期刊名称：Journal of Modern Applied Statistical Methods
出版年度：2020
卷号：19
期号：1
语种：English
出版社：Wayne State University
摘要：This paper uses inequality-measurement techniques to assess goodness of fit in income distribution models. It exposes the shortcomings of the use of conventional goodness of fit criteria in face of the big income data and proposes a new set of metrics, based on income inequality curves. In this note, we mentioned that the distance between theoretical and empirical inequality curves can be considered as a goodness of fit criterion. We demonstrate certain advantages of this measure over the other general goodness of fit criteria. Unlike other goodness of fit measures, this criterion is bounded. It is 0 in minimum difference and 1 in maximum distance. Furthermore, there is a consistency between this new goodness of fit measure and the other conventional criteria. A simulation study based on fitted distribution to real income data is performed in order to investigate some statistical properties of the new goodness of fit measure. An empirical study and comparisons are also provided.
