A procedure to test the suitability of a factor for stratification in estimating diversity

摘要

When we try to estimate the diversity of species or categories for a classification process in large uncensused populations, we can take advantage of the fact that they can usually be stratified by different factors to obtain a gain in the precision of the estimate. Should this be the case, we can first compare different factors to find the one which exerts the strongest influence on the distribution of species (more precisely, the one which presents the highest “mutual information” between the classification process and each factor). Then, the suitability of the “best factor” can be discussed by comparing the corresponding mutual information with its maximum value: the diversity of the classification process. In this paper, we develop a mathematical procedure, based on the quadratic indices of diversity and mutual information, to test such suitability when large samples are available. The application of the procedure is illustrated by means of a practical example.