January 4, 2011

Retraction Watch is watching you

Charles Day
Big, scandalous cases of scientific fraud are widely covered in the popular press. In the early 2000s Jan Hendrik Schön of Bell Labs published 21 papers about organic semiconductors: seven in Nature, six in Physical Review Letters, and eight in Science. All of them were withdrawn when it turned out that Schön had faked the results.
Schön's notoriety was so great that he became the subject not only of news reports, but also of books and even a BBC TV documentary, "
The Dark Secret of Hendrik Schön."
Trends in scientific fraud also make the news, although not as often. Last September a
report in Nature about a move to kill off China's weakest scientific journals began as follows:
Few Chinese scientists would be surprised to hear that many of the country's scientific journals are filled with incremental work, read by virtually no one and riddled with plagiarism. But the Chinese government's solution to this problem came as a surprise last week.

Low-key, "routine" cases of scientific fraud don't appear to be newsworthy. Like shoplifting, vandalism, disorderly conduct, and other misdemeanors, such cases are deemed of local, not national, interest. News of a plagiarized paragraph in a chemistry paper, say, might appear on a chemistry blog; less likely in the New York Times.
But a steady background of petty fraud harms the integrity of science more than sporadic spectacular outrages. A Schön or a Viktor Ninov, who faked evidence of a newly discovered superheavy element, can be excused as a pathological outlier. Widespread fraud suggests something intrinsically wrong with the science establishment.
So I was relieved to hear from Marty Hanna, a Physics Today copy editor, about
Retraction Watch. Founded by Adam Marcus and Ivan Oransky, the blog strives to publicize every fraud-prompted retraction that occurs in the scientific literature.
Marcus and Oransky aren't the only watchdogs. Academic publishers, both nonprofit and for-profit, are collaborating to implement a software tool,
CrossCheck, that screens for plagiarism when a paper is submitted.
Ideally, scientists shouldn't cheat. Realistically, some scientists, under pressure to succeed, will always succumb to temptation and commit fraud. When they do so, and when the watchdogs catch them, I hope they feel guilty and ashamed. That reaction would mean that efforts of the American Physical Society and others to
instill ethical behavior are working.


Popular Posts