Confidence intervals for the population standard deviation: simple random sampling vs. ranked set sampling
Abstract. Ranked set sampling (RSS) is the cost-efficient sampling procedure. This procedure gives more efficient estimators of the population parameters than the procedure based on simple random sampling (SRS), with the same sample size. In this paper, we compare the coverage probabilities of confidence intervals for the population standard deviation using the simple random sampling and the ranked set sampling. The following confidence intervals are considered: the exact, the Bonett, the Steve large sample normal approximations, the log asymptotic approximation and the adjusted degrees of freedom. The results for the Gamma, Log-normal and Exponential distributions and for the real data set are presented. The simulation study shows that the results obtained using the ranked set sampling are better than those using the simple random sampling.
engleski
2018
© All rights reserved
Keywords: population standard deviation; confidence interval; coverage accuracy; ranked set sampling; simple random sampling