Classification of Address Register Coverage Rates--A Field Study

• Activity Number: 309
• Type: Contributed
• Date/Time: Wednesday, August 11, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Gavin Thompson*+ and Claude Turmelle
• Companies: Statistics Canada and Statistics Canada
• Keywords: Canadian Labour Force Survey ; address register ; coverage rates ; sample frame

Abstract

Every 10 years the Canadian Labour Force Survey (CLFS) undergoes a review of its sample design. This time the use of the Canadian Address Register (CAR) was examined to replace the costly listing process. The CLFS has a two-stage design. For each Primary Sampling Unit (PSU) selected a complete listing of all dwellings must be compiled in order to select the final sample. This list is normally compiled by an interviewer touring the entire PSU and noting each dwelling. The use of the CAR would enable us to eliminate this step in some PSUs (type 1), and in others potentially improve the quality of the lists by being used as an aide (type 2). In some PSUs the coverage rates of the CAR is too poor to be used (type 3). A field study in five cities, totaling 55,000 households, was completed in December 2003. This paper will describe the design of the field test, review the overcoverage and undercoverage rates of the CAR and present the results obtained with both methods of listing. Also, we will discuss the methods used for determining which of the three types the 8,000 CLFS selected PSUs fall into.