|
Activity Number:
|
537
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #306394 |
|
Title:
|
Using Administrative Records with Model-Assisted Estimation for the American Community Survey
|
|
Author(s):
|
Robert Fay*+
|
|
Companies:
|
U.S. Census Bureau
|
|
Address:
|
Office of the Director, Stop 9001, Washington, DC, 20233-9001,
|
|
Keywords:
|
ACS ; survey estimation ; Greg ; calibration estimators ; generalized regression estimation
|
|
Abstract:
|
Full implementation of the American Community Survey (ACS) began in 2005. Among other purposes, the ACS will replace the decennial census long-form data, enabling a short-form census in 2010. A test implementation of the ACS in 36 test counties during 1999-2001 suggested that the initial ACS estimation procedure, while adequate at the county level, yields higher variances at the small-area levels of tract and block group relative to the decennial long form. Previously reported research argued the likely success of an approach combining administrative record data with a generalized regression estimator. The case for likely success was based on the R-square of the underlying regression. This paper reports detailed results on a full implementation in 34 test counties, showing the degree to which the predictions based on regression diagnostics have been confirmed.
|
- 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 2006 program |
