|
Activity Number:
|
71
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #308976 |
|
Title:
|
Handling Imperfect Overlap Identification in a Dual-Frame Survey
|
|
Author(s):
|
Jay Clark*+ and Marianne Winglee and Benmei Liu
|
|
Companies:
|
Westat and Westat and Westat
|
|
Address:
|
1650 Research Blvd, Rockville, MD, 20850,
|
|
Keywords:
|
Dual-frame design ; Record linkage ; Threshold selection ; Misclassification error
|
|
Abstract:
|
The analysis of dual frame surveys requires identification of which sampled units are included on both frames. However, this identification is imperfect when only limited and nonidentical matching data is available (except in the case of units sampled from both frames). This situation is encountered in the National Incidence Study of Child Abuse and Neglect. That study employs a dual-frame design that combines a list frame of all maltreated children investigated by Child Protective Services agencies and another sample frame compiled from maltreated children reported by sources such as the police and school staff. We handled overlap determination by using a record linkage procedure that aims to minimize the misclassification rate for matches and nonmatches. This paper describes our approach and also explores the impact of misclassification in a simulation study.
|
