New research on gambling survey estimates

A new study has been published exploring why different gambling surveys produce varying estimates of participation and Problem Gambling Severity Index (PGSI) rates. Led by Professor Patrick Sturgis, the research offers causal evidence on how survey design influences reported gambling behaviours and associated risks.

The study compares self-completion surveys, such as the Gambling Survey for Great Britain (GSGB), with traditional face-to-face surveys like the Health Survey for England (HSE) and the Adult Psychiatric Morbidity Survey (APMS). It investigates three key factors: the impact of mentioning gambling in survey invitations, the influence of interviewer presence, and the effect of updating the list of gambling activities.

Findings show that explicitly mentioning gambling in the survey invitation did not affect overall response rates but did increase reported gambling participation by four percentage points. This suggests that individuals with a personal interest in gambling were more likely to respond. The PGSI score for those scoring 1 or above was 1.8 percentage points higher in this group, although the difference was not statistically significant.

The presence of an interviewer had a more pronounced effect. Participants completing the survey online reported PGSI scores 4.4 percentage points higher than those interviewed by telephone. This nearly 50 per cent increase indicates that respondents may under-report undesirable behaviours when speaking to another person, aligning with broader research on social desirability bias.

Updating the list of gambling activities to reflect newer products, as done in the GSGB, did not significantly alter participation rates or PGSI scores. This suggests that the comprehensiveness of activity lists may not be a major driver of variation in survey outcomes.

Ben Haden, Director of Research and Policy, welcomed the findings, stating that the research enhances confidence in GSGB outputs and helps clarify differences between gambling surveys. He emphasised the importance of using multiple data sources to build a robust evidence base for policy and public guidance.

Professor Sturgis highlighted the value of the study’s experimental design, which allows for strong causal conclusions about the factors contributing to variability in gambling estimates. While no single survey can determine definitive values, the research contributes significantly to understanding how methodology shapes reported outcomes.

The study recommends that the Commission review its online guidance on interpreting GSGB estimates to better reflect the methodological factors influencing differences with earlier health surveys. This forms part of the Commission’s broader commitment to improving research development and public understanding of gambling-related harm.