3 rd ETHMIGSURVEYDATA TRAINING SCHOOL of COST ACTION CA16111
INTERNATIONAL ETHNIC AND IMMIGRANT MINORITIES’ SURVEY DATA NETWORK (ETHMIGSURVEYDATA), not open for the general public.

14–15 November 2019

Description of the Sessions

Aniko Bernat, TÁRKI Social Research Institute, Budapest
How to avoid the traps: conceptual and methodological issues of Roma and migrant surveys Abstract: The presentation provides an overview and comparison on how the main steps of a Roma and a migrant survey should be designed with a focus on the social, ethical and legal factors relevant for the two target groups. The lecture will thus highlight some methodological lessons on how to survey the Roma and migrants as a hard-to-survey and hard-to-reach group, covering the following topics: identification of the target group, sampling strategies, getting in contact, persuading respondents, interviewing and ethical concerns. The lecture will be supported with examples from recent Roma and migrant surveys from Eastern Europe.

Vera Messing, Centre for Social Sciences, Budapest
Dilemmas of identifying the population in the focus of our research and challenges of sampling Abstract: European countries as well as the EU are striving for comparable and reliable statistics about the ‘Roma’, which is a precondition to efficiently support the design and implementation of national and EU wide inclusion strategies as well as monitoring their outcomes. However, reliable data on Roma population is scarce because it is a difficult-to-survey population for various reasons. Firstly, the identification of the group – who should be considered as ’Roma’ by a survey - is a challenging dilemma. The other dilemma of the definition of the target group involves an answer to the following question: “What for do Roma need to be surveyed?”. If it is about their living conditions, attitudes or opinions then self-ascription of ethnic identity seems to be an appropriate method. But if it is about discrimination and persecution – one of the major sources of their vulnerable situation – then the definition of the direct environment has to be considered as another suitable approach to identifying Roma. And these two definitions result in highly divergent outcomes in terms of constructing a representative sample of the population, which then leads to greatly diverging outcomes in terms of the most essential indicators, such as the size of the population, geographical dispersion, level of poverty, level of education and employment rate. The first part of the session will consider challenges related to the identification of the target population of the research (‘the Roma’) and the dilemmas related to the sampling of a survey. The second part of the session will focus on a case study: the methodological challenges of researching a highly debated phenomenon, namely Roma migration and the referred subgroup of Roma: the “Roma migrants”. The aim of this part of the session is to deconstruct the concept of the “Roma migrant”, and to present the dilemmas researchers in the field need to respond prior to their research. In the past decade the migration of Roma people got to the forefront of political and media discourses within the European public sphere. In order to provide academic sanity and objectivity to this phenomenon, it is of utmost importance to overview the conceptual and methodological dilemmas of researching “Roma migration” and set the basic framework within which this phenomenon has been and can be academically discussed.

Stefano Piemontese, Institute for Research into Superdiversity of the University of Birmingham and CER-M Research Centre for Migration Studies, EMIGRA Research Group on Education,
Migration and Childhood Targeting “Roma” in policy and research: conceptual, methodological and ethical dilemmas
Abstract: With the adoption of an EU Framework for National Roma Integration Strategies in 2011, Member States were requested to develop inclusion plans targeted at their Roma populations. The scope and limits of these policies, however, were largely conditioned by the divides between colour-blind and identity-policy traditions, and between the recognition of Roma as a domestic or European issue. In other words, a variety of understandings and definitions exists about “Who the Roma are”, and this diversity triggers very heterogeneous political approaches and actions. Also, the knowledge production about Roma, which is supposed to inform the design and implementation of these policies, is affected by meaningful dilemmas: conceptual, methodological as well as ethical. Indeed, the adoption of rigid, policy-oriented categories for sampling and surveying populations with multiple and threatened identities, like the Roma, run the risk of essentialising these groups and to invisibilise other important sources of social differentiation. Using ethnographic cases as discussion material, the seminar summarises the various methodological decisions that policy-makers and researchers need to make when targeting Roma in their work.

Andrey Ivanov, European Union Agency for Fundamental Rights
What data for what indicators? Approaches to collecting data on the situation of Roma. 
Abstract: Data collection does not take place in a vacuum; it serves the purpose of populating specific indicators. Thus, the first part of the session will introduce the Roma inclusion policy frameworks and the indicators that are needed to monitor progress - and, respectively, the data needed to populate such indicators. The second part will address the challenges that make collecting data on the situation of Roma (and on measures targeting Roma) a particular challenge. These include the definition of the universe of study, sampling frames and constructing a sample, different levels of comparability and specific challenges related to the fieldwork. In part three of the session the pros and cons of different approaches will be presented using examples from UNDP, FRA and UNICEF surveys targeting Roma as well as EU SILC and LFS in countries using identifiers of ethnicity.

Steffen Pötzschke, GESIS – Leibniz Institute for the Social Sciences, Department
Survey Design and Methodology. Introduction to the use of Facebook as a sampling tool in migration research
Abstract: Sampling migrants into a survey is a challenging endeavor. This holds especially true in crossnational research (where comparative sampling frames might not exist in all countries) and in surveys of small or scattered populations (so-called hard-to-reach populations, where established techniques might be less efficient and, consequently, very cost-intensive). The advent of social networking sites presents scholars with new means to face these challenges. One such possibility consists in the use of Facebook advertisements to sample migrant populations. This approach is especially promising as it can be used to reach a multitude of target populations in a high number of countries at comparatively low costs. This session will discuss the advantages and limitations of this new technique. Furthermore, the participants will learn how to set-up a targeted advertisement campaign and be introduced to general survey design aspects they should take into consideration when using this sampling approach. The course builds on findings of a survey which used Facebook advertisements to sample Polish migrants in four European countries. A total of 1,103 individuals completed the relatively extensive questionnaire within a field period of eight weeks. The sampling costs amounted to €500.

Stefánia Toma, Romanian Institute for Research on National Minorities
The Romanian Roma in Quantitative Surveys: looking for Schrödinger’s Cat
Abstract: Research and evidence-based policy making aimed at Roma faces serious limitations that stems in methodological, epistemological, empirical, ethical, etc. considerations in collecting ethnic data. The session explores issues related to these problems in collecting data about the Roma population in Romania (and in other European countries generally) highlighting methodological difficulties and putting it into broader social (e.g. migration), political (e.g. identity politics) and policy context (e.g. social inclusion). After a short overview of several controversial issues (use of ethnonyms, types of identification, sampling, mixed methods, etc.), the presentation will provide specific examples of such methods from several researches in Romania that had Roma subjects.