Sociology, Department of


Date of this Version


Document Type



Presented at “Interviewers and Their Effects from a Total Survey Error Perspective Workshop,” University of Nebraska-Lincoln, February 26-28, 2019.


Copyright 2019 by the authors.


Substantial research and practical experience shows that a telephone interviewer is most successful at gaining cooperation and avoiding refusals when they are free to tailor their introductory pitch to the potential respondent or household informant they reach. However, survey designers are often uncomfortable allowing interviewers to work “off-script,” and instruct interviewers to read introductory text verbatim. Further, some interviewers report being more comfortable with a script than without one. To bridge this gap between research and practice we asked, “Can we create a scripted introduction that engages the potential respondent, gets a foot-in-the-door, and facilitates interviewer tailoring?” This paper reports on a randomized experimental test of two such scripts, each implemented within the Washington Behavioral Risk Factor Surveillance System (BRFSS), a random digit dial (RDD) phone survey. In both phases of the experiment, sampled phone numbers were randomized to the standard BRFSS introduction or the new script. Phase 1 (August, 2018) implemented a “conversational” introduction that added or revised three features of the standard BRFSS introduction: First, the introduction included three “hook questions” (e.g., “Have you heard of the survey?”). One hook question was randomly displayed each time a phone number was called. Second, the script displayed on the first three CATI screens was modified to sound more conversational and less abrupt. Third, pause points were created to make sure the interviewer slows down and listens to the potential respondent. Each of these features is hypothesized to increase tailoring, and thus cooperation, by encouraging interaction between the interviewer and potential respondent. Phase 2 (September, 2018) replaced the conversational introduction with a “progressive scheduling” script that instructed interviewers to ask for a good time to call back to complete the interview rather than asking for complete cooperation on the call. This approach encourages a dynamic that shows respect for the respondent’s time. It also changes a large, unexpected request to a small one that the respondent can plan. While call-backs are sometimes considered undesirable outcomes, they can be a good “toe-in-the-door” technique that leads to full cooperation later. Our primary outcomes are cooperation, scheduled callbacks, and refusals. Additionally, a more conversational introduction might influence answers to questions within the interview, such as if increased rapport depresses reports of sensitive behaviors. Thus, we will also assess the effect of the modified scripts on responses to assess the nonresponse / measurement error trade-offs of this approach.