|
Activity Number:
|
115
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 1, 2016 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Government Statistics Section
|
|
Abstract #320147
|
View Presentation
|
|
Title:
|
Using Auxiliary Marginal Information to Deal with Nonignorable Missing Data
|
|
Author(s):
|
Mauricio Sadinle* and Jerome Reiter
|
|
Companies:
|
Duke University/National Institute of Statistical Sciences and Duke University
|
|
Keywords:
|
Missing data ;
Log-linear model ;
Nonignorable nonresponse
|
|
Abstract:
|
We study a scenario where we have a sample with two sets of categorical variables, and we have access to auxiliary marginal information on the distribution of one of these sets. The incorporation of this marginal information allows us to identify the nonignorable response mechanism under certain assumptions. We present different ways of accounting for the possible uncertainty in the marginal information.
|
Authors who are presenting talks have a * after their name.
