The Science of Fraud

Two stories caught my attention in the past week or so, both about fraud in the scientific community. The first is this story about Annie Dookhan, who seems to be the major reason for " a burgeoning investigation that has already led to the shutdown of the lab, the resignation of the state's public health commissioner and the release of more than a dozen drug defendants." In addition to pretending to have a Master's degree from the University of Massachusetts, Dookhan now admits to having faked the majority of her tests for a Massachusetts state lab, most of which involved criminal narcotics testing. Perhaps thousands of accused were wrongfully convicted in the nine years in which Dookhan worked there. The article explains the situation better than I have, but makes only a feeble attempt to discover why Dookhan might have committed such comprehensive fraud, other than that she "wanted to be seen as a good worker."

The other story involves a larger fraud trend in the scientific community; according to a new study, fully two thirds of retractions of biomedical papers are due to fraud or suspected fraud. The study divides types of scientific misconduct into "fraud or suspected fraud (43 percent), duplicate publication (14 percent), and plagiarism (10 percent)." The report suspects that many other instances of fraud may go undetected to bring the numbers even higher.

The stories are not particularly distinct from others I have read, I suppose. It's quite possible that it was simply harder to detect scientific misconduct in 1975 than it is now, and that it has not actually increased "tenfold" as the study suggests. Annie Dookhan may be one of a series of impostors whose purposes for fraudulence will never be determined to anyone's satisfaction. I am struck by them, perhaps, because I still labor under the delusion that science is more objective, more factual than other disciplines. The forger of scientific data must be more brazen even than the literary forger, because he knows that someone will almost certainly check his data. In fact, Dookhan got around those checks by committing yet another fraud and signing others' names to her work. Does the fraud count on the scientific community to be lazier than he even is, or is his pressure to publish so great that he is scared and rushed into sending work out that is either not tested or not his own? Are similar trends happening in humanities publications, or is this type of misconduct not as possible there?