Reaching Further: Insights from a Factorial Survey with a Hard-to-Reach-Population

Wednesday, 9 July 2025: 00:00
Location: ASJE028 (Annex of the Faculty of Legal, Economic, and Social Sciences)
Oral Presentation
Anne-Kathrin CARWEHL, Research Centre, Federal Office for Migration and Refugees, Germany
Randy STACHE, BAMF Research Centre, Germany
Laura PEITZ, BAMF Research Centre, Germany
Respondent-Driven-Sampling (RDS) is a validated approach for targeting hard-to-reach populations, such as drug dealers, elderly immobile individuals, sex workers, sexual minorities, and migrants. Nevertheless, the accessibility of these populations does not guarantee the success of a survey, as the particular circumstances of the group may make it challenging to obtain valid and reliable responses to sensitive questions (reservations, traumatic experiences, etc.). Survey experiments, such as factorial surveys, are frequently employed to address complex issues, as they allow for a realistic representation of even more complex scenarios. However, it’s still unclear if they are particularly effective for gathering high-quality data on sensitive topics or hard-to-survey groups.

The BAMF Research Centre conducted a fully app-based RDS with rejected asylum applicants from anglophone West Africa (Nigeria, Ghana, Gambia, Sierra Leone) between June and December 2023. The app encompassed all the elements necessary for conducting an RDS, as well as a factorial survey, while maintaining participant anonymity. Our factorial survey comprised four hypothetical scenarios (vignettes), each describing the living circumstances of a fictional person with tolerated status. Respondents had to recommend whether the fictional persons should stay in Germany, return to their home country, or migrate elsewhere.

This contribution offers insights into our methodological approach, particularly in terms of whether factorial surveys are an appropriate method for addressing sensitive questions to a hard-to-reach population. To this end, we compare vignettes with traditional survey questions and use paradata (e.g. dropout rate and response time) to gain valuable insights with a view to enhancing future factorial surveys addressing hard-to-survey populations.