Activity Number:
|
350
|
Type:
|
Topic Contributed
|
Date/Time:
|
Wednesday, August 14, 2002 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Nonparametric Statistics*
|
Abstract - #301573 |
Title:
|
Nonparametric Inference About Divergence Based on Gini Index
|
Author(s):
|
Magdalena Niewiadomska-Bugaj*+
|
Affiliation(s):
|
Western Michigan University
|
Address:
|
3304 Everett Tower, Kalamazoo, Michigan, 49008, USA
|
Keywords:
|
two-sample problem ; divergence ; Gini index ; nonparametric estimation
|
Abstract:
|
Gini Index has been originally proposed as an inequality measure of income distribution. It has been shown that it can also be used as a measure of divergence of two probability distributions. Simple nonparametric estimator of Gini index is proposed as a statistic for inference related to two-sample problem. A simulation study compares proposed Gini test to other known two-sample procedures. Also, confidence intervals for Gini Index can be used to compare divergence in various pairs of distributions. Several examples will be provided.
|