Responsive Survey Design, Demographic Data Collection and Models of Demographic Behavior

William G. Axinn, University of Michigan
Cynthia F. Link, University of Michigan

Falling response rates have led survey methodologists to pioneer innovative ways to use process data ("paradata") to address non-response by altering the survey design. By improving representation of reluctant respondents, responsive design also changes our understanding of substantive issues studied by demographers. Using the National Survey of Family Growth (NSFG) Cycle 6 we illustrate how responsive survey design can improve both demographic estimates and models of demographic behaviors based on survey data. By juxtaposing measures from regular and responsive data collection phases, we document characteristics of the general population that are systematically under-represented in surveys not taking special effort to interview reluctant respondents. Using multivariate models already established in the literature and based on NSFG data, we demonstrate how adding reluctant respondents through responsive survey design changes model estimates. Results demonstrate the wide potential of responsive survey design to improve the quality of science in demographic research based on survey data.