Rescaling bootstrap technique for variance estimation for ranked set samples in finite population

Ranked Set Sampling (RSS) is preferred to Simple Random Sampling (SRS) when measuring an observation is expensive or time-consuming, while ranking small subset of observations is relatively easy. Estimating the variance of RSS estimator has been found cumbersome under finite population. In this stud...

Full description

Saved in:

Bibliographic Details
Published in:	Communications in statistics. Simulation and computation Vol. 49; no. 10; pp. 2704 - 2718
Main Authors:	Biswas, Ankur, Rai, Anil, Ahmad, Tauqueer
Format:	Journal Article
Language:	English
Published:	Philadelphia Taylor & Francis 02-10-2020 Taylor & Francis Ltd
Subjects:	Cluster based rescaling bootstrap Random sampling Ranked set sampling Rescaling Rescaling bootstrap Strata based rescaling bootstrap Variance
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Ranked Set Sampling (RSS) is preferred to Simple Random Sampling (SRS) when measuring an observation is expensive or time-consuming, while ranking small subset of observations is relatively easy. Estimating the variance of RSS estimator has been found cumbersome under finite population. In this study, we propose two rescaling bootstrap variance estimation techniques in RSS under finite population framework viz. Strata Based Rescaling Bootstrap (SBRB) and Cluster Based Rescaling Bootstrap (CBRB) methods. Simulation as well as real data application results suggest that SBRB method performs better than CBRB method for different combination of set size (m) and number of cycles (r).
ISSN:	0361-0918 1532-4141
DOI:	10.1080/03610918.2018.1527349