|
Activity Number:
|
247
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 8, 2001 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods*
|
| Abstract - #301264 |
|
Title:
|
A Two-Phase Sampling Approach for Selecting Samples from Large Administrative Files
|
|
Author(s):
|
Skip Camp*+ and K. P. Srinath and Susan Arday and Elizabeth Axelrod
|
|
Affiliation(s):
|
Abt Associates Inc. and Abt Associates Inc. and Health Care Financing Administration and Abt Associates Inc
|
|
Address:
|
1110 Vermont Ave., N.W. , Washington , District of Columbia, 20005-3522,
|
|
Keywords:
|
Two-Phase sampling ; Sample overlap ; Terminal Digits
|
|
Abstract:
|
In this paper, we describe a method of selecting a sample of Medicare beneficiaries from each of the 50 States and the District of Columbia from the Medicare Enrollment Database (EDB), a beneficiary level file maintained by Health Care Financing Administration (HCFA). Since this file is very large, a two-phase sampling approach is adopted. The first-phase selection of the sample of beneficiaries is achieved through the selection of terminal digits of Health Insurance Claim Numbers (HICN). Issues relating to the reduction of overlap in sample selected for other surveys from the same files are discussed. A stratified sample is selected at the second-phase based on the information in the EDB and the first phase sample. Alternative allocations to strata in the second-phase selection are considered based on the information from the sample and the entire EDB. An example illustrating the sampling procedure is presented.
|
