‘Statistical Proof’

These are musings – I have not yet fully recovered from the excesses of my holiday.

Chris Snowdon has drawn attention to how the statistics were fiddled to justify the new recommended alcohol limits. He shows how it was done in The Spectator here:


Very briefly, the ‘statistical model’, which Sheffield Uni had invented, showed that a male person, on average, could drink about 21 units of alcohol per week perfectly safely. The equivalent for a female person was 14 units. But there was more to it than ‘perfectly safely’. Such quantities could actually be beneficial for the functioning of the heart especially. But even if you disregard the beneficial effects, at least you can say that such quantities are harmless. Thus, the ‘Sheffield Model’ started to rate harmfulness at more than 21 units for men and more than 14 units for women. Those numbers were ‘the base’ – equivalent to zero on a graph. Thus, any harm from alcohol only started to accumulate ABOVE those base figures. Thus, roughly, 21 units is zero harm; 22 units produces a tiny bit of potential harm, which we can call ‘1 unit’ of harm.

Public Health England told Sheffield to get rid of the ‘harmless base’ from their model. Obviously, that demand was made in a roundabout sort of way. Sort of: “Do us a favour and amend your model, on your computer, to see what happens if the harmless base allowances of 21 units for men and 14 units for women were reduced to zero”. Under great pressure from their financial benefactor, Public Health England, and with great misgivings, they did as they were told. For some reason, the ‘adjustments’ to the model did not make much difference to the fate of women, but brought down the values for ‘safe’ limits for men.

But what matters is the that the ‘science’ was compromised, at the behest of PHE. Unrealistic parameters were introduced so that the ‘beneficial/safe’ level of drinking was removed from the calculations.

Sheffield Uni was compromised. Why did it comply? 1) There was money involved – several thousand pounds; 2) It needed not to be ‘struck off’.

Thus, it provided the ‘evidence’, very unwillingly, for Silly Sally to declare that a 7 stone woman can drink as much alcohol as a 15 stone man without danger.

If Silly Sally was asked to produce proof, she would cite Sheffield’s results.

We might reasonably ask how many other ‘statistical proofs’ have been engineered in much the same way? To what extent did DOLL ET AL fiddle the stats in the Doctors Study?

You see, the fiddling with the Sheffield model probably only involved a few hundred real people out of thousands.

I really must go through Doll’s Doctors Study again. My reason is to find out how many subdivisions of smokers there were and what were the actual number of doctors involved in the subdivisions. By subdivision, 30,000 can rapidly become 3000, and can then rapidly become 300. It is when a very large number, which is statistically representative, is divided into smaller groups that confounders become very important.

But the opposite is also true. Untested results, based upon fiddled stats as illustrated above as regards ‘base/safe’ starting points for alcohol danger, can envelope the whole world.

What can be done? None of the academics have committed a crime. All that they have done is told lies. It is not a criminal offence to tell a lie.

But such people have a problem – their lies can be countered by the lies of others. There is no truth.

The sadness of the fiddling of statistics as regards recommended drinking limits is that the new limits can be used to justify things like minimum pricing. It would be easy to show that people are ignoring the limits, which might be used to justify the persecution of the poorer segment of the people.

Somehow or other, our politicians must be made aware that smokers are not as pathetically subservient as they think.

But perhaps they are? Disregarding the Smoking Ban, how many smokers talk about the cost of cigs? Taxes on cigs should be more important than bans since those taxes are clearly, and without doubt, discriminatory.

Governments are lazy. Politicians are lazy. The Civil Service is lazy. Parliament is lazy. They need to be woken up.




2 Responses to “‘Statistical Proof’”

  1. Timothy Goodacre Says:

    Excellent points Junican ! I enjoy my 21 plus units a week.

Comments are closed.

%d bloggers like this: