How common is scientific misconduct? It's a tough question to answer. Scandals like Hwang Woo-Suk's faked stem-cell lines or Jon Sudbø's made-up cancer trials have demonstrated that fraudulent research is easy to publish, even in the most prestigious print journals, but are they deviations of a few "bad apples" or evidence that a great deal more is never discovered? The actual numbers are a matter of dispute.
In a PLoS ONE meta-analysis of surveys questioning scientists about their misbehaviors, Daniele Fanelli of the University of Edinburgh, a PhD in the behavior and genetics of Malaysian and Panamanian tropical wasps, suggests that altering or making up data is more frequent than previously estimated and might be particularly high in medical research.