The interplay of incentives and mode-choice design in self-administered mixed-mode surveys

Self-administered mixed-mode surveys are increasingly used as an alternative to face-to-face surveys for collecting data from the general population. However, little is known about how decisions regarding the incentive scheme and the mode-choice design jointly affect key outcomes such as response rates, net sample composition, and survey costs. To study this, we drew a probability sample of the residential population of the city of Mannheim, Germany (N = 2,980) and randomly assigned target persons to one of four incentive schemes (€0, €1, or €2 prepaid incentive on first contact, and €2 delayed prepaid incentive) and one of two mode-choice designs (concurrent or sequential [web-push]). Our results indicate that small prepaid monetary incentives work better in concurrent than in sequential designs. Moreover, a €2 prepaid incentive in a concurrent design proved particularly successful for older target persons, probably reinforcing their sense of trust and reciprocity, while also fitting better with their survey-mode preferences. Finally, a €2 delayed prepaid incentive in a sequential design primarily motivated target persons aged under 50 years. This combination of incentive scheme and mode-choice design also proved to be most cost-effective in that age group. Based on our results, we recommend using sampling frame information on age to address different age groups with different combinations of incentive scheme and mode-choice design. This may help to maximize response rates, achieve a balanced net sample composition, and minimize survey costs.