|
Activity Number:
|
479
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #304720 |
|
Title:
|
Multiple Imputation for Missing Items in Multi-Section Questionnaires
|
|
Author(s):
|
Rong Liu*+ and Joseph L. Schafer
|
|
Companies:
|
Penn State University and Penn State University
|
|
Address:
|
The Department of Statistics and the Methodology Center, State College, PA, 16801,
|
|
Keywords:
|
factor analysis ; item nonresponse
|
|
Abstract:
|
It is common in survey research to have multiple sections of a questionnaire, each of which contains multiple items that measure related aspects of the participants. Even with a modest number of sections, such data sets can have a large number of variables relative to the number of cases in the study, and even low missingness rates among the variables can result in a high proportion of incomplete cases. Here we present a new method for multiply imputing missing items in such survey data. Instead of attempting to model the covariance between each pair of items, we model the relationships among a small number of factors extracted from each section. The method is illustrated on a set of items from a large national survey on adolescent health.
|
