BAD BAD BAD reporting of stats: 70% of British drivers high on drugs
According to this news headline reported on NDTV.com:
Over 70% of British drivers [are] high on drugs.
Really? I mean really? Perhaps I should check that in the finer detail from the article.
A whopping 72 per cent of drivers in Britain admit being under the influence of narcotics while driving, according to a new study.
Yes, I read it right. Almost three quarters of all drivers on the roads in Britain say they are high on drugs.
No no no no no. I hope not at least, that seems a little dangerous. You’d think the police would do something about that. You’d think we might have heard about it before.
This is a classic example of bad reporting of research, by someone who clearly has no grasp of statistics.
In fact according to the Press Association:
7% of the UK’s 49.1 million adults admitted to [ever] driving having taken drugs.
That I can believe. But 72% (at any given time) compared to 7% (ever) is a massive difference. That’s quite a misrepresentation of the figures. It looks like NDTV.com just read the information wrong and then never questioned their ridiculous interpretation when ripping it from a brief article in the Daily Express. But what can you do, it is out there now.
Oh dear oh dear oh dear.
This press release first came to my attention when I read an article in the Metro that said that 35% of British drivers were high on cannabis. It didn’t look right to me, and upon further investigation it was clear that they’d misunderstood the data. In fact 35% of British drivers who had driven while under the influence of drugs were high on cannabis. That is 35% of 7% = 2.45% of all drivers had ever driven under the influence of cannabis. Not the same thing at all. Funnily enough that article does not seem to be available online. I hope that means someone pointed out their mistake. And as it turns out, their interpretation was not as ridiculous as NDTV.com so NDTV became the focus of my blog post instead!
Statistics are not for everyone, I’ll grant you that, and I do understand that percentages of percentages can be a bit tricksy. Even so, common sense should be applied. I’ve said it before but someone wise once told me “If it looks wrong, its probably wrong.”
I recently wrote a blog entitled why you can trust (some of) the research stories in the media. And I still believe that, because the actual survey relating to drug driving appears to stand up to scrutiny. It is the reporting that doesn’t. Perhaps I should have called my blog why you can trust some of the articles you read in the press that are not – at face value – complete nonsense.