Plenty of smoke but no fire

SBHD: National Science Week is upon us, comprising a range of initiatives to educate old and young alike but is it really in the interest of scientists to enlighten Joe Public?

This is National Science Week, and we are invited to marvel at the wonders of science and encourage our children to do the same.

The aim, of course, is to dispel a widespread and lamentable ignorance of matters scientific, and to inspire young scientists. While the latter objective might be fulfilled, the former will not. Most adults are implacably obtuse about even the simplest scientific facts, and in truth that’s the way the scientists like it.

If the objectives of National Science Week were to be fulfilled, there would be two casualties. First, the scientists themselves would suffer a blow to their authority, much of which at present derives from public ignorance. And secondly, the single-issue groups would lose the most effective weapon in their armoury, namely bogus science.

For it is the willingness of the public at large, and the media in particular, to swallow statistics unexamined that allows us to be hoodwinked and manipulated.

The smokers’ rights organisation, Forest, delights in exposing such statistical jiggery-pokery. For instance, it is frequently asserted that 300 smokers die every day, which is shocking until you put the figure in context.

In 1990, the daily death toll in Britain was 1,758. Since smokers comprise 30 per cent of the population, it is reasonable to assume that, if their habit causes their death, the 300 would account for 30 per cent of all deaths. But they don’t. Three hundred is only 17 per cent of 1,758. If smoking were an inevitable cause of fatality, based on the percentage of the population that smokes, some 527 smokers should die every day.

“Conversely,” says Forest mischievously, “if only 17 per cent of deaths a day are those of smokers, this means that 83 per cent are of non-smokers. But non-smokers are only about 70 per cent of the population. Does this mean it is healthier to smoke?”

Seldom are “scientific” statistics challenged in that way. For instance, a few days ago, a professor announced that a baby whose parents smoke is five times more likely to be a victim of cot death than one in a non-smoking home. This alarming statistic was based on a study of cot deaths which showed that slightly more than half the fatalities were in homes where parents smoked. Which means, of course, that, in almost half the cases examined, tragedy befell non-smokers. Why?

As far as I know, the professor stopped short of explaining the causal connection between parental smoking and sudden infant mortality. Perhaps there isn’t one. The cause of cot death remains a mystery. Just as mysterious is the process whereby a habit which it is politically modish to deride becomes the scientifically proven cause of just about every ailment to which man is heir. It is only historical bad timing that prevents tobacco being blamed for the Black Death.

Bogus science does not rely simply on bogus statistics, it also favours bogus presentation. The National Food Alliance, a preposterous body dedicated to determining the nation’s eating habits, published a manifesto called Children: Advertisers’ Dream, Nutrition Nightmare? It purported to prove that food advertising during children’s TV was wrecking the nation’s health.

Amusingly, the publication came fully attired in the garb of a scientific document with no fewer than 171 footnotes, a sure sign of dedicated scholarship. Or was it? The Advertising Association, knowing its own rebuttal would be dismissed as special pleading, commissioned two independent reviews.

The first, by Professor Patrick Barwise, of the London Business School, concluded that “the quality of the NFA’s analysis is so poor that its actual contribution to knowledge and understanding is very limited”. The second, by independent researcher Caroline Sharp, found that many of the references either did not support the argument or were taken from the lobby group’s own previous oeuvre.

To draw on one’s own authority in order to convey authority is a sure way of disappearing up one’s own random variable.

Another feature of bogus science is that it is immune to evidence that runs contrary to what it sets out to prove. Thus a publicly-funded survey in the US was suppressed by the American Lung Association because the results showed that 70 per cent of the population were “absolutely” opposed to a ban on smoking in restaurants.

Worse than the hiding of contrary data is the deliberate manipulation of published research. Dr John Luik, head of Canadian think-tank the Niagara Institute, condemns what he calls “corrupted science”. Once again, it is the scientists’ habitual addiction to anti-smoking that stunts their statistics.

Since it is necessary for the Government and the anti-tobacco lobby to demonstrate that passive smoking poses dangers to non-smokers, science is called upon to provide the proof.

“Only science can deliver dangers with the requisite pedigree,” says Dr Luik. “And with so much at stake, the pressure to adjust, shave, create, ignore, re-evaluate, even manipulate, is enormous.”

He goes on to demonstrate a few of the tricks used, including halving the “margin of confidence” and neglecting the “two-tailed test” which incorporates the opposite hypothesis to the one being tested.

“The fact that much of society believes tobacco to be dangerous,” says Dr Luik, “creates a strong incentive to establish, and indeed enlarge, the range of smoking-induced harms, while at the same time ignoring or suppressing research that questions these received orthodoxies.”

In this National Science Week, you won’t hear much about “corrupted science, science that has been politically laundered, science that because of its corrupted status actually ceases to be science”. More’s the pity.