Nonparametric Variance Estimation of Strata in Ranked Set Sampling
Authors
FARYAL GOHAR, MUHAMMAD HANIF, USMAN SHAHZAD and NASIR ALI
Abstract
This study is based on ranked set sampling (RSS) with nonparametric stratum variance estimation. To select an extensive number of respondents from the intended audience, RSS uses incorrect rankings for the relevant variable. The sample obtained, which consists of different judgement order statistics, reflects a stratified random sample. Utilizing an updated kernel estimate of distribution function, we modify the previous estimator. Through simulation, the performance of the designed estimator is evaluated and its consistency is shown. It indicates that, under the assumption of perfect or nearly perfect ranking, our approach can be significantly more effective than existing estimators.
Keywords: Ranked Set Sampling, Nonparametric Distribution Function, Estimation of Variance, Kernel Functions, Bandwidth Selectors.
