Face-to-Face Screening Interviews: Tradeoffs Between Coverage and Nonresponse Errors

• Activity Number: 232
• Type: Topic Contributed
• Date/Time: Monday, August 5, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Roger Tourangeau*+ and Frauke Kreuter and Stephanie Eckman
• Companies: Westat and University of Maryland and IAB
• Keywords: Coverage error ; Screeners ; Nonresponse error

Abstract

Prior research has shown that the same groups that are sought in screening interviews are often seemingly underrepresented in the screening data. This underrepresentation may be a product of deliberate misreporting by the screener respondents (or motivated underreporting) or may reflect interviewer motivation to deal with potentially difficult respondents by encouraging them to screen out. This paper reviews the results of an earlier study in which screening interviews were administered by telephone and of a new study in which screening interviews were done face-to-face. In both studies, different versions of the screening questionnaire were administered. The different versions of the screener varied in whether they used a direct screening question ("Is anyone who lives here between 35 and 55 years old?") or collected several characteristics for each household member (including their ages). The more detailed screening questions reduced coverage errors but also increased the nonresponse rate in the telephone study. We examine data from the new face-to-face experiment to see whether this same tradeoff is apparent when the screeners are done in person.