Researchers have shown that ranked set sampling outperforms simple random sampling in many situations by reducing the variance of a parameter estimator, thereby providing the same accuracy with a smaller sample size than is needed in simple random sampling. In this paper, we have explored the feasibility of using RSS (rank set sampling) within the framework of stratified sampling. Rather than selecting a simple random sample within each stratum, as is done in SSRS (stratified simple random sampling), we have taken a ranked set sample within each stratum. This procedure combines the variance reduction that arises from stratifying the population with the increased precision RSS holds over SRS (simple random sampling). Theoretical and simulation study are presented. It appears that using rank set procedure under stratification results in reduction of standard errors thus provides more efficient results than stratified random sampling.
展开▼
