Regression Estimators Using Stratified Ranked Set Sampling

Presenters/Authors

Arpita Chatterjee, Georgia Southern UniversityFollow
Hani M. Samawi, Georgia Southern UniversityFollow
Lili Yu, Georgia Southern UniversityFollow
Daniel F. Linder, Georgia Southern UniversityFollow
Jingxian Cai, Georgia Southern UniversityFollow
Robert L. Vogel, Georgia Southern UniversityFollow

Document Type

Presentation

Presentation Date

6-16-2014

Abstract or Description

This article is intended to investigate the performance of two types of stratified regression estimators, namely the separate and the combined estimator, using stratified ranked set sampling (SRSS), introduced by Samawi (1996). The expressions for mean and variance of the proposed estimates are derived and are shown to be unbiased. A simulation study is designed to compare the efficiency of SRSS relative to other sampling procedure under varying model scenarios. Our investigation indicates that the regression estimator of the population mean obtained through an SRSS becomes more efficient than the crude sample mean estimator using stratified simple random sampling. These findings are also illustrated with the help of a data set on bilirubin levels in babies in a neonatal intensive care unit.

Sponsorship/Conference/Institution

International Chinese Statistical Association (ICSA) and the Korean International Statistical Society (KISS) Joint Applied Statistics Symposium

Location

Portland, OR

Recommended Citation

Chatterjee, Arpita, Hani M. Samawi, Lili Yu, Daniel F. Linder, Jingxian Cai, Robert L. Vogel. 2014. "Regression Estimators Using Stratified Ranked Set Sampling." Biostatistics Faculty Presentations. Presentation 23.
https://digitalcommons.georgiasouthern.edu/biostat-facpres/23