Edinburgh, Scotland - Scientists are routinely falsifying their findings, according to the first meta-survey on the topic.
While this news may come as no surprise to many of our readers - see the comments on more or less any one of our science stories - Dr Daniele Fanielli of the University of Edinburgh said the problem was far more extensive than had previously been realised.
The findings are based on a review of 21 scientific misconduct surveys carried out between 1986 and 2005. "The point is to help people acknowledge that this happens and make things better," Dr Fanielli told TG Daily. "Science being a human enterprise, it is bound to have defects."
Of course, it's always the other guy what done it. While only two percent of those surveyed admitted to fabricating, falsifying or altering data to improve the outcome, 14 percent reckoned they knew someone else that had.
A further 34 percent of people admitted other dodgy practices, such as 'failing to present data that contradict one's own previous research' and 'dropping observations or data points from analyses based on a gut feeling that they were inaccurate'. But 72 percent said they knew someone else that had.
Medical and pharmacological researchers reported the most misconduct - possibly indicating that industrial sponsorship is severely distorting scientific evidence, says the report, in order to promote commercial treatments and drugs.
The findings will be published in the online, open-access journal PLoS ONE.
Dr Fanelli says he doesn't expect the report to make him popular amongst his colleagues. "It's probably going to be that way. But someone's got to do the dirty work," he said.
And the obvious question: how can we trust his conclusions? "I'm expecting people to look very, very carefully at my findings," he said.