Fresh experimental research led by Professor Patrick Sturgis has shed light on why gambling surveys see varying levels of participation and generate differing estimates of problem gambling. 

The news arrives just weeks after the Gambling Commission (UKGC) pledged to strengthen confidence in the outputs of its new Gambling Survey for Great Britain (GSGB) – a study that looks to determine the extent of gambling harm in the UK. 

The study’s main output is determining how many Britons can be considered suffering from problem gambling, based on the Problem Gambling Severity Index (PGSI), which their responses to survey questions are assessed against.

The GSGB, launched this summer, has already drawn comparisons with the Health Survey for England (HSE) and Adult Psychiatric Morbidity Survey (APMS), which have both been used to inform policy decisions around gambling harm in the past.

The new research shows that design choices in surveys – from how invitations are worded to whether questions are asked by an interviewer – play a major role in shaping the results.

Why survey mode matters

The study found that explicitly mentioning gambling in an invitation did not affect how many people responded overall, but did attract more people with a personal interest in gambling, resulting in a four-point increase in reported participation. 

Meanwhile, when it came to PGSI scores, this effect was smaller at 1.8 points and not statistically significant.

A bigger difference appeared when interview style was tested. Participants completing surveys online were almost 50% more likely to score one or above on the PGSI compared to those answering questions over the phone, underlining how people tend to under-report sensitive behaviour when speaking to an interviewer. 

Updating the list of gambling products to reflect new market entries, however, had little to no impact on results.

Ben Haden, Director of Research and Policy at the UKGC, said: “The research builds our confidence in the outputs of GSGB, helps to understand the differences between surveys published on gambling and will improve our guidance for users.” 

He stressed that no single study can ever definitively measure participation or harm, but that the regulator will continue to refine the GSGB, expand its use of datasets, and work with other survey providers to build a rounded evidence base.

Next steps for Commission guidance

Professor Sturgis emphasised the experimental design means “strong causal conclusions” can be drawn about the reasons for wide variability across gambling surveys, while noting that no single piece of research can provide the “true” value of participation or harm rates. 

The Commission has been advised to update its guidance on interpreting GSGB results to better explain the differences with health surveys such as the HSE and APMS.

The regulator has confirmed that this update will be part of its ongoing improvement programme as it looks ahead to the second annual GSGB report, due for release on 2 October 2025. 

With gambling harms still central to the White Paper reforms and under scrutiny in Westminster, the Commission has said it is keen to ensure the GSGB provides the most reliable picture yet of British gambling behaviours.

This data could prove critical as the debate around gambling regulation and the industry’s societal impact remains as heated as ever. Although the government seems committed to seeing out the recommendations of the Gambling Act review, it is facing calls for another regulatory overhaul from a large group of backbench MPs and local governments.