Research question: How to improve data quality and completion rates in online surveys? Design: We conducted an online experiment in which 1,750 students were randomly assigned to one out of 4 treatments and a control group. Group 1 received a postal prenotification of the survey along with a prepaid voucher. Group 2 received a postal prenotification and a postpaid voucher. Group 3 received a postal prenotification and a prepaid 5 EUR bank note. Group 4 solely received a postal prenotification. Group 5 was the control group and was invited via e-mail. Dependent measures were completion rate, item-nonresponse, straightlining and willingness to self-report sensitive information. Results: Logit models show Group 3 (prenotification & bank note) to significantly outperform all other groups with respect to completion. Treatment had no effect on straightlining. The overall amount of straightlining, however, was low. Regression models for count data revealed item-nonresponse to be lowest in Group 3 (prenotification & bank note). Most groups differed from one another in item-nonresponse, whereby all treatment groups were superior to the control group. Finally, treatment had no significant effect on respondents’ willingness to disclose sensitive information.

Implications of the study: This is one of the first experiments that tests the effect of a prepaid cash incentive on participation in an online study. We are able to show that including a bank note roughly doubles completion and also enhances the overall data quality. We therefore advocate the use of prepaid cash incentives. As there were no differences in the willingness to self-report sensitive information, the different treatments seem to be neutral to respondents’ perceptions of anonymity.