Tuesday 28 November 2023

Gambling Commission fiddling the problem gambling statistics

The Guardian's resident anti-gambling correspondent got excited about figures which suggested that there are eight times as many problem gamblers than previously believed last week. I had a look at this stat for The Critic. It seems that the Gambling Commission is usually a notoriously weak methodology to inflate the stats.

There are several reasons why the new survey inflates the statistics, but the main issue is selection bias. All surveys try to get a representative sample of the population, but you can’t force someone to participate. If there are systematic biases behind people’s reasons for participating or not participating, the data will be skewed.

And indeed there are. Firstly, online surveys appeal to people who are very online — and that includes a lot of problem gamblers. Older people, who are less likely to be problem gamblers, are under-represented. The Gambling Commission has acknowledged that an “online methodology means that the sample responding to the survey are more likely to be engaged online, thus skewing the data”. 

Secondly, people who gamble a lot are attracted to surveys about gambling. The Health Survey for England asks about a range of health issues, but the new Gambling Survey for Great Britain is just about gambling. The clue is right there in the name and a study published in 2009 found that “gamblers and problem gamblers are intrinsically more interested in “gambling” surveys and therefore participate at a much higher rate than nongamblers”. If a disproportionate number of problem gamblers take the survey, the survey will naturally identify a disproportionately high number of problem gamblers.

The Health Survey for England has a response rate of more than 50 per cent, but the response rate for the Gambling Survey for Great Britain is only around 20 per cent. Four out of five people simply refuse to take the new survey, leaving a relatively small group of self-selecting individuals who differ from the general population in various ways, not least in being more likely to have gambling problems.


No comments: