U.S. flag

An official website of the United States government

Department of Labor Logo United States Department of Labor
Dot gov

The .gov means it's official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Bureau of Labor Statistics > Office of Survey Methods Research > Publications > Browse Research Papers

Modeling the optimal contact mode in the Current Population Survey

John S. Dixon, Brian J. Meekins, and Polly A. Phipps

Abstract

The Current Population Survey (CPS) is designed to measure labor force characteristics of the United States, as well as other population characteristics. Households are interviewed eight times; four consecutive months, followed by an eight month break, then four more consecutive months. Interviewers contact households in person for the first interview and subsequent interviews are often done by telephone. Because of the eight-month break, conducting the fifth interview in person is encouraged; however, 40 percent are conducted by telephone. This study uses paradata from the Contact History Instrument (CHI), and demographic and household characteristics to determine which households are most likely to be successful using telephone data collection in the fifth interview. A successful interview is one that collects sufficient labor force information. A logistic model was used to identify the respondent and household characteristics which predict the best mode for conducting the interview in the fifth interview. The model included interviewer effort, differences in regions, age, race, ethnicity, gender, labor force status of the previous respondent (from the 4th interview), and respondent concerns (from the previous 4 interviews) to explore the impact of mode on nonresponse. The best mode for a successful household interview will be modeled using a competing risk model where interviewer effort and likelihood of nonresponse are minimized. The results can be used by managers and interviewers to decide which mode should be used.