Number of the Month

April   2005

Poisson d'Avril

Doesn’t time fly when you are enjoying yourself? Believe it or not, this is the sixth number of the month for April. And, believe it or not, the creator of the first ever number of the month is still hitting the headlines. Back in April 2000 he was Dr Richard Wiseman, but by the time he was providing numbers for April 2004, he became Professor Wiseman. Another April and another Wiseman event, which managed to bag two tabloid pages of The Times. This time he hit the British media with a study of smiles. Impeccable timing too, for fixed grins were in the news, with three princes uncomfortably facing the media pack and Rictus Man himself about to declare a general election. You can even take part in the research yourself, helping to create a source of data at luckfactor.

A scientific breakthrough in these days of the Einstein centenary requires synchronisation with a popular TV series. So it was when a new piece of “research” hit the headlines. Grumpy old men are a myth. It's the women who rage claimed The Times. I don’t believe it… old men are less grumpy said the Telegraph, managing to get in a reference to yet another TV series by quotation:

“Victor Meldrew was the exception and not the rule,” Jane Barnett, of Middlesex University, said in reference to Richard Wilson’s permanently cross character in the hit television show One Foot in the Grave.

She added that the Grumpy Old Women television programme, featuring people such as Janet Street-Porter and Germaine Greer ranting at a succession of irritants, was far better at reflecting reality than its counterpart Grumpy Old Men.

Miss Barnett, presenting a study of anger and the sexes to a British Psychological Society conference, said that she had looked at how levels of fury changed over the years.

We are given some details of the “research”. 153 people were interviewed. They were divided into two sexes, four age ranges and four anecdotal levels of rage; so the 153 people were put into 32 boxes, which on average is fewer than five per box. Using the Poisson approximation, the random variation on this number would be about 50%. Furthermore, the ratio of two such numbers would have a variation of about 100%. However, that is where they have got us, because no numerical results are presented. Never mind, at least we have the comfort of familiarity:

Miss Barnett told the conference in Manchester yesterday that more research was needed into why men calm down while women “remain simmering” through the ages.

It is quite extraordinary what passes for science these days, especially during the popularising festivals, while the British Psychological Society has always been good for a laugh. The Associations for Advancement of Science (British and American versions) give house room to the most bizarre unscientific theories, such as Lysenkoist  explanations for the development of man’s great brain (apparently it is now caused by eating meat and not tubers as reported in Sorry, wrong number!).

The stories above were both published on April Fool’s Day. The genuine traditional hoax took the form of a full page advertisement from BMW, claiming to have eliminated the steering wheel. Trouble is, the hoax is more believable than the “science”.

6/04/05

Step aside Einstein

They think it’s all over. It is now!
Kenneth Wostenholme, football commentator 

It must be an important scientific result, for it not only commanded half a broadsheet page in the Daily Telegraph, but also an editorial leader. The result of the experiment was recorded in the latter as “The men turned out to be much better at spotting the fakes.” How much better? Well, the score was men 72%, women 71%.

What sort of difference would we expect at random? Using the formula (shown last month) for the expected percentage variation, we can insert the values N=7500 and p=0.705 (the median value). As a result we get an expected range of variation of ±0.53%. Lo and behold, the published difference of 1% is within that range!

Yes, it is the number watcher’s favorite professor, Richard Wiseman, who gave us the usual April opening above. Naturally, he is able to make far reaching pronouncements on the human condition on the basis of these decisive numbers, especially after a bit of astute cherry picking based on the extreme value fallacy.

Expressed as a relative risk, the ratio 72/71 gives us a value of 1.014.  This will come as a great disappointment to number watcher Nicholas Hallam, who was, even at this early stage, considered a dead cert to win the “Spot the lowest RR of the year” competition with this paragraph:

"Each hour of television viewed per day at age 4 years
was associated with a significant odds ratio of 1.06
(95% confidence interval, 1.02-1.11) for subsequent
bullying."

which led the BBC to announce portentously that TV could create child bullies.

No one to die

It is one of the sad facts of human existence that when one champion arises another takes the back seat. Nigel (thousands to die) Hawkes, who has reigned for so long, has lost his golden touch. His latest headline Frequent use of mobiles is 'not linked to cancer' is hardly likely to scare the pants off anybody. Apart from the fact that number watchers have known this for years, it illustrates yet another law:

Law of epidemiology:

For every result in epidemiology there is another equal and opposite result.

Corollary:

The scary results are far more likely to get published.

Hawkes' other item of the day also illustrates one of our laws, that of beneficial developments:

Warning as doctors turn backs on arthritis drug

It’s a funny old world.

12/04/05

Pushing the pills

‘In any investigation, my Bunter, it is most damnably dangerous to have a theory.’
‘I have heard you say so, my lord.’
‘Confound you – you know it as well as I do. What is wrong with the doctor's theories, Bunter?’
‘You wish me to reply, my lord, that he only sees the facts which fit into the theory.’
‘Thought-reader!’ exclaimed Lord Peter bitterly.
Dorothy L Sayers, The footsteps that ran

In our forum Ian Reid draws attention to the new outbreak of aspirin mania in sections of the medical profession. They seem to get an idea into their heads and nothing will shake it out again. This one is the obverse of the Salt scare. They ignore all contrary evidence and present underwhelming evidence of their own, as if it were kosher science. Apart from the thrill of seeing their names in the papers, why do they bother? Perhaps it is the same kick that religious leaders get when legions of the faithful conform to their diktats. There is material for several PhD theses in psychology here, except that, as we have seen above, academic psychologists are too busy in their relentless pursuit of trivia.

14/04/05

 

People and computers

It is touching, astonishing, and rather frightening that people have so much faith in computers. Here is a piece from The Times:

Forecast is streets ahead

WEATHER forecasters will soon try to predict precisely where rain will fall and at what time, using a powerful supercomputer. 

Predictions will become so localised that within the next five years an “in-your-street” service will be available for mobile-phone users giving details of the weather to expect within a mile of their location. 

Steve Noyes, the chief technology officer at the Met Office in Exeter, said that the launch of the £27 million NEC SX-8 supercomputer represented a significant step towards far more localised weather predictions. It will be invaluable for the military, which needs to plot operations based on predicted weather patterns….. 

Oh yeah?

It used to be said that when two Englishmen gathered together the topic of conversation would be the weather. Now it is more likely to be the inaccuracy of the weather forecasts.

If your computer program reveals that two and two make five, doing in one microsecond rather than one millisecond adds nothing to the quality of the result.

The parties involved in that election also suffer from the same delusion. The Conservatives have plumped for a system called Voter Vault, which they obtained from the Republicans. Bush won, so it stands to reason that the program is infallible, doesn’t it. Using data from a variety of sources, it puts people into numbered categories, the names of which are straight out of the Marketspeak wordbook – Happy Families, Take-A-Break Couples, Urban Intelligence, Caravanners etc.

The truth is that electioneering is an area where there is no substitute for human intelligence (in both meanings of the term). One of the consequences of the huge majorities of the ’80s is that the arrogant and complacent Government allowed the party in the country to fall into disrepair through sheer neglect. When you do real electioneering, the location of your support is always a source of surprise – council estates contain true blue Tories and millionaire’s pads house life-long Labourites.

A perfect scam

Just above the weather forecast story in The Times is another story headlined It might not have felt like it but spring was early. The Telegraph has Early spring is a fact of life for the birds and bees.

It started with a bunch of global warming religious zealots calling themselves The Woodland Trust, but now it is the UK Phenology Network. By taking selectivity of evidence to an extreme and getting the ordinary punters involved, they have created an ideal propaganda tool. Blithely ignoring the fact that most of the Northern Hemisphere has experienced a sequence of harsh and fatal winters and seizing upon the fact that Britain has experienced a weather pattern that has blocked the cold north-easterlies that are a normal part of winter, they produce mounting evidence of the reality of Global Warming.

Primarily the scheme exploits the extreme value fallacy. If there is an early snowdrop and you have only one observer it is almost certain to be missed; have a million well-distributed and it is almost certain to be seen. The apparent earliness will tend to vary with the logarithm of the number of observers.

Then there is the trick of ratchet reporting. For example, if you only mention cases of species diffusing northward, such as the inevitable premature swallow, and ignore those coming in the other direction, then you achieve the required distortion of reality.

Waxwings, as we mentioned in December, invaded Britain early this winter. They do this every few years, when the Rowan berry crop fails due to especially harsh weather in Russia and Finland. There was a particularly massive invasion in 1965-6. But they were not on the list, so they do not count as “Phenology”.

The Telegraph tells us:

A record number of unseasonal events has been submitted by members of the public for a database that monitors the times of annual events in the natural world.

What an odd coincidence, when we also have a record number of observers!

Granny not

When you get the zeal of the anti-tobacco propagandist combined with the arrogance and insouciance of the epidemiologist you can expect something special in the way of junk. One of the dilemmas of the propagandists is that the rate of occurrence of asthma has risen as the exposure to tobacco smoke has fallen dramatically. All that was needed for a solution was a bit of tawdry statistics accompanied by a bit of Lysenkoist  theorizing, demolishing the accepted biology of  reproduction and genetics on the way. Then you have a story that was seized upon by the media editors.

The Trojan Number was 338 and the relative risk 1.8, leading to the cast iron conclusion that granny’s smoking causes childhood asthma. Apart from the fact that it is total nonsense, as you would expect to emerge from the Dark State of Insanity, it trivializes the very serious matter that over a third of the sample contracted asthma before the age of five. When we oldies were children and had more exposure to tobacco smoke in one visit to the cinema than Californian children get in a lifetime, asthma was virtually unknown, apart from a few virtual invalids of the kind beloved by Victorian novelists. What have epidemiologists got against other factors, such as central heating and fitted carpets?

What is the bloody point?

As suggested last month we have had daily opinion polls, with learned professors explaining variations that are within the expected random range, but it is worse than that. Tories close the gap to one point yells the Sunday Times. Blair heads for third landslide with ten point lead screams the Sunday Telegraph. There was a time when two groups of people measuring the same phenomenon would be rather disturbed when their estimates of a difference were discrepant by a factor of ten.

Let’s go a-Maying

Look, look: a May-mess, like on orchard boughs!
Gerard Manley Hopkins,
The Starlight Night

Talking of factors of ten, Tim Curtin has been listening to a lecture given in Australia by our embarrassingly innumerate President of the Royal Society. Apparently he gave the same talk in Sweden. Among other gaffs his Lordship equates ten percent to a factor of ten. Tim suggests another Number Watch annotation exercise. Only too pleased if someone can provide a copy of the original.

17/04/05

Lettuce prey

You could not have a clearer example of  a health guru getting carried away with himself than the attack mounted by Dr Thomas Stuttaford in The Times (March 31st). The headline was Lettuce: green, mean and dirty. He goes on:

A portion of lettuce contains only a twentieth of the fibre in a helping of broad beans, and a fraction of that found in peas or parsnips. The amount of vitamins that lettuces contain is trivial. There is a little bit of beta carotene in the darkest lettuce leaves, and a trace of folate, but compared with that found in peppers, tomatoes, spinach, red cabbage or baked beans, it is insignificant. Lettuce is an expensive way to buy water. 

It so happens that your bending author is currently growing eight varieties of lettuce, together with a dozen other salad plants, including three varieties of the holy tomato, to which Stutterford pays due obeisance as the fount of all good, as laid down by the Priests of the Church of Harvard (see the books of the web site). The lettuce plants cost about a penny to grow. They are among the things that make one so look forward to summer. They provide more than just a background to the wonderful range of flavours, colours and textures of a well prepared and dressed summer salad. 

Stuttaford blames lettuce for food poisoning, because some restaurateurs fail to wash it properly. That is like putting the blame for all road accidents on the vehicles. To prepare a lettuce correctly, you wash each leaf individually under running water and then dry them thoroughly in a kitchen centrifuge. This ensures that your dressing, made of course with home-grown garlic, adheres properly.

What a desolate world these gurus live in! They care not for colour or flavour or texture. As long as their little grey pills have all the right nutrients (and, of course, the sainted fibre) they can live happily in their little grey cells.

The fat in the fire , again!

The e-mails among number watchers were buzzing over the unique U-turn by the CDC over its obesity claims. Junkscience.com quoted several different commentaries. It certainly was an astonishing event. This organisation has sailed on, regardless of the fact that its various scary claims have been thoroughly debunked, promoting its campaigns to dragoon the population into obeying the new ten commandments, so they can receive their just reward at life’s end. The previous claim of 400,000 deaths due to obesity has been drastically watered down 26,000. The latter number is almost certainly also an exaggeration, based on weak epidemiological evidence, but an admission from such a source of exaggerating by a factor of more than fifteen is truly remarkable. Even more amazing, this admission is also reported:

… for individuals aged 25-59 the risks of premature death from being underweight are substantially greater than those of being overweight and they are also slightly greater than those of being obese. For those aged 60-69 the risk of dying from being underweight is much higher than from being even significantly obese, that is with a BMI > 35. Again, the total number of premature deaths due to obesity is 25, 814, while the mortality attributable to being underweight is 37, 746. If anything this points to an epidemic of not fat but thin caused death.

If you think this is something new, here is a sentence from Sorry, wrong number!, which went to press almost exactly 5 years ago:

One of the things I have learned from my researches for this book is that a certain amount of obesity, particularly in older people, is not a health danger, despite all the propaganda.

So what could possibly prompt an outbreak of near honesty from a propaganda organisation that has shown little such inclination since it lost most of its main function of fighting infection? Could it be that its members suddenly realised that they were going to run out of corpses? Their claims for tobacco and obesity alone already accounted for more than a third of  USA deaths. 

Is it likely that there will be a similar correction to the lies about tobacco? Not a chance! That campaign is driven by people who live in terror of their own cravings; so, even though the evidence shows that smokers do not on average die early, it will be actively suppressed, and the campaign continued with vehemence. It is no coincidence that the two biggest frauds in the history of statistics involved tobacco.

And what of the huge industry that has been built up on the back of the CDC obesity scam? You can bet your sweet bippy that it will not fade quietly away.

25/04/05

Junk mail

Your bending author and spouse have just deregistered from postal voting, not wishing to be involved in something clearly designed to encourage electoral fraud. Of all the humiliations heaped upon a benighted nation by The Great Leader this must be one of the worst, with observers from the likes of Ukraine and Serbia casting a critical eye over the elections to the Mother of Parliaments. It is just another example of the fundamentally corrupt approach of New Labour to just about everything. The voters don't like it, yet they inexplicably continue to favour this bunch of crooks. The old style socialists might have been objectionable, but at least there was a certain straightforward honesty about them.

Of course, New Labour BBC would have to be involved as well.

Should there be any remaining doubt , your bending author will be casting his vote in favour of the candidate most likely to beat New Labour.

Swedish rhapsody

rhapsody, an orderless, unconnected composition or collection of things said, beliefs etc. (obs.):
Chambers Dictionary

The Karolinska Institute is still set on becoming the Scandinavian Harvard. Our Man in Puerto Rico keeps an eye on its activities (among many others that contribute to his regular offerings of the Junk du Jour). Your bending author is not so indefatigable and is becoming wearied of the rôle of Canute.

OMIPR's latest take 

Here is the latest from those Karolinskians. I think they are losing their magick touch:

"... no evidence that the hardness of the water or its magnesium or calcium content has any effect on the risk of MI".
"An inter-quartile range increase in the daily intake of drinking water magnesium was associated with an adjusted odds ratio for myocardial infarction of 1.01 (95% CI 0.87-1.17)."


But they regained their true form, rapidly shrinking away from this outbreak of common sense:

1.
Exposure to energy-averaged aircraft noise above 55 dBA was associated with an adjusted odds ratio for hypertension of 1.59 (95% CI 1.00-2.53), and maximum aircraft noise above 72 dBA was associated with an adjusted odds ratio of 1.76 (95% Cl 1. 12-2.77)

2. spousal exposure of 20 cigarettes per day or more was associated with an adjusted odds ratio of 1.58 (95% CI 0.97-2.56), and cumulative combined exposure at home and work was associated with an adjusted odds ratio of 1.55 (95% CI 1.02-2.34) in the top decile of exposure.

3. A difference in the 30-year average exposure to traffic-generated N02 of 30 µg/M3 was associated with an adjusted odds ratio for fatal myocardial infarction of 1.51 (95% CI 0.96-2.16) and 2.17 (95 CI 1.05-4.5 1) for out-of hospital death.

Time to thin out some lettuce seedlings.

26/04/05

Lest we forget!

Earth Day 1975

 

How refreshing!

Sp!ked has now published the results of its survey to commemorate the Einstein centenary. 250 scientists have each stated the one thing they would like to teach the world. It is all inspiring stuff and there is barely a handful of eco-theologians among them. Despite the impression given by the media and politicians, real science is still alive and kicking.

27/04/05

Number of the month 1.008

For all knowledge and wonder (which is the seed of knowledge) is an impression of pleasure in itself.
Francis Bacon

Readers of the Number Watch Forum will have noted that relative risks are still heading downwards. This coincides with the Swedish bid to replace Harvard as the home of junk science (see Swedish rhapsody above).

This one made the Daily Telegraph, with the headline Toxic chemical exposure 'could affect the sex of your children. This time it was from the University of Malmö and, believe it or not, offered a relative risk of 1.008 from a Trojan number of 149 seamen divided into quintiles, i.e. about thirty in each group. The number of spermatozoa examined in each case was not specified. We learn from the press release that (emphasis added):

The researchers found that larger amounts of both CB-153 and DDE in the blood of the fishermen was associated with a statistically significant increase in the proportion of Y chromosome bearing sperm in semen, and that age, smoking and hormone levels had no effect.

Is it something to do with that strange and ineffable sense of humour of that dour nation? Are they pulling our leg (or any other member for that matter)? What counts as a statistically insignificant result in that part of the world?

Sperm is certainly in vogue this year. The Italians blame traffic for a decline in quality, no numbers given.

30/04/05

Index

moncler outlet usa Moncler outlet hermes outlet prada outlet gucci outlet dior outlet lv outlet chloe outlet