Interactional Influences on Survey Participation

University Of Wisconsin-Madison, Madison WI

Investigators

Abstract

This research focuses on the critical challenge of recruiting participants for social research surveys, a problem highlighted by the recent National Research Council?s Panel on the Future of Social Science Data Collection. The refusal component of nonresponse has grown steadily in recent years, threatening the ability of surveys to represent the populations from which the samples are drawn. The proposed research examines the moment-by-moment unfolding reciprocal influence between interviewers and sample members. For example, when an interviewer engages in persuasive actions, is it because she has somehow perceived cues that this sample member is persuadable? When the interviewer engages in actions that appear to be successful in persuading the sample member, is it simply that a persuadable sample member has provided her the opportunity to do so? The research uses new substantive and methodological approaches, drawing on theories of social exchange and reciprocity; conversation analysis and the interaction order; and content analysis. The empirical investigations exploit a unique collection of pairs of acceptances and declinations in a case-control design that uses observations from the Wisconsin Longitudinal Study that are matched on the sample members' propensity to participate. The matching of observations provides for comparison between calls that end in acceptances and declinations of the request for the interview. The project fills a gap in knowledge about whether sample members' voices provide interviewers with information that they can use to rapidly form accurate assessments of the likelihood that a call will lead to an interview. An acoustical analysis of the sample member's initial speech provides variables such as pitch that can be used with the case-control design to estimate whether these properties of the sample member's voice predict acceptance of the request for participation. By selectively and strategically adding to the existing data transcripts of subsequent contacts and attempts to obtain cooperation, the research also expands previous descriptions of ways interviewers tailor and maintain interaction to describe whether empathic responsiveness and interactional responsiveness (as specific forms of tailoring) make a difference for acceptances of the request to participate. The intellectual merit and importance of the project are rooted in its contributions to improving the accuracy of social scientific research, which is key to the nation's data infrastructure. The quality of information that describes an entire population can be compromised if participation is restricted to highly selected segments within a sample. This research expands theories of survey participation and their application to the development of practical methods for increasing participation in survey interviews for social research. The broader impacts of the research include identifying specific techniques for improving recruitment to participation in social research, disseminating information about these techniques in refereed journals and presentations at conferences, and training of both graduate and undergraduate students as they participate in the research effort. The project is supported by the Methodology, Measurement, and Statistics Program and a consortium of federal statistical agencies as part of a joint activity to support research on survey and statistical methodology.