|
Activity Number:
|
190
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Government Statistics
|
| Abstract - #306871 |
|
Title:
|
Inferences on Two-Stage, Multiply-Imputed Data
|
|
Author(s):
|
Satkartar Kinney*+ and Jerome Reiter
|
|
Companies:
|
Duke University and Duke University
|
|
Address:
|
211 Old Chemistry Bldg., Box 90251, Durham, NC, 27708,
|
|
Keywords:
|
synthetic data ; missing data ; data confidentiality ; multiple imputation
|
|
Abstract:
|
Partially synthetic datasets, where some values are replaced with multiple imputations, usually are generated for disclosure limitation. It is often the case that missing values also must be multiply imputed. We will discuss the generation and analysis of two-stage, multiply-imputed datasets, particularly significance testing for estimands from such data.
|
