Testing the Convenience of a Variate for Stratification in Estimating the Gini-Simpson Diversity

摘要: In the literature on quantification of diversity in a population, one indices that has been proven to be useful dealing with uncensused populations is Gini-Simpson one. On other hand, whose respect certain aspect required quantified, often arise naturally stratified and large samples from them are available. Targets this paper are: approximate optimum allocation estimating sampling so asymptotic precision maximized; introduce relative gain (with allocation) over random as measure adequacy variate for stratification approximating diversity; construct procedure test convenience given samples.