Speaker
Ingram Olkin, Professor Emeritus, Department of Statistics, Stanford University.
Abstract
Heterogeneity, diversity and inequality have much in common. None are based on a probabilistic model; none are defined as a parameter. All are ill-defined, yet all exist. Diversity generally relates to biodiversity, and inequality generally relates to economics. But the question of when one group is more diverse or more unequal than another are similar. Many measures have been proposed to measure diversity. In meta-analysis these measures focus on the Q-statistic, or its transform. Although single measures describe certain aspects of diversity, they are inadequate to capture the full range of the distribution. In this talk I will review many of the measures used, and then discuss a new method based on the Lorenz curve, which captures more of the characteristics of the distribution than single measures.