BACK TO SCHEDULE REGULAR

Session 10

Quality as a Shield Against Misinformation

3 June 2026
14:15 – 15:45
ŠIBENIK VI

Presentation title
Quality of environmental data: ethical challenges and the fight against disinformation
Quality of environmental data: ethical challenges and the fight against disinformation

Read more Read less

Public statistical services have the task of producing reliable data to inform public debates and policy decisions. To fulfill this mission, they observe the European Statistics Code of Practice, ensuring principles of impartiality, transparency, and methodological rigor in data production. However, in the era of social media, even the most robust figures can be quickly misrepresented, oversimplified, or weaponized. This can erode user trust, undermining the legitimacy of public statistics in societal debates and thus weakening institutional credibility. Environmental data are particularly vulnerable to media distortion due to their complexity and association with polarizing issues. This article aims to illustrate the dangers of environmental disinformation through case studies, and explore concrete solutions by leveraging the European Statistics Code of Practice and quality frameworks specific to statistical services.

Main author / Presenter
Manuela Beaudoin
Service des données et études statistiques

Read more Read less

Manuela Beaudoin is a data protection (GRPD) and Quality Officer at the French Statistical Service for Sustainable Development (SDES, French Ministry of Ecology). A trained lawyer, she has worked in the French civil service since 2021 and joined the SDES in 2025. She leads the quality management process at the SDES and advises on legal issues.

Presentation title
Think TQV: how Trustworthiness, Quality and Value provide an ethical framework for building confidence and combating misinformation
This presentation will show how the Code of Practice for Statistics, developed by the Office for Statistics Regulation (OSR), helps anyone producing or using official statistics to work in ways that build trust and public confidence.

Read more Read less

OSR is responsible for regulating the whole government statistical system in the UK. OSR’s vision is that statistics should serve the public good: this means they should be public assets that provide insight that allows them to be used widely for informing understanding and shaping action. Statistics are more than just numbers: how things are done can matter as much as the numbers themselves. This is where the Code of Practice for Statistics comes in. The Code ensures that statistics are not just numbers, but reliable tools for understanding the world.

The Code of Practice for Statistics is structured around three core principles: Trustworthiness, Quality, and Value (TQV).

Trustworthiness is about building confidence in the people and organisations that produce statistics and data. It covers the integrity and professionalism of the individuals and organisations, and it ensures that statistics are free from manipulation and undue influence.

Quality is about using data and methods that produce assured statistics. It encompasses the use of appropriate data sources and robust methodologies to produce reliable and accurate statistics, but also involves continuous improvement, transparency, and rigorous quality assurance.

Finally, Value is about ensuring statistics support society’s needs for information. It covers the relevance and accessibility of statistics to users, ensuring that they can inform decision-making and public debate.

Together, TQV provide an ethical framework that supports the production of statistics that are technically sound, socially beneficial and that inspire confidence among users.

The latest version of the Code, launched in October 2025, includes new Standards for the Public Use of Statistics, Data and Wider Analysis. These Standards direct public bodies to use statistics transparently, and with integrity, accuracy and clarity. They help reduce the risk of misinformation by ensuring people have equal access to information and that they can hold public bodies to account. OSR uses these Standards to stand up for the role of statistics in public debate, actively fighting misinformation through our casework function, which investigates and takes action to correct misleading use of statistics in the public domain. This presentation will share recent examples of our work and the impact it has had.

Main author / Presenter
Penny Babb
Office for Statistics Regulation

Read more Read less

Dr. Penny Babb – Head of Policy and Standards, Office for Statistics Regulation, UK Penny joined the Office for Statistics Regulation team in 2008. She has reviewed a wide variety of statistics since 2009 and co-authored the UK Statistics Authority’s Quality Assurance of Administrative Data framework. She led on delivering and maintaining the refreshed Code of Practice for Statistics, supporting its wide adoption across the UK. Penny has worked in official statistics for over twenty years, on births, cancer, social inequalities and police recorded crime.

Presentation title
Strengthening Trust and Governance in Official Statistics in the Age of AI
The 2020 European Data Strategy and the Data Governance Act established the legal basis for EU data governance, which is being implemented differently across Member States to ensure organised access to data.

Read more Read less

Although data governance has always been at the core of statistical organisations, national statistical offices (NSIs) are now seeking to position themselves within a new and more complex data ecosystem. Regulation 223/2009, as amended in 2024, explicitly promotes the active participation of NSIs in data governance, while Member States are appointing data authorities, opening data spaces, setting up data offices across public administrations, and defining interoperability rules, security tools and ethical guidelines.

As part of this data ecosystem, the EU Artificial Intelligence Act (Regulation 2024/1689), adopted in June 2024 and applicable from 2025, introduces a minimum governance framework for AI, mainly focused on high-risk systems, but also provides references to ethical principles and best practices. The regulation builds on the Ethics Guidelines for Trustworthy AI adopted in 2019 and encourages the development of codes of practice, including the General-Purpose AI Code of Practice adopted by the European Commission in August 2025. Although the use of AI for official statistics falls mostly outside the scope of the AI Act, these developments raise important questions for the European Statistical System (ESS): to what extent is the ESS affected by these new rules, and should it align with the same ethical and governance frameworks?

Furthermore, the forthcoming 2025 Data Union Strategy is expected to further reshape the data governance landscape by expanding and amending existing legislation, with the aim of strengthening AI innovation through improved access to high-quality, well-organised data. This paper explores whether NSIs are ready to take part in this evolving framework and examines the implications of the Ethics guidelines for trustworthy AI and the General-Purpose AI Code of Practice for the European Statistics Code of Practice.

Main author / Presenter
Yolanda Gómez
INE

Read more Read less

Yolanda Gómez has a Degree in Law, a Master in European Law, a PHD studies Certificate in International Law and a Degree in Art History. Ms Gómez works as Legal Advisor at INE since the year 1999. She is an expert in legal statistical issues such as Statistical laws and Statistical principles, ethics, Code of Practices, privacy and confidentiality or Data Science and statistics, dealing with the preparation of high level international meetings and providing support and advice to the Board members on different national and international legal issues. She is also member of different international statistical groups of UN and EU and have several publications on legal statistical issues. She has participated as speaker in the Quality conferences 2016, 2018, 2022 and 202.

CO-AUTHORS:

Presentation title
Combating data management and misinformation issues arising from the evolution of calendar effects over time in seasonal adjustment
The common assumption when using calendar effects in seasonal adjustment (SA) is that calendar effects are constant over time.

Read more Read less

Present study discusses that the longer the time series, the higher the possibility of economic and social change, which can result in calendar effects evolving over time. In some areas, this evolution may be more apparent: consumers may change their preferences towards different goods in a matter of decades or even years, while company production may be more stable in terms of calendar effects.

Changing consumer habits affect the SA and the indices calculated from it. Yet, SA in official statistics is often considered a routine exercise: statisticians at subject matter departments run it automatically, often without paying attention to the changes in effects calibrated by the methodologists when setting up the models. Firstly, this emphasizes the need for more transparent communication between departments. Secondly, to avoid misinterpretations and to keep data users informed, it would be useful if statistical offices also published the methodology for SA, especially if there are major changes compared to the previous year’s models.

Consumption behavior can be subject to change due to policy changes (such as compulsory shop closures on Sundays that happened in 2015-2016 in Hungary), social tendencies (e.g. the changing popularity of products connected to Easter traditions due to people becoming less religious or less tradition-oriented then before), crises (such as the several waves of the COVID-19 pandemic in 2020-2021 disrupting industries and restructuring the consumption of goods and services in various ways), or due to changes in companies’ behavior (e.g. changing marketing campaigns, the spread of shopping malls, or the surge in online shopping possibilities). Moreover, there may be substantial differences in how the above factors affect different products.

This study is an exploratory analysis investigating the extent to which calendar effects may change over time by turning the lense on the most consumer behavior-related topic, retail sales. Specifically, doing so by analyzing calendar effects in separate time periods in retail sales volume indices in Hungary over the years 2000-2025, deploying the TRAMO-SEATS method in JDemetra+ in accordance with Eurostat regulations. Based on my research, the evolution of calendar effects (Easter effect, leap year effect, trading day and working day effect) is remarkable and differs between different product categories, which is caused by economic and social changes that occurred over the past 25 years.

Main author / Presenter
Zsolt Csáfordi
Hungarian Central Statistical Office (HCSO), Methodology and Innovation Department

Read more Read less

Zsolt Csáfordi is a methodologist at Hungarian Central Statistical Office (HCSO), Methodology and Innovation Department responsible for the seasonal adjustment, temporal disaggregation and modeling of economic and social time series of Hungary, and a member of Confidentiality and Seasonal Adjustment (COSA) – Time Series Analysis Expert Group, with a background in economic research. He earned his PhD degree in Applied Economics at Erasmus School of Economics and Tinbergen Institute, Rotterdam with research in labor mobility, productivity spillovers and inter-industry relatedness, and obtained more research experience in innovation studies and data science at Aalborg University Business School. Previously, he worked for 3 years as Assistant Research Fellow at the Institute of Economics CERS HAS, Economics of Networks Research Unit, and prior to that in economic policy / industry / labor market analyst positions (London, Budapest).