Asking for Panel Consent in Web Surveys: Choice, Opt-in, or Opt-out?

Abstract

Some surveys ask respondents for consent to be recontacted for follow-up surveys after the initial  part of the survey has been completed. Based on an experiment, we compare three options for asking for this panel consent: choice (yes/no), opt-in, and opt-out. We analyse panel consent rates and compare consenters with non-consenters against a comprehensive set of socio-demographic characteristics, political attitudes, and survey-related variables in a probability-based web survey. In a second step, we analyse consenters’ actual participation in the first follow-up wave. The opt-out option yields higher panel consent rates than the other two options. Based on socio-demographic variables, panel consenters and non-consenters are most similar to each other in the choice design, and most different in the opt-out design. Based on typically biased variables, such as political interest or how the survey was perceived, the opt-out design performs better than the opt-in design in terms of consent, followed by the choice design. When it comes to actually participating in the first follow-up wave, the three panel consent options work in a similar way to giving consent. Overall, these findings speak in favour of the opt-out design, followed by the opt-in design