False Data Charge Stirs I.Q.-Heredity Controversy

Recent allegations that a famous English psychologist falsified the data that led him to conclude intelligence is highly hereditary have led to a sharp controvery among experts at Harvard and other universities over the validity of those conclusions.

The doubts about the work of Cyril Burt, an English researcher who concluded that intelligence is 80 per cent inheritable, were first expressed by Leon Kamin, a professor at Princeton University, in 1972. Kamin has said Burt's reported data is far too consistent to be true.

Burt's research has been quoted by many scientists who argue that low test scores of black children result from genetic inferiority rather than environment, David R. Layzer '47, professor of Astronomy, said yesterday.

A Major Blow

Supporters of Kamin's theories and those who believe I.Q. scores to be largely determined by environment, not heredity, point to Burt's works as the only substantial scientific evidence contradicting their "environmental influence" theories, and see its apparent refutation as a major blow to its credibility.

Even though some adherents of the "hereditary influence" view admit that some of Burt's actual numerical data may have been falsified inadvertently, they still contend that his conclusions are correct.

Scientists who have received public attention for their statements on heredity include Layzer, Kamin, Richard C. Lewontin '50, professor of Biology and Stephan J. Gould, professor of Geology. Agreeing with Burt's conclusions, if not his data, are Richard J. Herrnstein, professor of Psychology, and Arthur Jensen, professor of educational psychology at the University of California at Berkeley.

In the portion of his work in question, Burt tested the I.Q. of pairs of identical twins that had been placed in different homes in infancy. In papers written at three different stages of his work, Burt reported identical statistical correlations between the I.Q. scores of the pairs of twins. He concluded that this high correlation (79 per cent) proved the dominant influence of heredity in determining intelligence.

Burt, who received British knighthood for his work, died in 1971.

Recently Kamin and a British newspaper found an unusually high degree of consistency among Burt's data. The biggest question mark, Kamin said, is how Burt arrived at exactly the same correlation coefficient from three different sets of data.

Gould said yesterday that the odds against that result in actual research data are astronomical.

Gould also mentioned doubts concerning the existence of Burt's two research assistants. He said that "as far as anyone can tell," the assistants were completely fictitious, fabricated by Burt to lend credence to his results.

Gould accused Burt of "conscious fraud" in fabrication of his data. He cited studies refuting hereditary influences on intelligence and said that there is "nothing really substantial" left to support Burt's theories.

Burt Defended

Herrnstein said yesterday, however, that although he considers Burt's consistency of data "moderately implausible," he does not believe Burt intentionally falsified data.

More importantly, he cites three studies published since Burt's first reports that give correlation figures of 77 per cent, 77 per cent, and 73 per cent. He said it has been established that at least one of the assistants named by Burt is a real person, not a fabrication.

Herrnstein said that the work of the researchers who produced the three studies with conclusions similar to Burt's has not been called into question "probably because they are still alive to defend themselves."

He said another study by Burt, published after his death, found a 49 per cent correlation in parent-child I.Q. scores and that a review by Christopher S. Jencks '58, professor of Sociology, of five other parent-child studies found a 48 per cent correlation.

Doubts Remain Strong

But the doubts among skeptics remain strong. Gould believes Burt's data "are unusable." Even Herrnstein admits he would no longer use them in papers of his own, since "the appearance of trust-worthiness is as important as trust-worthiness itself" in academic documentation.

Layzer said that most adherents to the heredity view have "some political ax to grind," citing their support of tougher immigration laws and other "exclusionary legislation." He said that "these people believe the I.Q.-heredity correlation is high. Nothing will shake that belief. They believe those numbers.