Rescaled Spatial Bootstrap Variance Estimation of Spatial Estimator of Finite Population Parameters under Ranked Set Sampling

Abstract

Ranked Set Sampling (RSS) is preferred over Simple Random Sampling (SRS) when measuring an observation is expensive or time consuming, but can be easily ranked at a negligible cost. Biswas et al. (2015) proposed a Spatial Estimator (SE) of population mean under RSS through prediction approach incorporating spatial dependency among sampling units of a spatial finite population. In this present article, an attempt has been made to propose bootstrap techniques viz. Rescaled Spatial Stratified Bootstrap (RSSB) and Rescaled Spatial Clustered Bootstrap (RSCB) methods for unbiased variance estimation of the SE under RSS from finite populations. Simulation study reveals that both the proposed methods give approximately unbiased estimation of variance of the SE under RSS for different combination of sample and bootstrap sample sizes, but while considering relative stability, RSSB method was found to be more stable