Determination Of An Efficient Sampling Design For Rare And Clustered Population Using Design Based Estimators

ABSTRACT

A sampling design that provides estimates of population mean and abundance with small

variance is important to researchers. Estimates that are accurate even with minimal sampling

efforts allow researchers to easily and confidently investigate rare populations. In the

determination of efficient sampling design for rare and clustered population, mean square errors

have been applied in many previous research works. However, this method only captures the

variability of the estimator and fails to capture their reliability. This study obtained the interval

estimates based on the design based estimators, the HT and HH estimators. The study examined

the behavior of the Horvitz Thompson (HT) and Hansen Hurwitz (HH) estimators under the

ordinary adaptive cluster sampling design (ACS) and adaptive cluster sampling with data driven

stopping rule (ACS’) design using artificial population that is designed to have all the

characteristics of a rare and clustered population and another population that does not have those

characteristics. The efficiency of HT and HH estimators were used to determine the most

efficient design in estimation of population mean in rare and clustered population. The coverage

probability confidence intervals of population mean based on HT estimators and the HH

estimators were examined. Results of the simulated data show that the adaptive cluster sampling

with stopping rule is the more efficient sampling design for estimation of rare and clustered population than ordinary adaptive cluster sampling.