|
Activity Number:
|
16
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 3, 2003 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Health Policy Statistics
|
| Abstract - #301308 |
|
Title:
|
A Statistical Match Between the National Health Interview Survey and the Current Population Survey
|
|
Author(s):
|
Deborah D. Ingram*+ and Chris Moriarity
|
|
Companies:
|
Center for Disease Control and Prevention and U.S. General Accounting Office
|
|
Address:
|
4411 Elm St., Chevy Chase, MD, 20815-6054,
|
|
Keywords:
|
statistical matching ; exact matching ; record linkage ; health policy analysis
|
|
Abstract:
|
This paper describes the statistical match of the 1995 National Health Interview Survey (NHIS) and the March 1996 Current Population Survey (CPS) and presents the results of an evaluation of the quality of the match. Statistical matching is a method used to combine two files when it is very unlikely that an individual is included on both files.This is in contrast to exact linkage where one attempts to link a record for a specific individual on one file with that for the same individual on the other file. The purpose of the NHIS-CPS match was to obtain a data set with measures of health status, health care utilization and family resources on the same file. Such a file is of interest to policy analysts who wish to examine the effect of changes in health policy, particularly with regard to effects on health care utilization. The NHIS-CPS statistical match was carried out using fully constrained predictive mean matching (PMM) with partitioning. Two matches were performed, one with the NHIS as Host and total annual family income from the CPS as the PMM variable, and one with the CPS as Host and number of doctor visits from the NHIS as the PMM variable.
|
