Scientists should resist the temptation to hype their results
According to an Editorial in today's Nature ( 461, 1174; 2009 - free to read online), "the temptation for scientists and their institutions to spin their research to the media, or to go publicity-mongering, is always there. And — as illustrated by the excessive public-relations campaign surrounding "Ida", a fossil presented as a missing link in human evolution (see Nature 459, 484; 2009 and Nature 461, 1040; 2009) — too many in the media will buy into the initial hype. Such behaviour is corrosive to the process of scholarly scientific communication. Research institutions must not allow it to become the norm."
The Editorial discusses the recent announcement of results from an HIV vaccine trial in Thailand, in which the trial's sponsors announced that it had been a success in that the vaccine had a statistically significant effect on preventing infection. But the full data for the claim were not made available for almost a month after the announcement - and included two other data sets in which the effects were not statistically significant.
Fortunately, states the Editorial, such stories are still rare in science. "Witness the way scientists have behaved since the beginning of the current H1N1 flu pandemic, in which the urgent threat to health creates legitimate tensions between getting results out fast and respecting peer review. Most researchers have negotiated this tension well, through a combination of fast-track publication by journals and online pre-publication sharing of preliminary data — but not through hyping their results."
