Abstract #301114

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2002 Program page
JSM 2002 Abstract #301114
Activity Number: 133
Type: Invited
Date/Time: Monday, August 12, 2002 : 2:00 PM to 3:50 PM
Sponsor: Section on Survey Research Methods*
Abstract - #301114
Title: Issues in Design-Based Weighted Analysis of Survey Data
Author(s): Barry Graubard*+ and Edward Korn
Affiliation(s): National Cancer Institute and National Cancer Institute
Address: EPS 8024, 6120 Executive Blvd, MSC 7244, Bethesda, Maryland, 20892, USA
Keywords: sample weights ; mixed models ; random effects ; variance components
Abstract:

It is well established that weighted analysis of survey data provides nearly unbiased estimates of population parameters such as means or regression coefficients. However, offsetting this reduction in bias often increases variance due to inefficiency of the sample weighting. Several issues arise in deciding when and how to weight survey data in statistical analysis that depend upon the extent of inefficiency, evidence of bias reduction, and the type of model assumptions about the relation of the sample weights to the analytic variables. In the area of how to weight, we discuss potential problems that can come from using standard inflation weight estimators. We give several examples from regression, estimation of variance components, and categorical data analysis, using data from national surveys.

  • 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

JSM 2002

For information, contact meetings@amstat.org or phone (703) 684-1221.

If you have questions about the Continuing Education program, please contact the Education Department.

Revised March 2002