Practical guidelines for developing a smartphone-based survey instrument

The increasing relevance of mobile surveys makes it important to gather empirical evidence on designs of such surveys. This research note presents the results of a test study conducted to identify the best set-up for a smartphone-based survey. We base our analysis on a random sample of Danish citizens (n=196) who participated in mobile surveys during three subsequent days of investigation. First we investigated whether survey invitations sent via an app-based push message or via text message result in different response rates. Second, we tested for differences between the two platforms regarding response time-lag. Third, we tested whether send-out time (evening vs. morning) affects the number of respondents taking part in the survey. Our findings show that choice of platform as well as send-out time influence response behavior. Based on the good performance of the survey across platforms, we suggest that in certain contexts mobile-only surveys are a viable option compared to more traditional surveys.