Address-based sampling frames generated from the U.S. Postal Service Computerized Delivery Sequence file (CDS) are considered by many to be the gold standard source for a national household sampling frame. The literature to date suggests that the CDS provides nearly complete coverage of the national household population. Prior research has shown that the under-coverage is not random and can vary by geographic area, survey mode, and address type, all of which can introduce significant bias to survey outcomes.
Missing from the literature are methods of evaluating how much of the CDS our vendors qualify for. This document outlines preliminary methodology for determining how much of the USPS Computerised Delivery Sequence (CDS) file a vendor qualifies.
These data will give researchers a better understanding of what they can expect from their own ABS vendors while at the same time help vendors and their client's demonstrate the coverage of their ABS frame.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.