Key points are not available for this paper at this time.
Abstract In-the-moment surveys administered when respondents participate in a predetermined event of interest—such as visiting a particular location or completing a specific activity—allow researchers to measure behaviors and experiences as the event occurs, situating data collection within the relevant context and reducing reliance on retrospective self-reports. However, data quality can be compromised if individuals are unwilling to participate or provide low-quality responses. In particular, the timing of survey prompts and the structure of incentives may affect data quality. We evaluated these issues in an experimental study of 610 individuals who used a custom smartphone app to answer a survey about each of their public transit trips over 2 weeks. Participants were randomly assigned to receive either daily or per-survey incentives, with surveys administered either during trips (in situ) or immediately afterward (near-time). Both daily incentives and in situ prompts significantly increased participation. In situ prompts also reduced “speeding” and increased the variation in survey responses across trips, suggesting improved accuracy in capturing event-specific details. These findings demonstrate that prompt timing—specifically, in situ prompts as opposed to near-time prompts—plays a key role in enhancing data quality for in-the-moment surveys.
Antoun et al. (Tue,) studied this question.