BACK TO SCHEDULE SPECIAL

Session 6

Towards More Inclusive Data Collection for European Statistics

3 June 2026
14:15 – 15:45
ŠIBENIK II

Presentation title
Measuring Discrimination: Enhancing Data Quality and Reducing Respondent Burden by analyzing Paradata. The Istat Discrimination Survey
Measuring discrimination in the general population poses substantial methodological challenges due to the sensitive nature of the topic, the difficulty of reaching specific population groups and the risk of placing a high burden on respondents.

Read more Read less

These challenges may negatively affect both participation rates and data quality.

This paper presents the efforts of the Italian National Institute of Statistics (ISTAT) to ensure the application of the SDG's core principle 'Leave No One Behind' in redesigning and implementing the second edition of the national Survey on Discrimination (covering prejudices, stereotypes and experiences of discrimination). The paper mainly focuses on the methodological innovations introduced to enhance the survey’s quality. It highlights the survey's evolution from its first edition in 2011 to the 2025 edition, drawing on empirical evidence from a pilot study conducted in 2022-23. The pilot study was designed to test the feasibility of introducing Computer-Assisted Web Interviewing (CAWI) as a data collection method for the first time and to evaluate its effect on response behavior and data quality. Particular attention was given to sub-populations that are typically characterized by lower response rates (e.g. older people, foreign residents, and individuals living in southern Italy).

In addition to analyzing response and non-response rates, drop-outs and qualitative feedback from respondents, the pilot survey emphasized systematically analyzing paradata. Indicators such as access frequency, device and browser type, completion time, screen size and patterns of questionnaire navigation were used to improve understanding of respondent behavior and survey usability. The results provided concrete evidence of the growing importance of clear and accessible wording, and mobile-friendly design to reduce the cognitive and time burden.

Based on these findings, the 2025 survey used a mixed-mode design combining CAWI (Computer-Assisted Web Interviewing), CAPI (Computer-Assisted Personal Interviewing) and paper SAQ (Self-Administered Questionnaires). This approach was supported by reminder protocols and comprehensive sampling strategies enabling analytical depth to be preserved while limiting the response burden on individuals.

Although data processing is ongoing, preliminary results suggest promising response rates across modes. Overall, the Italian experience demonstrates that paradata-informed survey design, mixed-mode data collection and modular questionnaires can foster more inclusive participation, reduce respondent burden and improve the quality of data in surveys that address sensitive topics, such as discrimination.

Main author / Presenter
Francesca Brait, Vanessa Ioannoni
Italian National Institute of Statistics

Read more Read less

Francesca Brait is a statistician and senior researcher at the Italian National Institute of Statistics (ISTAT). Since 1997, she has led projects in demographic and educational statistics and has gained extensive expertise in CAWI household and individual surveys, publishing numerous scientific papers. She currently coordinates the Discrimination Survey (second edition, 2025) and represents Istat in the Praia Group Task Team on Non-discrimination and Equality and in the Eurostat’s Equality and Non-discrimination Task Force. Vanessa Ioannoni is a senior researcher at the Italian National Institute of Statistics (ISTAT). She also holds a Specialization (equivalent to PHD) in Methods and Techniques for Social Research. In ISTAT since 2001; her research mainly deals with equality data and “hard-to-reach” populations, producing national and international scientific publications. She represents ISTAT within the EC Subgroup on Equality Data, the UN Praia Group Task Team on Non-discrimination and Equality and the Eurostat’s Equality and Non-discrimination Task Force.

CO-AUTHOR:

Presentation title
Challenges of measuring ethnic origin in the Hungarian LFS
There is a strong legal and policy framework at EU level for promoting equality and combating discrimination based on racial or ethnic origin, however, available data to monitor this type of inequality is very scarce.

Read more Read less

Therefore measuring racial or ethnic origin in social surveys is of outmost importance in Europe and existing country examples play a pivotal role in harmonising data collection.

The Hungarian Central Statistical Office (hereinafter HCSO) has got long-standing experience in measuring ethnic origin namely ethnic identity in the census, and since 2013 in the Hungarian Labour Force Survey as well.

First, we introduce why and how HCSO started to measure ethnic identity in the LFS, and discuss briefly, why migrational background as a standard variable in ESS cannot be used to monitor disrimincation against the Roma and therefore it is an inefficient proxy of ethnic origin in Eastern Europe. We introduce how the questions of HCSO allow the possibility of double (self-)identification which was received very well among the Hungarian ethnic minorities, and discuss in a nutshell how the wording of the questions changed to be in line with Hungarian census last year.

Next, and most importantly, we examine the quality challenges HCSO faces and the opportunities it can explore by measuring ethnic identity in LFS touching upon the following topics: sampling and estimation strategy (and the absence of these), decreasing response rate of the Roma, proxy answers and publication strategy.



Finally, we present some key figures to demonstrate why collecting this type of equality data is inevitable to monitor labour market discrimination of the Roma population in Hungary - and in Europe.

Main author / Presenter
Réka Kemény
Hungarian Central Statistical Office

Read more Read less

Réka Kemény holds a Master’s degree in Ethnic and Minority Studies from Eötvös Loránd University, Budapest. She works as a senior statistician in the Employment Statistics Section of the Hungarian Central Statistical Office (HCSO), where she specialises in the adaptation of ad hoc and regular modules of the EU Labour Force Survey. Her professional interests include labour market statistics, questionnaire design and cognitive interviewing, as well as equality and non-discrimination statistics, with a particular emphasis on gender inequalities and the discrimination of the Roma population. She participates in several international working groups and task forces, including the Eurostat Equality and Non-discrimination Statistics Task Force, and coordinates EU-funded projects within HCSO. Holding a Bachelor’s degree in adult education, she actively supports training and lifelong learning initiatives within her organisation.

Presentation title
A review of methods and participatory approaches to improve data quality and coverage on people at risk of discrimination
Enhancing and expanding statistics on people at risk of discrimination, alongside advocating for more inclusive, participatory, and representative data collection methods are crucial for creating evidence-based public policies.

Read more Read less

It is vital to have more granular data to monitor policies and meet community needs in line with the 2030 Agenda for Sustainable Development and EU equality strategies, which stress the significance of disaggregating indicators by protected grounds of discrimination.

The primary challenges in surveying people at risk of discrimination include: (a) conceptual issues or insufficient information for identifying target populations, (b) low representation and relatively low incidence in the population, (c) absence or incompleteness of sampling frames, (d) obstacles to disclosing group membership, (e) methods and techniques of data collection, (f) recruitment, and (g) gaps in administrative data coverage.

This papers draws from the work of the Eurostat Equality and Non-discrimination Task Force to explore key methodological solutions for surveying groups at risk of discrimination, including: (1) sampling techniques such as non-standard probability sampling, hybrid survey designs, data sources integration and the use of new data sources like big data; (2) strategies to enhance response rates and minimise survey representation and measurement errors (e.g. employing methods like paradata and mixed-mode surveys); and (3) engaging of civil society organisations (CSOs) and civic participation at different stages of the data collection process, including dissemination and use of data.

The discussion on improving statistics for diverse populations intersects with current debates in official statistics concerning recent trends, such as declining response rates in probability surveys and the increasing demand for ‘real-time’ statistics alongside the rise of non-probability data sources and data produced by citizens.

Concrete examples and methodological proposals are explored highlighting the possibilities and challenges of different solutions. Ultimately, the aim is to identify the most effective strategies and practices for improving statistics on people at risk of discrimination. This involves ensuring high-quality data characterised by participatory approaches, robust statistical representativeness and informative insights.

Main author / Presenter
Eugenia De Rosa
Eurostat

Read more Read less

Eugenia De Rosa, Ph.D. in Methodology of Social Science, is currently Seconded National Expert specializing in Equality and Non-Discrimination Statistics, as well as and Gender Statistics Coordinator at Eurostat, European Commission. Since 2013 she has been working as researcher at the Italian National Institute of Statistics (Istat) where her work focuses on gender analysis and groups at risk of discrimination (e. g., LGBTIQ+ people, Roma people, homeless people). From 2018 to 2024 Dr. De Rosa was research manager of the Istat-UNAR (National Office against Racial Discrimination) project on ‘Labour discrimination against LGBT+ people and the diversity policies implemented in enterprises’. Since 2024 she teaches demography at Tor Vergata University in Rome. Dr. De Rosa's extensive expertise encompasses gender issues, SOGIESC indicators, inequality and discrimination, and methodology of social science.

CO-AUTHOR:

Presentation title
Towards Inclusive European Statistics: Community Engaged Data Collection
Robust European statistics on equality and fundamental rights require not only sound methodology but engagement with communities and CSOs across the research cycle—from design and data collection to dissemination.

Read more Read less

Drawing on FRA’s flagship surveys—EU MIDIS, successive Roma and Travellers surveys, the EU LGBTIQ surveys, and FRA’s antisemitism surveys—this presentation shows how participatory approaches enhance data inclusivity, relevance and impact for policy and practice.

We present a practical, evidence informed approach for engaging communities in survey work, grounded in the EU fundamental rights framework and applicable across national statistical systems. Early, structured collaboration improves policy relevance, reduces nonresponse and measurement error, and strengthens legitimacy. A five phase engagement model structures the approach: (1) co-define problems and indicators; (2) co-design instruments via iterative cognitive testing, multilingual adaptation and accessibility checks; (3) co-produce sampling and fieldwork that combine probability designs with targeted outreach to hard to reach groups; (4) co-deliver fieldwork through trained community mediators and culturally competent interviewers; and (5) co-interpret and disseminate results with feedback loops, open microdata where feasible, and shared ownership of recommendations. Governance safeguards—conflict of interest rules, GDPR compliant consent, do-no-harm protocols and transparent documentation—prevent tokenism and protect participants.

FRA operationalises this model through expert panels, CSO consultations and piloting, ensuring questions reflect lived experiences of discrimination, victimisation and unequal treatment. In the LGBTIQ surveys, civil society co-shaped outreach, language and recruitment pathways. Dissemination via trusted networks, online platforms and social media enabled participation among groups often missing in household surveys, including non-binary persons, ethnic minority LGBTIQ individuals and those in less visible or hostile environments. Clear guarantees on anonymity and data protection build trust. Comparable methods informed antisemitism surveys, conducted with Jewish organisations and community representatives to co-design communications and recruitment materials. Partnerships spanning formal institutions and informal networks increased visibility and participation. Multilingual instruments and endorsed messaging improved coverage and response. Roma and Travellers surveys demonstrate long standing engagement: local mediators, interviewers from within communities and tailored training foster cultural sensitivity and ethical practice. EU MIDIS integrated civil society and academic expertise into sampling and fieldwork to improve accessibility for migrants and minorities and bolster the validity of self-reported discrimination data.

Engagement extends beyond data collection. FRA involves CSOs in disseminating and interpreting results through co-presentations and dialogues, which strengthen transparency, ownership and evidence-based policymaking. FRA’s experience shows that participatory, community engaged survey methods are essential to producing policy relevant European statistics and practice.

Main author / Presenter
Rossalina Latcheva
EU Agency for Fundamental Rights (FRA)

Read more Read less

Rossalina Latcheva, is currently Head of the Equality and Non-Discrimination Sector in the Equality, Inclusion and Sustainability Unit at the EU Agency for Fundamental Rights (FRA). In this role, she coordinates and oversee sector’s main activities: research and data collection, legal and policy analysis, capacity building and technical assistance to EU institutions and Member States. This includes undertaking qualitative and quantitative research, carrying out EU-wide large-scale surveys, strategic thinking and initiatives by monitoring and analysing EU policy and legislative files, as well as economic and social developments and trends. Rossalina is experienced in developing, collecting and analysing primary and secondary survey data, as well as in applying and mixing diverse quantitative and qualitative research methodologies. Rossalina holds a Ph.D. in sociology from the University of Giessen in Germany (awarded with Summa Cum Laude), and a Master of Sociology from the University of Vienna.

CO-AUTHORS:

Presentation title
Advancing inclusivity in European statistics: piloting SOGIESC variables in a large-scale cross-national survey
Ensuring that official statistics leave no one behind requires data that are disaggregated along relevant protected grounds of discrimination.

Read more Read less

While sex-disaggregated data are generally well established in European statistics, information on gender identity and sexual orientation remains largely absent, limiting the capacity to identify and address intersecting inequalities. This paper presents a pilot exercise carried out by the European Institute for Gender Equality (EIGE) in the second wave of its Survey on gender gaps in unpaid care, individual and social activities (CARE Survey), which for the first time incorporated variables on sex, gender identity and sexual orientation.

The survey was conducted in all 27 EU Member States through non-probability online panels, covering over 65,000 respondents aged 16–74. It provided an unprecedented opportunity to test the feasibility, data quality and analytical value of enhanced SOGIESC data collection in a large general population-based survey.

This work assesses the practical feasibility of collecting these variables at a large scale, including response patterns, item non-response, and discusses implications for overall survey quality.

Methodologically, the analysis evaluates the use of a two-step approach to measuring sex and gender identity, and discusses terminology, question wording and response options. Particular attention is paid to measurement challenges specific to SOGIESC variables in large scale surveys, including small sample sizes, disclosure risks and potential misclassification.

The work also addresses key considerations for analysis and dissemination, including responsible reporting and the balance between visibility and statistical robustness.

The findings contribute to ongoing discussions on improving inclusiveness in European official statistics while maintaining high standards of data quality and usability.

Main author / Presenter
Irene Rioboo Leston
European Institute for Gender Equality

Read more Read less

Dr Irene Rioboo is a researcher at the European Institute for Gender Equality (EIGE), where she works on the Gender Equality Index and the monitoring of the Beijing Platform for Action in the EU. Previously, she worked at Eurostat and Eurofound on policy-oriented research on socio-economic inequalities and has over 15 years of experience in academia as a Professor of Applied Statistics.

CO-AUTHOR:

Presentation title
Enhancing the Availability and Quality of Equality and Non-Discrimination Statistics: Insights from the Finland National Data Ecosystem
This paper examines Finland’s national data ecosystem on equality and non-discrimination, focusing on the structures, coordination mechanisms and methodological approaches that support the production of high‑quality statistical data.

Read more Read less

In Finland, multiple institutional actors generate population‑level data on equality and non‑discrimination, complemented by targeted surveys addressing specific population groups, such as students and immigrants. Strengthening national coordination is a shared priority among decision‑makers, data producers and NGO’s. The ministerial level coordination is facilitated through a national system for monitoring discrimination. The paper provides an overview of the Finnish data ecosystem, highlights effective practices in inter‑institutional collaboration, and discusses key challenges related to data availability and quality. Particular attention is given to the use of paradata in monitoring the accessibility of the survey process, as well as adaptive design in large-scale surveys—such as the Gender-based Violence Survey (GBV) and the Labour Force Survey (LFS)—to reduce bias, increase response rates and enhance the robustness of equality-related statistics.

Main author / Presenter
Marjut Pietiläinen
Statistics Finland

Read more Read less

Senior Researcher MP is an expert in equality, non-discrimination and family statistics, and serves as the gender statistics team leader at Statistics Finland. She is the co-chair of the Praia Group Task Team on Non-Discrimination and Equality and participates in several international expert groups, including the Eurostat Task Force on Equality and Non-Discrimination Statistics, the FRA Steering Committee on Surveys, the EIGE Experts’ Forum, the UNSD IAEG-GS, and the UNECE Steering Groups on Gender Statistics and Statistics on Children. At the national level, she contributes to multiple advisory bodies, such as the Committee for Combating Violence against Women and Domestic Violence and the Expert Group on Monitoring Discrimination. Her latest research projects include the Gender-Based Violence (GBV) Survey, the Gender Equality Barometer and a national study on pregnancy discrimination in Finland. Her personal research interests focus on discrimination in working life, with a particular emphasis on improving data quality.

CO-AUTHORS: