SummaryThe willingness to participate in population-based research has been declining for decades. It is known that nonresponse in health studies may introduce biases in subsequent analyses, because it is often differential with respect to relevant socioeconomic characteristics, lifestyles, and the health status of potential participants. Typically, this phenomenon is countered by intensifying the recruitment, hence, by also recruiting persons who are more reluctant to participate (henceforth, late responders). However, the evidence on whether extended recruitment efforts are actually successful in diversifying study samples is mixed and evidence on their long-term effects on cohort attrition remains scarce. The large body of study data and paradata from recruitment and examinations in NAKO offers a unique opportunity to investigate differences between early and late responders. Since NAKO, as a multicenter study, offered to the study centers some degrees of freedom during recruitment, the effects of different recruitment intensities on the sample composition can be compared within as well as across study centers. The results of this project may help to investigate response and attrition in the follow-up examinations and, in the future, may inform strategies to refresh the NAKO sample. Furthermore, findings of this project may help in the design of the recruitment for future population-based studies.

Keywords attrition differential-nonresponse early-respon hard-to-reach-populations late-responder nonresponse recruitment representativity

InstitutionsLeibniz Institute for Prevention Research and Epidemiology - BIPS, Universitätsmedizin Greifswald, GESIS, Universitätsklinikum Freiburg