Six Recommendations Are Met on Britain’s Gambling Survey. What Still Needs Work?
Six of nine recommendations have been met for the Gambling Survey for Great Britain. The Office for Statistics Regulation’s November 19 follow-up review also adds a new one. That leaves readers with a more useful question than whether the survey passed a review: what can you reasonably compare when you use a figure from it?

The Gambling Commission’s response says the survey collects information from around 20,000 respondents a year. These are estimates from what people say about their gambling and its effects. The sample size describes the reach of the survey. It cannot tell you that respondents remembered every transaction or that a result should match records held by an operator.
Where to look before comparing two figures
Start with the release itself and read how the measure was collected. OSR says the Commission has published more about validation and quality assurance, plus useful guidance on when the data can be used. The review still wants the potential biases and their effects explained more clearly inside the statistical release. A headline figure can travel much further than its supporting notes.
Then check what the comparison is trying to show. OSR points to work comparing patterns in the gambling survey with the Health Survey for England, along with work under way on differences from industry data. It asks for further investigation, including comparisons involving the Adult Psychiatric Morbidity Survey. When two sources use different samples or ask different questions, a gap between them alone does not establish that participation has changed.
That distinction also mattered in our coverage of Britain’s youth gambling survey. The measure of age-restricted play and the broader measure of gambling answer separate questions. Check the population, dates and definition before putting those results beside an adult survey estimate. Neither a survey participation rate nor an estimate of harm describes stakes, revenue or profit.
The review concerns statistical practice. There is no finding here about a company’s compliance with a licence. The practical changes readers should notice are in the context around the estimates, especially where a rate is likely to be picked up and quoted on its own.
The next date to watch
OSR still wants clearer reporting of bias, more work on comparability, and better feedback to people challenging the statistics. Its new recommendation asks the Commission to explain how it interpreted the Sturgis research and how that interpretation informed its decisions. The Commission expects to give a public progress update by February 2027. For readers, look for changes to the release notes and the comparison guidance alongside the next set of figures.
