Scientific Fraud

It is nearly thirty years ago that I wrote a piece for the Christmas edition of the British Medical Journal, entitled Fake! In it I detailed some well known fakes, frauds and mistakes made by scientists and doctors. Among them were the misconception that spinach was a good source of iron (a German chemist put the decimal point in the wrong place), the Piltdown fraud (and there are many others concerning ape-men) and Gregor Mendel's statistics (they were too perfect - he must have fiddled them).

At that time there had been a rash of scientific frauds, particularly from the North Eastern states of the USA, that had hit the headlines. The whole paper is available on line so I won't list them. The most remarkable concerned a Jordanian (or perhaps Iraqi) research fellow, called Elias A K Alsabti, who had plagiarized 63 papers. He made use of the fact that many scientific articles are hardly ever read. (I have to confess that of more than 700 articles that I have written, more than 100 have never been cited - though many of them like book reviews and polemical pieces were not written to be cited.) Alsabti used to raid his boss's filing cabinet at night and photocopy articles that had been sent for review. A new title and new authors, and it was apparently a new article, ready to be sent to an obscure journal. The amazing thing was that he went on for so long without being caught.

But the fraud that I want to draw attention to was that committed by Sir Cyril Burt. Burt was the doyen of educational psychologists. He was editor of a major psychology journal and an advisor to the British government. He believed that intelligence was inherited and set about to prove it. He reported on pairs of identical twins separated at birth by being adopted by families of different social class. He tested their IQs and demonstrated that they were virtually identical despite one being brought up by, say, a factory worker and the other by, say, an academic.

His conclusion encouraged the government to design an examination at the age of 11 which would segregate children into those who would benefit from an academic education and those who would not. Since intelligence was genetically determined there would be no need for multiple tests at different ages. In America his work was cited by those who believed that intelligence was racially determined.

Years later, a psychologist called Leo Kamin, a man of the Left who had different prejudices, investigated Burt's work. What he found was astonishing. At different stages Burt had reported on 21, more than 30, and 53 pairs of identical twins raised apart, but in each report the correlation factor was precisely 0.771, while the correlation for the twins raised together, no matter how many, was always 0.944. This is clearly an impossibility. A little later, in 1976, Oliver Gillie, a British journalist, sought out Burt's co-workers, Jane Conway and Margaret Howard. They could not be found. They were just as much inventions of Burt's fertile mind as his separated twins.

Why do I drag up these old cases? Because we may be witnessing a fraud of even greater proportions. The Climategate e-mails have shaken people's confidence in anthropogenic global warming. In the UK more people disbelieve it than believe it. I am not a climate scientist and I would not be any good at disentangling the data that has been presented, but have some experience of scientific fraud.

The e-mails seem to be saying that there was a conspiracy to deny doubters access to the raw data on climate measurements. Freedom of information requests can be troublesome for researchers. They may involve a good deal of time which would be better spent on getting on with the job. But the e-mails seem to be saying that data were destroyed rather than allowing them to be seen by sceptics. It wasn't only the University of East Anglia involved; other climate change researchers seem to have been collaborating in what must be called a cover-up. There seems to have been a conspiracy to 'peer-review' particularly harshly any papers suggesting an alternative explanation for the data so as to prevent their publication.

I find it hard to come to any other conclusion about what was going on.

On these two points alone I am extremely worried for the reputation of this field of science. The purpose of publication is to invite criticism. How ever much evidence you accumulate to support your hypothesis, it cannot be proved to be true. It may be contingently true; contingent on someone not coming up with contrary evidence. But if you refuse to look for that contrary evidence then your hypothesis is very weak indeed.

Consider the hypothesis that ALL SWANS ARE WHITE. You might travel up an down the land from Cornwall to the Shetland islands and demonstrate that every swan you see is white, but unless you travel to Australia, you really haven't tested your hypothesis. The climate conspiracy is worse. It is like they have seen early stamps form South Australia that were adorned with a black swan, but they refuse to go and look there, declaring that the picture was a figment of the artist's imagination.

Worse still is the hiding of data. Several frauds have involved fictitious or manipulated data. Publication of raw data in electronic form has become the norm in many scientific fields, and it should always be available to journals if a question is raised about the article. Malcolm Pearce, a British obstetrician reported that he had rescued an ectopic pregnancy by reimplanting the fetus in the womb. He was exposed as a fraud because he could not produce the patient's record. Imagine if he had said, "The records are confidential. I am not going to reveal them to those silly individuals who don't believe me."

Science involves two things: 1] imagining an explanation for an observation; 2] designing an experiment to prove the explanation wrong. I am worried that climate scientists are very good at 1] but instead of doing 2] they have become polemicists whose role is to convince others that their view is right. The Himalayan glaciers were an instance of polemicism rather than science.