Transforming disability data collection in Wales: Embracing the social model

I am excited to share the results of this project, commissioned by the Welsh Government, to provide an alternative way of measuring and collecting data on disability based on the social model of disability. The new questions are grounded in disabled people’s lived experience and aim to improve data about disabled people in Wales.

The medical model of disability – used in the Equality Act (2010) - regards disability as a health condition or impairment inherent to an individual. The social model, by contrast, says people are disabled by society, not by their impairments or health conditions. Welsh Government formally adopted the social model of disability in 2002, prompting a search for new ways of collecting data that capture disabled people’s lived experiences.

The initial discovery phase of the project involved workshops with data users, a review of existing survey questions measuring disability, a rapid literature review, and deliberative workshops with disabled people. This respondent-centred design approach ensured we gathered a wide range of insights.

What we found was that the existing survey questions did not capture the nuances of the social model. No single set of questions could meet everyone’s needs – different stakeholders had conflicting priorities, making it impractical to create a comprehensive question set. These early findings, published in the interim report, echoed issues raised in an Office for National Statistics (ONS) review of the current GSS harmonised disability standards.

Designing inclusive questions

The ‘alpha’ phase built upon the findings and recommendations of the discovery phase. It involved developing and testing:

1. an updated version of the GSS harmonised impairment standard and;
2. new questions capturing disabled people’s experiences of societal barriers.

The design process was iterative and collaborative, involving feedback from disabled people, the Welsh Government and a wider research Steering Group. Their guidance and feedback were invaluable in refining the questions.

The new questions were designed to be inclusive in the following ways:

Mode: They work for both self-completion (e.g. online surveys) and interviewer-led formats (face-to-face).

Audience: They are understandable to adults answering for themselves and to parents answering on behalf of young children.

Languages and formats: They are available in Welsh and English and in British Sign Language (BSL), and Easy Read versions to support wider participation.

Developing the impairment question

The starting point was the existing GSS harmonised impairment question with the aim to modify it to align with the social model. In practice, this meant broadening the scope of what is asked. For instance, we added a category for ‘being in pain or discomfort’ – a commonly reported issue that wasn’t explicitly covered before. Additionally, an instruction was added to the question stem to better capture fluctuating conditions. The updated question aims to capture a wide range of impairments without implying that the person is ‘the problem’.

Survey questions on experience of barriers

We ended up with two new questions to capture people’s experiences of barriers. The first question asks whether disabled people face difficulties across various domains of life – for example, in healthcare, employment, transport, or public spaces – similar to some international surveys. The second question focuses on the type of barrier: it asks if people experience attitudinal barriers, institutional barriers, environmental barriers, or communication barriers.

Cognitive testing

We conducted cognitive interviews, in both Welsh and English, to see how people understood and answered the draft questions. Our sample included disabled adults as well as parents of disabled children aged 0–15. We also carried out smaller ‘proof of concept’ tests with specific groups: one with d/Deaf BSL signers, and another with people who use Easy Read communication. The pretesting allowed us to check that the questions were workable and clear and to catch any problems with how the questions were interpreted or administered, before moving on to a larger-scale survey.

Quantitative field testing

The final stage of our project was a quantitative field test using NatCen’s Panel survey. We collected responses from over 2,500 adults, including an extra boost of participants in Wales to ensure robust data for that population. In this survey, everyone answered the new social-model-aligned questions on disability as well as the existing harmonised standard questions. This allowed us to calculate how common different impairments and barriers are (using the new questions) and compare those results to estimates from the current harmonised standards.

Key findings from the field test

• Higher measured impairment rates: The new social-model-aligned impairment question identified 57% of respondents as having an impairment, compared to 35% under the current harmonised standard question.
• New issues captured: About half of disabled people reported experiencing “being in pain or discomfort” – a category not explicitly covered by the harmonised standard.
• Multiple barriers: Barriers in healthcare and in public spaces or facilities were among the most commonly reported. Indeed, most disabled people said they face barriers in more than one aspect of life.
• Types of barriers: When asked about what kinds of barriers they encounter, attitudinal barriers (like people’s prejudices or stereotypes) and institutional barriers (like policies or systems that aren’t accessible) came up most often. This underscores how much social attitudes and structures can disable individuals.

Our approach to estimating disability is novel in that it defines someone as ‘disabled’ if they have both an impairment and experience one or more barriers. This combined criterion is a significant departure from the status quo (which was rooted in the medical model’s focus on health conditions alone). Using this new measure of disability raises methodological questions about who is counted as disabled and who is being excluded. It also provides policy makers with data that may help them formulate new policies aimed at breaking down the societal barriers disabled people face.

Moving forward

This research is a positive step towards more inclusive and accurate data collection on disability in social surveys. The new survey question on impairments addresses many of known issues with the current GSS harmonised impairment standard. The new questions on barriers are innovative but new to the field. We encourage researchers to try out these new questions and to continually review how they perform. Wider use will build evidence, and ongoing feedback will ensure the measures remain relevant and effective.

We’re already seeing an impact from this work. The question set is informing the survey design feeding of a DWP survey, with findings feeding into the Timms Review of Personal Independence Payment (PIP) (due to report in autumn 2026). Our research findings will also provide useful insight to the ONS when it resumes its review of disability data. In short, we hope these questions mark the beginning of a shift in how disability is measured.

Further developments and next steps

There is great potential for this research to be the beginning of a more inclusive way of involving disabled people in the development of survey questions. For example, further co-design and testing with d/Deaf BSL signers would ensure the new questions work well in BSL. Similarly, doing some additional testing with people who benefit from written content being presented in an Easy Read format would strengthen our confidence that these questions are effective for this group. Developing equivalent questions for children and young people would address a current gap in data collection and ensure their experiences are captured too. 

Acknowledgements

We want to thank our partners at Welsh Government, the project’s steering group, and our collaborators Social Research with Deaf People (SORD), University of Manchester, and Prof Sally McManus, City St Georges, University of London. Most of all, thank you to all those who participated in this project – your input was invaluable. 

For more insights into our approach and the practical challenges we faced, read Alessandra Radicati's blog post about the deliberative workshops held with disabled adults in Wales. It describes how we navigated accessibility requirements to gather valuable input from participants to inform the question design. Additionally, the full research reports from this project are available here – and we encourage anyone interested to read them.