This study examined the consent to linking survey data to LinkedIn accounts among participants in the UK Understanding Society Innovation Panel. An experiment that crossed the placement of consent requests in a questionnaire and the inclusion of a statement about linking importance showed the highest consent rate when making the request early in the questionnaire with an addition of the importance statement. LinkedIn users were not a random subset of the respondents. LinkedIn users were more likely than nonusers to have higher education, to be employed and to use social media actively. Among the users, consenters to the linkage request were more likely to have higher education, to be employed and to use social media actively, compared to nonconsenters. This demonstrates potential compound selectivity issues with education, employment and social media use in survey-LikedIn linked data.
