How ASH turned the views of just SIX ex-smokers into a headline-grabbing 300k
Tuesday, May 5, 2020 at 18:20 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?'
Reader Comments (4)
Well done. Simon.
This ASH poll was only 831 people over 16 in a telephone poll, goodness knows who they were, but they were used to get us banned from every enclosed space open to the public and from some that weren't.
New Poll Shows Public Back Health Select Committee Amendment on Smokefree Law
31st January 2006
“An all-Party group of ten members of the Select Committee, led by the Chair of the Committee Kevin Barron MP, has tabled an amendment to the Health Bill that would ensure that all workplaces and enclosed public places, including all pubs and membership clubs, would be smokefree. The Bill as it now stands would allow smoking to continue in clubs and in pubs that do not serve prepared food.
The Government has now announced a free vote for its MPs on the issue, as has the Conservative Opposition. The Commons Report Stage of the Health Bill may take place in the week beginning February 13th.
Asked whether they support the Committee’s proposal 70% of those polled said yes, with only 18% saying they were opposed”
[1] The survey was conducted by BMRB International using the BMRB Access Omnibus (telephone) survey between 20-22 January 2006. It involved 831 adults aged 16+ in England.
The survey asked:
“A Government Bill in parliament will make enclosed workplaces smokefree, with exemptions for pubs not serving food, and private members clubs. The Chair of the Health Committee supported by committee members from all parties has tabled an amendment to make ALL enclosed workplaces, including all pubs and clubs smokefree.
How strongly would you support or oppose this amendment to make ALL enclosed workplaces smokefree?”
http://www.ash.org.uk/media-room/press-releases/new-poll-shows-public-back-health-select-committee-amendment-on-smokefree-law
A confidence trick indeed.
1 smoker a month did quit because of the ASH fearmongering about Covid19... hurray
It's even simpler than that. They rounded the figure up to 300, and then just added all the 000s.
To make the lie believable, ASH has to make it a big one.
Why are we forced to fund this hate group that manipulates figures to incite hatred, exclusion, and abuse of egitimate consumers who are currently the only best paying any decent amount of tax during lockdown to keep the country going.
What would the margin for error be with the published numbers? For a newspaper to publish the numbers without reference to the margin for error seems disingenuous.