The Use of Diversity Analysis to Assess the Relative Influence of Factors Affecting the Income Distribution
使用基于二次熵的多样性度量，分析年龄、性别和教育等因素对收入分配的影响，发现性别与教育对总体收入不平等的解释力相当，且该方法可将总不平等分解为组间和组内成分。
This article uses diversity measures based on quadratic entropy to analyze the relative effects of factors such as age, sex, and education on the income distribution. In particular, sex appears to be as important as education in explaining the overall income inequality in the data collected by the U.S. Bureau of the Census. An advantage of the diversity measures is that they possess a decomposition of the total inequality into between-group and within-group components. Moreover, analogs of partial association enable one to study the effects of several factors. By using several measures, our inferences are shown to be robust.