Online Program Home
  My Program

Abstract Details

Activity Number: 80 - Inference Methods for High-Dimensional and Complex Data
Type: Contributed
Date/Time: Sunday, July 30, 2017 : 4:00 PM to 5:50 PM
Sponsor: Section on Statistics in Imaging
Abstract #324847
Title: Bounded-Width Confidence Interval for Gini Index Under Complex Survey
Author(s): Francis Bilson Darku* and Bhargab Chattopadhyay
Companies: University of Texas at Dallas and Indian Institute of Information Technology - Vadodara
Keywords: Gini Index ; Complex Survey ; Sequential Analysis ; Research Design ; bounded-width confidence interval
Abstract:

Gini index is an income inequality measure in determining the inequality in the distribution of income or assets among individuals or groups within a society or region. For any geographical location, the computation of Gini index is helpful in evaluating the performance of different economic policies. However, Gini index computed based on census data is available once in every 10 years or more, since many countries cannot afford to collect data from all households annually. In order to estimate Gini index for periods in-between two census years for a region with large number of households, complex sampling designs involving stratification and clustering are often used to ensure adequate representation of groups of interest. Fixed-sample size methodologies exist for constructing confidence intervals for Gini index under complex household survey scenarios but it cannot be used to find bounded-width confidence interval for Gini index. This article therefore develops a two-stage sequential procedure for estimating and constructing a bounded-width confidence interval for Gini index under a complex survey design using the smallest possible cluster sizes.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association