This think piece forms part of our World Statistics Day 2025 celebrations. In line with this years theme, “quality statistics and data for everyone”, our guest writer, Rochelle E. Tractenberg, gives her take on moving forward with ethical statistical practice. Learn more about Rochelle below this think piece.
Every five years, World Statistics Day is an event that celebrates all that the practice and profession of statistics and data science contribute to evidence-informed decisions, democratic accountability, human dignity and flourishing, and sustainable development worldwide. These contributions underpin an ever-widening array of applications in science, industry, and government, as well as our everyday lives. The ubiquity might lead to lower recognition by the public of the importance and impact of statistical work. As professionals, we rightly applaud methodological advances, successful projects, and new data sources; World Statistics Day 2025 challenges us to make their public value visible. In 2025, the theme is “quality statistics and data for everyone”. This theme suggests that “quality statistics and data” are already available for everyone, and we’re celebrating that. However, the quality dimension and the availability of quality statistics and data don’t just happen; these require diligence, care, and competence at all levels by those who work in and with statistics and data science.
Those working in science, industry, and government settings are all dependent on the outputs of statistics and data science for a wide range of decisions. Within a specific context, trust in these outputs may be established in context-specific ways, for example, by ensuring objective criteria are used to select and hire ethical and competent practitioners, and to support a workplace culture that prioritises excellence and ethical statistical practice. But when it comes to earning the public’s trust in statistics and data science, we all recognise that trust in our work does not flow from numbers alone. It must be earned, day by day, by practitioners who are committed to ensuring the quality of both their statistical outputs and the data that underpin them. World Statistics Day 2025 is an important opportunity to consider how to increase the visibility of this commitment, the work and outputs, and their accessibility and utility for everyone.
The UK Statistics Authority exists to promote and safeguard official statistics that serve the public good. The Office for Statistics Regulation (OSR) is the Statistics Authority’s independent regulator: it assesses statistics against its Code of Practice. OSR conducts assessments of official statistics as needed to ensure that they comply with the Code’s core principles of Trustworthiness, Quality and Value (TQV) and crucially, identify any that do not. This transparency serves many purposes, one of which is to help those producing official statistics monitor the potential for drift in either their data or methods used to collect or analyse it. “Drift” may occur in the data, where variables being collected remain the same but either the meaning/implications or the sources of the data change over time in important ways, or where there are changes in the concepts underpinning data collection in the mental models of respondents, or in the capabilities of the data to reflect the concept itself. If any of these types of drift occur, the most direct effect would be on the Value of the data (no longer reflecting what is of interest). It wouldn’t be immediately apparent that the Trustworthiness of the data – specifically, trust in its representativeness – would also be impacted in the face of drift. Ongoing assessment of statistical products against the TQV criteria, which are concretely outlined within the Code, is an essential part of any ongoing data enterprise.
The Code is being refreshed in 2025 for clarity and technological advances – to examine and correct any drift that has occurred. The core principles do not change, but support for those who follow the Code can be made clearer. Compliance with the Code is mandatory for official statistics produced by UK Government organisations. Any Crown body that is producing statistics is actually producing “official statistics”. This means that all statistics produced by Crown bodies that earn the OSR accreditation will be high quality. Since these statistics and data are designed and intended to serve the public good, OSR and the Code of Practice together support the 2025 theme for the UK.
The Code of Practice is also useful for voluntary application within and beyond government – inside and outside the UK – as a practical framework for behaviours and standards that earn and promote public confidence in statistical work. That utility is clear in the 2025 Royal Statistical Society Award for Statistical Excellence in TQV, awarded to the non-profit organisation Mental Health Innovations for their use of the Code’s principles to clearly communicate to their users – the public – their commitment to ethical work with data. Any individual or group can utilise the Code to help document their commitment to quality statistics and data for their stakeholders. This is one way that any individual or entity can build the public’s trust in the practice and profession of statistics and data science.
Recognising TQV as a first step towards ethical statistical practice
It is important to recognise that “ethical statistical practice” is actually more than just practicing statistics and data science ethically – although this is a critical dimension that can and should be supported widely. The second dimension is the ability to recognise practices that are not ethical, and also the ability to respond to those practices, or requests, that are not ethical. The OSR Code of Practice offers a concrete path towards ethical practice; those who would go further – for example, individuals who lead, supervise or mentor teams of statistical practitioners and data scientists – can look to the American Statistical Association’s Ethical Guidelines for Statistical Practice (which will be refreshed in 2026 according to their 5-year schedule) or the International Statistics Institute’s Declaration on Professional Statistics. In order to more actively step onto the path towards ethical statistical practice, engagement with the 2025 Code of Practice is worthwhile at the start of, and iteratively throughout, any data collection and analysis endeavour.
Trustworthiness
Trust begins with behaviour. This means committing to integrity and independence (collecting and analysing data free from political interference; separating analysis from advocacy; speaking clearly about uncertainty); transparency (publishing sources, methods, limitations, revisions policies, and logging changes); and accountability (setting out responsibilities and routes for redress so users know how standards are upheld). Our behaviour is the first and most consistent way to promote the public’s trust in the outputs of our work.
Quality
Quality is about fitness for purpose across the whole lifecycle. This is reflected by proportionate, well-documented methods aligned to intended use of any data; robust metadata and provenance; appropriate linkage and access controls, including for external or novel data sources; and operational excellence for timeliness, consistency, reproducibility, and secure processing. These aspects of quality may be the least perceptible to the public, but only the highest-quality data and statistical products have the potential to improve everyday life.
Value
Value is realised when statistics meet real information needs. Perhaps one of the least-recognised opportunities to show the public how important our work is to them, and how important their needs are to our work, Value requires active user engagement and co-design; clear explanations of what the statistics mean for decisions in plain language, supported by visuals designed for understanding rather than only for experts; and accessible releases in open formats with reusable metadata so others can build on the work. Effective collaboration with our stakeholders requires, and can also build, trust. Documentation of how our statistical work effectively and efficiently informs policy concretely reflects the value of statistics and data science for society, as well as the value of official statistics for the public good.
World Statistics Day is not only a celebration; it is an invitation. Around the globe, institutions are navigating heightened scrutiny, fast-moving data ecosystems, and challenges to the public’s trust in official statistics. The 2025 OSR Code of Practice offers an actionable, practice-ready framework for organisations and individuals to utilise. OSR accreditation safeguards Trustworthiness, strengthens Quality, and maximises Value. Whether you are in a national statistics office, a local authority, a research institute, a charity or a private data lab, commit publicly to the Code’s principles, map your current practice to TQV, and publish your improvement plan. That is how, on World Statistics Day and beyond, we make the public-interest contribution of official statistics visible, valued, and worth celebrating by the people they are meant to serve.
Related:
Blog: Quality Data, Shared Purpose: World Statistics Day 2025 and the Refreshed Code of Practice
Rochelle E. Tractenberg is a tenured professor at Georgetown University (Washington, DC). Her applied ethics work focuses on strengthening trustworthiness in statistics and data science across research and policy settings. A biostatistician since 1997, she serves on the UK National Statistician’s Data Ethics Advisory Committee (NSDEC), ISI Advisory Board on Ethics, and the Association of Computing Machinery Committee on Professional Ethics. She has written two books on ethical practice, and has contributed to standards for statistics, data science, and mathematics – as well as the forthcoming UN guide, “Ethics in Official Statistics: Foundations and practices to foster public trust”.