How ASH turned the views of just SIX ex-smokers into a headline-grabbing 300k
Tuesday, May 5, 2020 at 18:20
Simon Clark

Good news ... and bad news.

The good news is that earlier today I received the data tables for the ASH/YouGov survey that was published yesterday.

The poll generated quite a lot of headlines including 'More than 300,000 UK smokers may have quit owing to Covid-19 fears' (Guardian), 'More than 300,000 Britons quit smoking over Covid-19 fears' (ITV News), 'Surge in motivation' sees 300,000 smokers kick the habit (Yorkshire Post) and 'More than 300,000 Brits quit smoking amid coronavirus fears, study shows' (The Sun).

According to the Guardian:

The survey of 1,004 people suggested 2% of smokers had quit because of concerns about coronavirus; 8% were trying to quit; 36% had cut down; and 27% were now more likely to quit.

As I explained yesterday I was struck by the fact that, if this was a standard poll, the 1,004 respondents would be a representative sample of the general public.

Approximately 15 per cent of adults in the UK currently smoke so the odds were that the number of respondents who smoke would have been around 150.

That's a pretty small sample with a pretty large margin for error. Having read the data, however, the number of smokers polled was even less than that.

Of the 1,004 people polled, 56% (562) were never smokers; 31% (307) used to smoke but have given up; 9% (90) smoke every day; and 4% (40) smoke but not every day.

In other words, only 13% of the sample (130) were current smokers, significantly less than the national smoking rate.

The most important group however were the 310 people (or 307 according to the data) who used to smoke because it's from their responses that ASH extrapolated the headline figure of 300,000 smokers who have allegedly been driven to quit 'over Covid-19 fears'.

I'm no mathematician and I've been struggling to get my head round the figures, but with a bit of outside help this is how I think they did it.

First, let's round up the figures so the sample size is 1,000 people of whom 130 were smokers, and 310 ex-smokers.

Two per cent of the 310 ex-smokers cited concerns about Covid-19 being the reason they quit smoking in the last four months.

That's six people, or 0.6% of the 1,000 people polled.

The adult population of the UK is approximately 50 million and 0.6% of 50 million is ... 300,000.

The bad news therefore is that the headline-grabbing figure of 300k appears to be based on the responses of just SIX (6) ex-smokers who have quit in the last four months.

This mathematical exercise also seems to result in the claim that 550,000 smokers 'have tried to quit'.

According to the data, 8% of the 130 smokers polled said they had 'tried to quit smoking'.

8% of 130 is 10 (approximately). Therefore, ten out of the 1000 people polled said they had attempted to quit smoking.

Ten out of 1000 is one per cent. One percent of 50 million is ... 500,000.

OK, it's not 550,000 but it's in the same ballpark. I'm guessing that ASH used a different number for the adult population of the UK but I haven't got the patience to work out what it was so our figures differ slightly.

The point is, ASH/YouGov polled a relatively small number of smokers and extrapolated the most extraordinary figures.

According to ASH, 300,000 smokers have quit in the past four months, worried that smoking makes them more vulnerable to Covid-19, while an additional 550,000 have attempted to quit.

To remind you, these estimates are based on the responses of SIX and TEN people respectively.

Someone may tell me I've got this all wrong so do check the data tables for yourself and post your own analysis.

Or perhaps someone from ASH would like to explain how they arrived at their interpretation of the figures.

(Stop laughing at the back!)

Update: Chris Snowdon writes, 'Have 300,000 smokers quit because of Covid-19?'

Article originally appeared on Simon Clark (http://taking-liberties.squarespace.com/).
See website for complete article licensing information.