Over the course of his career Burt published research on the heritability of intelligence as measured in IQ tests using twin studies. It is now generally accepted that, at least a majority of, this research was fraudulent. The possibility of fraud was first brought to the attention of the scientific community by Leon J. Kamin in 1974. The smoking gun that Kamin cited was the constancy of Burt's correlation coefficients to 3 decimal places even when new data were twice added to the sample of twins.
The investigation that was conducted once Kamin had raised the alarm uncovered a number of other disturbing issues. In particular, two coauthors of Burt's published papers could not be traced, strongly suggesting that Burt had invented them.
The correlation coefficient given by Burt for the IQs of identical twins reared apart is within the range that has been found by reputable studies conducted since his death. This has led some commentators to the conclusion that Burt may have conducted some genuine research before he took to fabricating further data to match the results he had already obtained.