Rethinking Disability Measurement in Surveys

The social model of disability recognises that people are disabled by barriers experienced in society, not by their impairments.

Despite the growing popularity of the social model, most national and international social and health surveys continue to measure disability using medical model-based definitions, which frame impairments as the disabling factor.

These data are used to inform policy and decisions relating to disabled people, and the delivery of services across Wales.

The Welsh Government commissioned the National Centre for Social Research (NatCen) to develop and test a suite of social model-aligned survey questions, offering a viable alternative to medical model approaches that currently dominate disability data collection in the UK and beyond.

The methodology followed the Respondent Centred Design Framework, putting the participant at the centre of the research.

The discovery phase established the needs of data users; reviewed existing UK and international survey measures; and, through deliberative workshops, explored disabled people’s requirements for any new survey questions.

The alpha phase commenced with iterative question design to improve on the current UK harmonised standard on impairments and to draft new questions on the barriers experienced by people with impairments.

Questions then underwent cognitive testing, with 56 interviews conducted in English and Welsh, with adults, and parents of disabled children. This included early stage British Sign Language (BSL) and easy read testing.

Quantitative field testing (n=2,507; 57% response rate) on the NatCen opinion panel enabled comparisons between social model and the more conventional harmonised medical model-based estimates of disability. Two, alternative, but complementary, social model-based definitions of disability were constructed and measured, collectively capturing the types of barriers experienced as well as the areas of life these are experienced in.

No significant differences were observed in the percentage of the population that were categorised as disabled according to each model.

Findings demonstrate the efficacy of the social model approach in collecting more inclusive impairment data, while providing new insight into the types of societal barriers faced by disabled people.

The new questions are recommended for wider adoption in surveys and data collection to help build evidence on how they perform.

Further developments are proposed for BSL, easy read versions, parent versions, and face to face data collection.