Adaptive sampling designs and their applications

Abstract

Consider a study of stigmatized populations such as drug users , people enganged in illegal activities or people suffering from AIDS. Such a study encounters difficulties because the members of the population are rare, hidden ,or hard to reach. Conventional sampling methods detect relatively few members of the populations and therefore estimates of the population parameters have high variances. Similar problems are encountered in enviromental surveys since most of the populations of animals and plants tend to cluster over space. Adaptive sampling designs are alternative sampling methods which have been found to work well in such populations . This study looks at adaptive cluster sampling methods and their application to Enviromental surveys and Behavioral studies. The Hansen - Hurwitz estimator and the Horvitz - Thompson estimator are considered and their relative efficiency in estimating the population parameters is illustrated. The study also looks at the advantages of adaptive sampling designs over the conventional designs Other issues mentioned albeit briefly are; Adaptive sampling in a multivariate setting, Improved Rao -Blackwell estimator and efficiency of adaptive sampling designs