The scientists were under career pressure, and thought they knew the right answer.

In every case of scientific fraud I knew of, I realized that three factors were present: the scientist was under career pressure; he thought he knew the answer, and didn't need to go to all the trouble of obeying the scientific method; and he was working in a field where reproducibility was not precise. The last of these explained why fraud was almost always in biomedicine, where the truth is generally more statistical and less directly causal.

Then, in 2002, two cases of scientific misconduct by physicists came to light, one involving Jan Hendrick Schön at Bell Laboratories, and the other Victor Ninov at Lawrence Berkeley National Laboratory (LBNL). In both cases, the scientists were under career pressure (as most scientists almost always are) and both thought they knew the right answer. The test of my hypothesis would be the third factor.

The Schön case fits like a glove. He apparently made a breath-taking series of discoveries in MOSFETs (metal-oxide semiconductor field-effect transistors), a field that is notoriously sample-dependent: the fact that nobody could reproduce his results could just have meant that they had bad samples.

Victor Ninov was a leader of the group at LBNL using the Berkeley Gas-filled Separator (BGS) to sort through the debris of nuclear collisions. In 1999, the group announced the observation of three instances of decay chains characteristic of the element 118. By international agreement, new elements are not official until their discovery has been reproduced, which groups in Germany and Japan immediately undertook to do — but both failed.

In 2001, the BGS group produced a fourth signature decay chain. By now, suspicions had been aroused. A series of investigations ensued, in which it was found that the data for all four decay chains had been fabricated, and that Ninov was the only person in a position to have done it.

Thus he had turned my third factor on its head. Ninov had assumed that his result would be reproduced — and that he would get the credit for discovering it first.