|
Activity Number:
|
465
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #306113 |
|
Title:
|
On Generalized Variance Functions
|
|
Author(s):
|
Donsig Jang*+ and Amang Sukasih and Xiaojing Lin
|
|
Companies:
|
Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc.
|
|
Address:
|
600 Maryland Ave., SW, Suite 550, Washington, DC, 20024,
|
|
Keywords:
|
design effect ; variance approximation ; complex survey ; SESTAT
|
|
Abstract:
|
Generalized variance function (GVF) techniques would provide a few parameter estimates for a certain domain so that analysts can approximate variance estimates for estimates of their own interest. In this paper, we review GVF formulation and the underlying assumptions for GVF models. A typical GVF model assumes: (1) design effects are homogeneous for each GVF domain; and (2) proportions of people having attributes considered for GVF fitting are unrelated to the total number of people in the domain. GVF parameters are often estimated via least squares fitting methods. As an alternative, design effects, sample size, total domain size, and the variance of the total domain size estimate can be used to calculate GVF parameter estimates directly. Using the 2003 SESTAT data, we will compare the regression based GVFs with the design effect based GVFs.
|
- 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 |
