IJPDS International Journal of Population Data Science

A new UK-wide consultation has revealed public views on synthetic data and its role in research. The findings, published in the International Journal of Population Data Science (IJPDS), offer ten practical recommendations for organisations who create and share synthetic datasets.

Synthetic data mimics the structure and patterns of real data but doesn’t contain any actual personal information. It’s increasingly used by researchers to test ideas, train staff, and develop computer code - without the risks of handling real, sensitive data. But until now, the public hadn’t been asked what they think about this emerging technology.

To address this gap, researchers at the Centre for Trials Research in Cardiff University undertook the ‘Discussing Data’ project with funding from Administrative Data Research (ADR) UK. Over the summer of 2024, 39 members of the public from across the UK took part in a series of online workshops to explore synthetic data, ask questions, and co-develop recommendations for data owners.

“People came into the workshops with little or no knowledge of synthetic data,” said Dr Rob Trubey, who co-led the study. “But they quickly engaged with the topic and raised thoughtful, sometimes challenging questions about privacy, access, and public benefit.”

Public members expressed concern about the term “synthetic data,” which many associated with something fake or unreliable. They called for clearer, simpler definitions and explanations - especially around how synthetic data is created, what it can and can’t be used for, and how it benefits society.

The idea of researchers and others having open access to synthetic data was met with caution. Most workshop attendees preferred a light-touch registration process, where users explain their intended use and agree to basic terms of use. Public members also wanted reassurance that real people - not just automated systems - are involved in checking synthetic datasets before being shared for use.

The final recommendations cover five key areas: introducing synthetic data, explaining its purpose, creating datasets, managing access and misuse, and communicating clearly with the public. They include calls for real-life case studies, engaging visuals, and inclusive communication formats to help build public understanding and trust.

“This project shows that the public are ready and willing to engage with complex data topics,” said Dr Claire Nollett, co-author and public involvement lead. “But they need clear, honest information and a genuine opportunity to shape how these technologies are used.”

The study highlights a growing need for public dialogue as synthetic data becomes increasingly common in health and administrative research. It also signals a shift in how data governance should be approached - placing public expectations and transparency at the heart of policy.

The research team hopes their findings will guide data owners across the UK in releasing synthetic data responsibly, in ways that align with public values.

“Synthetic data offers huge potential - but only if people trust the way it’s created and used. These recommendations are a step toward earning that trust,” concluded the project co-lead, Dr Fiona Lugg-Widger.

For more information, visit the Discussing Data project page.

 

Click here to read the full article

Dr Fiona Lugg-Widger, Director of Data, Senior Research Fellow, Centre for Trials Research, Cardiff University, Wales, UK

Nollett, C., Lugg-Widger, F., Stock, J., Brookes-Howell, L., Fitzgibbon, J., Johnson, S., Munnery, K., Robling, M., Yameen, F. and Trubey, R. (2025) “Discussing Data: A UK-wide public consultation on the use of low fidelity synthetic data for research”, International Journal of Population Data Science, 10(1). doi: 10.23889/ijpds.v10i1.2962.