Scotland's Carers, 2026: Main Report

Annex 2: Data and methodology

Data sources

Scotland's Census 2022 methodology

The census is a mandatory national population survey, with an overall response rate of 89.9%. Results were subject to a process called “Edit and Imputation”, to account for the percentage of the population who did not respond. Broadly speaking, however, the census data is viewed as population-level counts. The Census 2022 identified nearly 628,000 carers. Based on this large set of responses, there is value in considering this data in addition to the SHeS, so in this report we have also included results of analysis from the census, for comparison and additional context.

More information can be found on the Scotland’s Census 2022 website.

Scottish Health Survey (SHeS) 2019-2023 methodology

The Scottish Health Survey (SHeS) is a voluntary questionnaire designed to survey a representative sample of the population. SHeS has been conducted annually since 2008 and prior to this was conducted in 1995, 1998 and 2003. An annual report is published for each year of the survey (the latest published being 2024).

The 4-year combined SHeS dataset (inclusive of 2019, 2021, 2022 and 2023 – data from 2020 was excluded because the COVID‑19 pandemic affected comparability with other years) is used throughout the majority of this report. This is because combining the four years of survey data gives a larger total sample size, which enables us to draw more robust statistical conclusions. The 2023 annual data can provide useful context (see section 4.4.3 for further detail); however, the sample sizes are often too small to support more detailed breakdowns. As such, this report predominantly relies on the aggregated 2019-2023 dataset for the majority of the analysis.

There were just under 19,000 adult responses to the unpaid caring prevalence question in the 2019-2023 combined SHeS dataset, with just under 2,900 respondents indicating that they provided unpaid care. Survey weights were then applied to adjust for non-response and to ensure SHeS results reflect the age distribution of the population of Scotland and Scottish areas.

More information can be found on the Scottish Health Survey website.

National Records of Scotland (NRS) Mid-Year Estimates (MYEs)

Population estimates are produced by NRS, based on the census and are updated each year to account for population change from 1 July to 30 June. They are based on the usually resident population which covers people living in Scotland for a period of at least 12 months, whatever their nationality. 

The MYEs are used in this report to produce population-level estimates based on the SHeS data, by applying SHeS percentages to the relevant population according to the MYEs. The MYEs are also used to transform confidence intervals associated with SHeS estimates into population-level limits.

It is worth noting that the population estimates for mid-2023 were revised in August 2025 to reflect updated estimates of long-term international migration from the Office for National Statistics. This planned revision estimated Scotland’s population to have been 5,506,000 at mid-2023, an increase of 15,900 from the previous estimate published on 8 October 2024.

As such, readers may notice that the population-level estimates vary slightly from those reported in the Scotland’s Carers annual update release (March 2025), although estimates are based on the same SHeS data (2019-2023). This is due to the revision by NRS being published between the publication of the annual update, and the production of this report.

More information can be found on the NRS website.

Confidence intervals

Taking into account the complex stratified and clustered survey design, all estimates from the SHeS presented in this report are central estimates, which have associated 95% confidence intervals.

Confidence intervals provide a way of quantifying sampling uncertainty. A 95% confidence interval means that, over many repeats of a survey under the same conditions, we would expect that the confidence interval would contain the true population value 95 times out of 100. The confidence intervals associated with all of the SHeS estimates calculated for this report are available in the supplementary tables.

Statistical significance

A result can be described as statistically significant if it is unlikely to have occurred by random variation. Testing for statistical significance allows us to assess whether there have been significant changes in estimates between two time periods or different groups (e.g. between two different age-groups, or different caring intensities).

For each question where a comparison is possible in the SHeS – for example, between two different survey periods or between men and women – a test statistic is calculated to test the statistical significance of the difference between the two results.

The test statistic is calculated as: 

p2 = the comparable second point estimate or proportion

where n1 = the number of valid responses for point estimate 1

n2 = the number of valid responses for point estimate 2

deff1 = the design effect for point estimate 1

deff2 = the design effect for point estimate 2.

For point estimate 1 not to be statistically significantly different from point estimate 2 with 95% confidence, then the test statistic must be between ±1.96.

All differences between SHeS estimates that are presented in this report are statistically significant at the 95% level, unless stated otherwise. Due to large sample sizes for some breakdowns, even small changes (of under 1 percentage point) may be statistically significant.

As Scotland’s Census 2022 produces population-level counts (compared to the weighted estimates from the SHeS), there are not associated confidence intervals readily available for the census data. As such, formal statistical significance testing on the differences between the census and SHeS data has not been undertaken. Instead, census figures have been compared to the confidence intervals of the SHeS data. Where the census figure is outside of the confidence intervals of the SHeS estimate, this is taken as indicative of a potentially notable difference. This approach is consistent with the methods implemented by National Records of Scotland (NRS) during data validation of the census. To find out more, please visit the Quality Assurance report produced by NRS.