|
Activity Number:
|
390
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #305915 |
|
Title:
|
Regression Diagnostics for Survey Data
|
|
Author(s):
|
Jianzhu Li*+ and Richard Valliant
|
|
Companies:
|
University of Maryland and University of Michigan
|
|
Address:
|
JPSM, College Park, MD, 20742,
|
|
Keywords:
|
regression ; diagnostics ; complex sampling
|
|
Abstract:
|
Regression diagnostics is an important area in regression theory which intends to explore the characteristics of a fitted regression model for a given data set. Techniques for regression diagnostics have been developed theoretically and methodologically for conventional OLS linear regression models. However, in survey sampling, diagnostics have not been extensively studied. The application of conventional techniques to survey data becomes less straightforward because of features of complex sampling designs like stratification, clustering, and weights. The goal of this research is to adapt and extend some of the standard regression diagnostics to the survey setting.
|
