ESRA 2025 Preliminary Program
All time references are in CEST
Surveying Ukrainian Refugees in Europe: Implementation, Methods, Experiments, and Challenges 2 |
| Session Organisers |
Dr Jean Philippe Décieux (University of Bonn and Federal Institute of Population Research (BiB))
Ms Silvia Schwanhäuser (Institute for Employment Research)
|
| Time | Friday 18 July, 11:00 - 12:30 |
| Room |
Ruppert 002 |
Since the start of the war in the Ukraine, many Ukrainians became internally displaced people or sought refuge in the surrounding European countries. According to Eurostat, around 4.3 million war refugees from Ukraine had received protection under the Temporary-Mass-Protection-Directive in Europe at June 2024. Such a massive displacement and inflow of refugees poses a significant challenge for local and national politics, administration and society of the refuge giving regions. Hence, there is a need for appropriate empirical evidence, in order to take efficient actions, grant needed support, and helping effective social integration.
In response to this growing demand, a large number of survey projects have been initiated in Europe. While initially, the focus was on rapid and cost-effective ad-hoc surveys to meet urgent data needs, there has been a notable shift towards the development of more sustainable panel infrastructures and cross-national or comparative studies in recent time. All these projects surveying Ukrainian refugees as hard-to-reach population face special circumstances and conditions. We would like to bring these projects surveying Ukrainian refugees together for an exchange of their experiences and to discuss survey methodological and practical challenges. We particularly encourage submissions that offer a perspective on the following dimensions of survey research:
• Different sampling strategies and approaches
• Comparative perspectives
• Approaches to reach the target population
• Different survey designs and modes
• Questionnaire design and translation
• Fieldwork organization and monitoring
• Attrition, follow-up rules, and experiences in tracing respondents’ return or onward migration
• Innovative tracking techniques for longitudinal designs
• Experimental approaches
• Cross-national comparisons and approaches
Keywords: hard-to-reach, refugee survey, sampling, probability vs. non-probability sampling
Papers
(Non-)Probability-Based Sampling of Refugees: A Comparative Analysis of Ukrainian Refugee Samples in Germany
Dr Steffen Pötzschke (GESIS - Leibniz Institute for the Social Sciences) - Presenting Author
Dr Bernd Weiß (GESIS - Leibniz Institute for the Social Sciences)
Dr Andreas Ette (Federal Institute for Population Research)
Dr Manuel Siegert (Federal Office for Migration and Refugees)
Professor Sabine Zinn (German Institute for Economic Research)
Ms Silvia Schwanhäuser (Institute for Employment Research)
Recently arrived refugees are notoriously hard-to-sample and survey. Yet, political realities indicate that the need for research on and with refugees might rather increase than decrease in the foreseeable future. Hence, the question of how survey sampling design might impact sample composition is highly relevant.
In refugee research, non-probability sampling has long been used, for example, due to the absence of reliable sampling frames, budget constraints, the need for quick data availability, or a combination of all these reasons. Despite the inherent analytical limitations of non-probability-based samples such data can provide crucial insights. However, it is important to better understand how such samples might differ from probability-based ones.
Against this background, we focus on comparing the sample composition of a probability-based (n ≈ 11,000) and non-probability-based (n ≈ 9,000) survey of Ukrainian refugees in Germany. While the former stems from a project carried out by a consortium of German research institutes that used register-based sampling (Brücker et al., 2022), the latter comes from a survey whose respondents were recruited through social media advertisements. Both surveys were conducted within a time frame of 10 months. Since the second survey largely replicated the questionnaire of the first, our data sources show a high level of measurement equivalence. We investigate how the samples differ from each other and from the distribution of the target population’s key sociodemographic information in administrative data at the respective time of data collection. Furthermore, we will highlight the specific advantages and shortcomings of both sampling approaches.
Effects of Socio-Demographic Composition on Attrition between the First and Second Waves in a Refugee Panel
Mr Johannes Schuett (DIW Berlin e. V. / SOEP) - Presenting Author
Mrs Elena Sommer (DIW Berlin e. V. / SOEP)
Previous research has highlighted attrition as one of the key challenges for probability panels of migrants and refugees (e.g. Jacobsen & Siegert 2023; Lynn et al. 2018; Kalter 2006). Among other factors, attrition can be attributed to residential mobility, language barriers, and the specific socio-demographic composition of migrant and refugee samples. The IAB-BAMF-SOEP Refugee Survey, which has been conducted annually in Germany since 2016, regularly draws refresher samples to account for panel attrition and changes in refugee migration patterns. In 2023, a refresher sample M9 was added to the IAB-BAMF-SOEP Survey of Refugees, consisting of two subsamples – refugees from Ukraine and refugees from other countries. In the first wave, participation was similar in both subsamples among households that had completed a short self-administered contact form before being invited for a face-to-face interview. However, in the second wave in 2024, participation was much higher in the Ukrainian refugee subsample than in the non-Ukrainian refugee subsample, despite the use of the same questionnaires, contact methods and incentive strategy. We expect that differences in the socio-demographic composition of these two subsamples – such as differences in gender and age composition, legal status and labour force participation – account for the differences in attrition between the first and second waves. In our study, we analyse the composition effect on participation in the second wave by considering household, respondent, and interviewer characteristics as well as interview aspects (e.g. change of interviewer, language, and mode preferences). Our study makes an additional contribution to the research on attrition in migrant panels by offering a comparative perspective within the broader refugee population in Germany and going beyond a descriptive comparison of survey participation by analysing the factors relevant to panel attrition across different groups of refugees.
Violence and related human rights abuses against women fleeing the war in Ukraine
Ms Rebecca Pearce (Ipsos UK) - Presenting Author
Mr Sami Nevala (EU Agency for Fundamental Rights)
This paper will describe our approach to researching hard-to reach populations in a time of geo-political and humanitarian crises. The project, conducted by Ipsos on behalf of the EU Agency for Fundamental Rights, interviewed women aged 18-74, who left Ukraine after the 24th February 2022, and were living in Czechia, Germany or Poland. Comprising of two elements - a face-to-face survey and qualitative interviews - it was designed to provide insights into the prevalence of physical, sexual, and psychological violence against women, while in Ukraine, during transit, or in their EU country of residence.
The paper will focus on the key considerations involved in delivering the project:
Developing a random probability sampling approach for the survey, based on available and accessible sampling sources. The most suitable approach proved to be a location sampling method, which derived representative samples of the target population across all three countries. In each country, recruitment took place in at least 35 location centres, which taken together, were likely to be visited by most members of the target group.
How to conduct ethical research on sensitive subjects with women who have been displaced, whilst minimising the risk of harm. Considerations included the interviewer’s role, how self-completion was used for particularly sensitive survey questions, and the need to ensure privacy when conducting recruitment in public places.
How to recruit and train Ukrainian interviewers - some of whom may also have been displaced - whilst also ensuring their well-being. A consideration was to employ native language counsellors for the duration of fieldwork, so that respondents and interviewers could receive a counselling session if required.
The paper will close with a reflection on how the data will contribute to a better understanding of the experiences and needs of women who have fled the Ukraine war.
