ESRA 2027 sessions by theme
Back to Overview of Sessions
Maintaining Engagement and Preventing Attrition in Longitudinal Studies: Evidence, Innovation and Practice |
| Coordinator 1 | Dr Pablo Cabrera Álvarez (University of Essex) |
| Coordinator 2 | Professor Nicole Watson (University of Melbourne) |
| Coordinator 3 | Dr Matt Brown (Centre for Longitudinal Studies, UCL) |
| Coordinator 4 | Professor Lisa Calderwood (Centre for Longitudinal Studies, UCL) |
Session Details
Maintaining engagement across waves is central to the quality of longitudinal data. This matters because panel attrition can bias survey estimates when those who drop out differ from those who remain on the variables being estimated. Panel attrition also accumulates over time, reducing the base available for long-term longitudinal analyses. Several strategies have been tested to mitigate panel attrition, such as survey incentives, using multiple modes to contact sample members, or tailored communications. This session is a forum to share and discuss the latest evidence on preventing panel attrition. We invite contributions that approach attrition from three perspectives:
•Analyses of panel attrition and its causes with a view to intervention. Analysis of attrition patterns, relative contribution of non-contact, refusal and inability to participate, and the role of respondent characteristics, life-course transitions, interviewer and fieldwork effects, and prior-wave survey experience. The consequences of attrition for representativeness and for substantive estimates.
•Evidence on interventions to mitigate attrition and improve representativeness. Incentives, contact strategies, approaches to tracking and tracing movers, between-wave participant engagement activities, responsive designs, and mixed-mode designs. We are particularly interested in novel approaches to maximise wave response rates and improve sample representativeness overall and among specific hard-to-reach subgroups, including re-engaging prior non-respondents. We welcome evidence from experimental designs, quasi-experiments and observational evaluations.
•Targeted and adaptive designs in longitudinal studies. Designs that use information from earlier waves, paradata and administrative records to prioritise cases and optimise the allocation of effort and modes. We also welcome contributions on targeted procedures for specific subgroups such as tailored communications or differential incentives, aimed at improving response rates and representativeness within cost constraints.
Papers on household panels, cohort studies, and other longitudinal studies are welcome, as are replications and comparative work across studies.
