Two-stage restricted adaptive cluster sampling

摘要

Adaptive cluster sampling can be a useful technique for parameter estimation when a population is highly clumped with clumps widely separated. In this design, however, the size of the final sample cannot be predicted prior to sampling, thus leading to design problems. In this paper a new version of adaptive cluster sampling which allows the sampler to know prior to sampling the exact upper limit of the final sample size and consequently the highest total sampling effort, is suggested. It is drawn from a combination of the restricted adaptive cluster sampling of Brown (1994) and a two-stage adaptive cluster sampling. For this new sampling design an unbiased estimator of the total and its sample variance are also suggested. The results of a simulation study performed to gather insight as to the performance of the proposed methodology are promising.