London School system as a source of twin data. But the most serious problem with the finally published twin data is the discrepancy between the numbers Burt sent to Jencks in 1969 and Burt ( 1966) and Conway ( 1958) earlier statements, coupled with Burt's diary entry in which he talked of "calculating twin data for Jencks'. It is this that suggests that the correlations reported by Burt and Conway were not based on real numbers, and that Burt had been forced to work out some new numbers in 1969 in order to provide the scores which Jencks had requested. And a final cause for suspicion, as Jensen notes in his chapter, is Burt's claim to Sandra Scarr in 1971 that he had recently obtained data on three new pairs of twins.
None of these cases, taken alone, is necessarily decisive--sufficient to establish guilt, in a court of law, beyond reasonable doubt. Of each, taken in isolation, the defence could reasonably say that there must be an alternative explanation--even if we do not know what it is. And since Burt is no longer alive to provide that defence, the only reasonable verdict must be: not proven. But we are not trying a case in a court of law. We are simply attempting to arrive at the most plausible explanation of the available evidence. And when all these cases are considered together, a defence which can only insist that there must be some other, more innocent explanation begins to lose its force. One can appeal to an unknown, even somewhat implausible, explanation once. To have to do so two or three times makes one's case rather less persuasive. We may say that it is implausible to suppose that a man of Burt's stature and reputation should have committed fraud. No doubt it is. No doubt it is implausible to suppose that Newton did. But if the defence is equally implausible, the case for the prosecution starts to look that much more convincing. We know that Burt could be devious and dishonest in small things, and was sometimes determined to win at any cost. We know that he had a relatively cavalier attitude to the reporting of empirical data. We know that he was prepared to conceal things about his data that would have made those data very much less persuasive, and we can be confident that he was sometimes prepared to adjust his data, and at other times make false claims about them, in order to make them appear more convincing. On balance, I believe that the evidence makes it more probable than not that some of the data he reported existed only in his imagination, in other words that he fabricated them.
Bouchard T. J. Jr. ( 1983). "Do environmental similarities explain the similarity in intelligence of identical twins reared apart?" Intelligence, 7, 190-1.