A Taint of Scholarly Fraud

The late British psychologist Cyril Burt was eminent in his profession: he held the psychology chair at London's University College, was knighted by King George VI and won the Thorndike award from the American Psychological Association. As a government adviser, he helped restructure the British educational system in the 1940s. Now, five years after his death, Burt is the object of a growing scandal. He has been accused of doctoring data and signing the names of others to reports that he wrote. If the charges are proved true, said Science magazine last week, "the forgery may rank with that of the Piltdown man."

Much of Burt's reputation rested on his prominent role in the debate about heredity and intelligence. His studies of identical twins who grew up apart indicated that heredity--rather than environment--explains most of the differences in IQ scores. But shortly before Burt's death in 1971 at the age of 88, there were academic murmurs that the psychologist's data were suspect. For one thing, the statistical correlation between IQ scores of his identical twins remained the same to the third place after the decimal point as more and more twins were studied--an extraordinary and highly unlikely coincidence. Yet most experts assumed it was an honest and unimportant mistake. "As he got old," said British Psychologist G.C. Drew, "he was remembering old figures that got stuck in his mind."

The doubts became public knowledge when the London Sunday Times reported that Burt's co-authors of the later twin studies--Margaret Howard and J. Conway--are not listed in London University records and are unknown to 18 of Burt's closest colleagues. The revelation is crucial: the two women were presumably Burt's field investigators on the twin research at a time when the psychologist was becoming feeble and deaf. It thus seems increasingly possible that the women never existed, that their investigations were never carried out and that Burt invented them and their reports.

Since the Sunday Times story, a Manchester professor has recalled meeting a Margaret Howard in the 1930s, but the only other traces of Conway and Howard are their signatures on reviews in the late 1950s published in the British Journal of Statistical Psychology. Those writings, mostly attacking Burt's enemies, stopped around the time Burt stepped down as the journal's editor. Says Princeton Psychologist Leon Kamin, an opponent of Burt in the heredity-intelligence debate: "It was a fraud linked to policy from the word go. The data were cooked in order for him to arrive at the conclusion he wanted."

Burt's allies prefer to believe the psychologist was careless but honest. The suggestion of fraud "is so outrageous, I find it hard to stay in my chair," says Harvard Psychologist Richard Herrnstein. "Burt was a towering figure of 20th century psychology. I think it's a crime to cast doubt over a man's career." Professor of Educational Psychology Arthur Jensen of the University of California at Berkeley adds: "If Burt was trying to fake the data, a person with his statistical skills would have done a better job. It is a political attack. The real targets are me, Herrnstein and the whole area of research on the genetics of intelligence."

At best, Burt's methods were incredibly sloppy. The raw test sheets on the twin studies were among papers stuffed into half a dozen tea chests and later destroyed. Many of his professional articles do not give primary data, referring readers to unpublished reports. Some of those reports, says Kamin, are at least as hard to find as are Howard and Conway.

Why did Burt's work go unchallenged during his lifetime? Says Philip Vernon, a collaborator of Burt's now at Alberta's University of Calgary: "There were certainly grave doubts, although nobody dared to put them into print because Burt was so powerful." In fact, he was powerful enough to see his ideas on heredity and intelligence translated into educational policy. As a government adviser in the 1940s, he played a prominent role in setting up the three-tier British school system that pigeonholed students on the basis of an IQ test given at age eleven.

That system has since been dismantled, and the controversy over Burt is unlikely to have much effect on educational policy. It will also make little impact on American psychologists who believe that heredity is crucial to intelligence; they have produced several twin studies similar to Burt's. Says Herrnstein, "I know of no correlation of Burt's which is seriously challenged in the literature." But Harvard's Richard Lewontin, a population geneticist, says that Burt's work with twins "is the only large study which is methodologically correct, so its loss is no trivial problem for the heritability people. It is also not nice for them to have this mess in their backyard."

It is also a mess for the entire field of psychology, which is still struggling to be taken seriously as a rigorous science. When a leader in the field is shown to be either a fraud or spectacularly inept, it is psychology's loss.

This file is automatically generated by a robot program, so viewer discretion is required.