|
Activity Number:
|
16
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 11, 2002 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods*
|
| Abstract - #301968 |
|
Title:
|
Replication Variance Estimation for Nearest Neighbor Imputation With Application to the 2000 Census Long Form Data
|
|
Author(s):
|
Jae-Kwang Kim*+
|
|
Affiliation(s):
|
Westat, Inc.
|
|
Address:
|
1650 Research Blvd., Rockville, MD, 20850, USA
|
|
Keywords:
|
Cell mean model ; jackknife ; nonresponse ; sequential hot deck
|
|
Abstract:
|
Item nonresponse frequently occurs in surveys of human respondents. To compensate for item nonresponse, various hot deck imputation procedures have been used to replace missing values in the sample. Kim and Fuller (2001) proposed a class of unbiased variance estimators for the total variance of the imputed estimator under the cell mean model. Using a simple example, we demonstrate that the proposed method can be extended to multivariate data with different missing patterns. The proposed variance estimator is general in the sense that it does not depend on the specific hot deck imputation method used. Thus, it can be naturally applied to the nearest neighbor imputation. An application with nearest neighbor imputation for long form census data is illustrated.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2002 program |
