ESRA 2019 Programme at a Glance

Lessons Learned from Studies on Refugees 2

Session Organisers Dr Christoph Homuth (Leibniz Institute for Educational Trajectories (LIfBi))
Dr Gisela Will (Leibniz Institute for Educational Trajectories (LIfBi))
Dr Roman Auriga (Leibniz Institute for Educational Trajectories (LIfBi))
TimeWednesday 17th July, 16:30 - 17:30
Room D24

Due to the increasing refugees’ movements over the world and the increasing influx of refugees into Western countries (especially Europe), many studies on refugees have been conducted or started.

These studies range from narrow topics on specific groups and/or their situations at specific times or places, their motivations or reasons to flee their countries, to large-scale surveys that are either newly started or that try to integrate refugee populations into their existing samples.

Studies on refugees face special problems due to the specific (legal, economic, social, medical etc.) conditions refugees are living in, e.g. conducting interviews in crowded places like group accommodations, ethical challenges of the researcher as a non-intervening observer, misunderstandings of the aims of the researchers to the interviewed refugee. But for all that it is crucial to interview refugees adequately, if one aims to gain meaningful results.
Therefore, in this session, we want to discuss possible solutions or research strategies that were applied to deal with these special conditions and challenges of studies on refugees. We are especially interested in experiences made by applying these strategies, so that researchers can learn from each others' trails and errors during their work with/on refugees. This session concentrates on solutions that can be implemented in or transferred to quantitative-empirical studies.

We invite contributions from researchers who report how they dealt with specific problems and how these survey technics worked in practice. Possible thematic foci can be:

- Translation and multi-language-instruments
- Handling of left- vs. right-aligned languages (e.g. in the presentation to the interviewee)
- Questioning illiterate persons (e.g. audio-supported interviewing)
- Interviewer effects (e.g.: cultural challenges due to the gender of interviewer and interviewee; native speaking vs. non-native speaking interviewers)
- Conducting interviews in group accommodations
- High mobility (e.g. use of special communication channels like social media)
- Social desirability
- Construct validity (e.g. cultural differences, systematic response bias, established scales vs. simple language items)
- Consideration of traumatic experiences and posttraumatic stress disorder (PTSD)

Furthermore, we appreciate contributions, which analyze, how different survey techniques may influence research results.

Keywords: refugees, translation, illiteracy, construct validity, social desirability

Probabilistic Sampling for Hard to Find Populations with Frame Deficiencies: Can Syrian Refugees be Sampled Using Registration Data in Turkey?

Dr Tugba Adali (Assistant Professor) - Presenting Author
Dr Sinan Turkyilmaz (Professor)
Ms Melike Sarac (Research Assistant)

Ever since the outbreak of the conflict in Syria in 2011, Turkey has witnessed a major influx of Syrians arriving to Turkey. The official figure as of November 2018 is 3,594,232 Syrians registered under the temporary protection status. Address registration for this special population is handled by the Directorate General of Migration Management, whereas the institution responsible for nationals is the Directorate General of Civil Registration and Citizenship Affairs.
This dual system is a potential frame problem for surveys. Not all cities completed the new round of registrations for Syrians that began in 2016, and not all updated addresses fit the format of the National Address Database. Accordingly, only a part of Syrian households are in the ABRS. This being said, how many Syrian addresses appear on the registration system frame is not officially declared.
We conducted a listing operation and updated 3 random blocks of approximately 100 addresses of occupied dwelling units each in a large district of Ankara where Syrians are known to reside. Address blocks are made up of proximate streets, and are used as enumeration areas in Turkey.
We found that a part of the addresses that were not included in the frame as occupied dwelling units were in fact occupied by Syrian households. We identified a total of 30 Syrian households in three blocks, and only 6 appeared on the frame.
We concluded that obtaining a representative sample of Syrians as a subgroup of a household survey in Turkey that uses ABRS based frames is not possible. Listing of random clusters to cover Syrians households, even in areas where they tend to live more, does not seem to be a cost effective method either. We suggest designs to incorporate Syrian households would need to deviate from classical methods of sampling households in Turkey.

Establishing a Panel Study on Refugees – Insights from the IAB-BAMF-SOEP Survey of Refugees

Dr Manuel Siegert (Research Center “Migration, Integration and Asylum” at the Federal Office for Migration and Refugees) - Presenting Author
Mr Jannes Jacobsen (German Institute for Economic Research (DIW Berlin) )

In 2015 the research consortium of German Socio-economic Panel (SOEP), Research Centre of the Federal Office for Migration and Refugees (BAMF-FZ), and the Institute for Employ-ment Research (IAB), started the IAB-BAMF-SOEP Survey of Refugees. It is a panel study representative for asylum seekers that migrated to Germany between the years 2013-2016. The net sample of the panel started with a total of 4,527 face-to-face individual interviews with adult respondents living in 3,336 households. The response for the first wave was striking compared to cross-sectional surveys (over 50%), whereas the second wave had a comparably high panel attrition (66% turnout), a trend that seems to continue with wave 3 which is currently in the field.

Due to the fact that much policy advice regarding migration relies on survey data, panel attrition calls into question the quality of such surveys and increases the need for solu-tions on how to acquire panel stability. Due to the fact that refugees are still a new and barely surveyed target population there is no gold standard on how to achieve this.

Against this background, we will illustrate the development of the panel and the special conditions that are conducive or hindering in this respect. For this purpose we will assess the survey in regard of burden and questionnaire design. Further, we report qualitative feedback from the field, which points toward the direction that refugees are increasingly suspicious and unwilling to participate due to (attempted) fraud or additional requests from other surveys. We will show which sub groups are why especially prone to drop out and discuss possible solutions to ensure panel stability.

Panel Attrition and Accessibility in Longitudinal Surveys on Refugees

Ms Katharina Sandbrink (infas Institute for Applied Social Sciences) - Presenting Author
Mr Michael Ruland (infas Institute for Applied Social Sciences)

Panel surveys face a number of challenges when it comes to maintaining the cooperation of respondents in longitudinal surveys. For the quality of a panel study, the highest possible degree of coverage and avoidance of selective refusals are particularly important. How these goals can be achieved depends on study design, target group, and survey content. Above all, the accessibility of the respondent is a central aspect next to their willingness to participate.
This paper focuses on measures to stabilize the participation of a hard-to-reach population to be analyzed by means of a refugee panel study’s fieldwork strategies.
Recent studies on refugees indicate that in this specific population the cooperation rates within the initial survey are above average. At the same time accessibility throughout the course of the panel survey is a significant challenge. This is caused by the highly mobile target population and their frequent change of contact details.
Based on the ReGES study (Refugees in the German Educational System) we describe the communication strategies used in the follow-up waves in order to increase panel loyalty, thus minimizing panel attrition and selectivity.
The first wave was conducted face-to-face by native-speaking interviewers. The second wave was conducted by telephone about 6 months later. About 60 percent of the panel-willing cases had provided a telephone number during the initial interview. The cover letter of the second wave and additional communication measures (address update letters, emails, push messages) asked the respondents without an existing telephone number to submit their current telephone number.
It was possible to increase the proportion of panel-willing respondents with telephone numbers to nearly 75 percent solely on account of these tracking measures. This paper analyzes which communication strategies were implemented in the study to which success. Selectivity analyzes are used to rate the effect of those additional communication measures.