Linking Survey and Facebook Data: Mechanisms of Consent and Linkage

Using a German online panel, we investigate respondents’ propensity to consent to the linkage of publicly available Facebook data to their survey data and to enable linkage post-consent. We analyze the effects of three experimental variations on consent and actual linkage: (1) the consent question wording, (2) the position of the Facebook consent question in a series of consent questions, and (3) the incentive offered. To describe potential selection bias, we consider respondents’ attitudes and sociodemographic characteristics. We found effects of consent question placement on consent and linkage, and effects of higher incentives on linkage, but we did find no effects of question wording. Our analysis showed that consent and linkage is dependent on several respondent characteristics.