técnica de conteo de ítems y se muestra su varianza teórica; no obstante, se propone que para estimar muestralmente la varianza de la técnica de conteo ítems se ha de usar bootstrap.
Abstract. Sensitive questions have been one of the biggest drawbacks for statistical departments because of the law enforcement in favor of the privacy of every citizen. For instance, statistical literature have been developing the item count technique and randomized response in order to deal with this problematic issue. However, these methodologies have not been developed on different occasions. This document presents an estimator of the randomized response technique on two-occasions, its theoretical variance and an estimator of its variance for finite populations. Likewise, it is proposed an estimator for the item count technique when the sampling on successive occasions is used. Notwithstanding, it is suggested to use bootstrap in order to estimate the variance of this estimator
Abstract. Sensitive questions have been one of the biggest drawbacks for statistical departments because of the law enforcement in favor of the privacy of every citizen. For instance, statistical literature have been developing the item count technique and randomized response in order to deal with this problematic issue. However, these methodologies have not been developed on different occasions. This document presents an estimator of the randomized response technique on two-occasions, its theoretical variance and an estimator of its variance for finite populations. Likewise, it is proposed an estimator for the item count technique when the sampling on successive occasions is used. Notwithstanding, it is suggested to use bootstrap in order to estimate the variance of this estimator