Variance estimation using memory type estimators based on EWMA statistic for time scaled surveys in stratified sampling.

ABSTRACT: In this article, we have proposed memory-type exponential and non-exponential estimators for population variance based on exponentially weighted moving average (EWMA) statistic in stratified sampling. We drive mathematical expressions for mean square errors of the proposed estimators using Taylor and exponential expansions. Our analysis demonstrates that the proposed memory-type estimators outperform the conventional estimators under stratification, particularly, when the information of previous sample is utilized. The performances of the proposed estimators are evaluated mathematically by deriving the conditions in which the memory-type estimators would perform better than their corresponding conventional estimators under stratification. Through an extensive simulation, we evaluated the p

SUBMITTER: Tariq MU 

PROVIDER: S-EPMC11535251 | biostudies-literature | 2024 Nov

REPOSITORIES: biostudies-literature