Scientific misconduct and participation rates in population-based epidemiological research: the NOWAC study

Abstract

To investigate the possible effect of a scientific misconduct case and the subsequent immense publicity on the participation rates in a population based epidemiological study. We investigated willingness to participate, by donating a blood sample, among random samples of 500 women included in the Norwegian women and cancer study—NOWAC. The women were asked to participate during a period before, under and after the discovery of misconduct January 2006. This population based sampling was independent of the case. Letters of invitation and a questionnaire were regularly mailed from 2003 till 2006, and here we use the period from November till April of two subsequent years; 2004/2005 and 2005/2006. Population based in Norway 30,216 Norwegian women born 1943–1957 invited to participate. Response rates (%) for returning a blood sample and for actively responding with a “No” to the invitation. The active response of a “No” increased immediately after the publicity of the scientific misconduct, but lasted only for 2 months. The response rate for blood samples was only marginally affected in the same period, and increased over previous level 3 months after the misconduct was publicly known. The widespread publicity related to the scientific misconduct did not affect the response rate in this population based study. This might imply that the public trust increased as a consequence of the public discussions that followed.