Rotten to the core? Far from it

Cheats fascinate us, particularly those who are at the top of the pile. The textbook example is Lance Armstrong, who beat cancer then went on to win the Tour de France seven times - and burnished his celebrity by dating Sheryl Crow, a singer who has sold 50 million albums despite being as familiar with the middle of the road as Armstrong himself.

When he was finally forced to put an end to the lies, one of his justifications for taking performance-enhancing drugs was that he had been operating in a world in which doping was the norm, so all he was really doing was levelling the playing field.

“I went in and just looked up the definition of cheat and the definition of cheat is to gain an advantage on a rival or foe that they don’t have. I didn’t view it that way,” he said.

The main problem is that reviewers are often ill-equipped to pick up fraud: how are they supposed to spot when data have been faked?

There’s no doubt that cycling was dirty far beyond Armstrong’s immediate circle, but it’s also clear that his fame contributed to the long shadow cast over the sport as a whole.

This week we look at the impact a rotten apple can have within an academic discipline.

Diederik Stapel, like Armstrong, was a star in his field: social psychology. Like Armstrong, he suffered a spectacular fall from grace.

Stapel’s modus operandi was to pick a topic, decide on findings, then fabricate results to fit. The details have been picked over at length, but the case, coupled with a small number of others, has also led to wider questions about whether something is fundamentally rotten in his field.

It’s a legitimate question, but in our cover feature this week (‘Social psychology is primed but not suspect’), two professors of social psychology - Wolfgang Stroebe, who holds posts at the University of Groningen and Utrecht University, in the Netherlands, and Miles Hewstone, from New College, Oxford - argue that an objective review of the discipline does not support claims of general malaise.

It would also be a mistake to whitewash the problems that exist: across all disciplines, the rate of retractions is rising faster than the publication rate, and the five worst offenders - all but one of them in the field of medicine - have published a total of 5 articles based on falsified results.

But Stroebe and Hewstone argue that social psychology has suffered unfairly as a result of misjudged interventions in the fallout from the Stapel affair by people who should have known better than to make generalisations on the basis of the misdeeds of a tiny minority.

One issue addressed by the committees that investigated Stapel was the failure of peer review but, as the feature points out, this is hardly unique to social psychology.

The main problem, highlighted in this case, is that reviewers are often ill-equipped to pick up fraud: even as experts in their fields, how are they supposed to spot with consistency when data have been faked? There will sometimes be warning signs but, as with doping controls in sport, there are many ways to avoid detection.

And yet, as is often pointed out, peer review, like democracy, remains the worst system we have, except for all the others. What the Stapel and Armstrong cases do is reinforce how essential it is to guard against complacency (which can creep into any well-established system) and to ensure that everyone is treated without fear or favour.

john.gill@tsleducation.com