Science: Lab misbehavior: not so rare?
The high-profile cases are as shocking as they are rare: a physicist at Bell Labs fired for fudging data in more than a dozen research papers; a menopause researcher at the University of Vermont College of Medicine admitting to faking data in grant applications and published articles; a German evolutionary biologist busted for claiming credit for the work of his students. But according to a team of researchers from Minnesota, focusing on the most egregious examples of scientific misconduct misses a potentially even more damaging phenomenon.
Writing this week in the journal Nature, researchers from the University of Minnesota and a nonprofit research foundation present the results of a survey on unsavory behavior in labs that receive funding from the U.S. National Institutes of Health. More than 3,000 researchers respondedanonymouslyand fully 33 percent of them admitted to engaging in at least one of the survey's "top 10" infractions, from fudging research data to ignoring confusing or contradictory results. The most serious lapses were relatively rare; fewer than 2 percent of respondents admitted to falsification, fabrication, or plagiarism. (As with any survey of wrongdoing, the numbers are almost certainly an underestimate.)
But two of the most commonly admitted infractions are also among the most disturbing. Science is largely a self-policing enterprise, relying on supposedly rigorous "peer review" of grant applications and research results to enforce standards of experimental design and objective interpretation. Yet 12.5 percent of the survey respondents admitted to overlooking the use of questionable data or suspect interpretations by other scientists. And in an era of increased concern about corporate and political interference in the research enterprise, 15.5 percent of the researchers admitted to bowing to that pressure, changing the design, methodology, or results of a study in response to pressure from funding sources.
Straying from the straight and narrow is hardly a new phenomenon in scienceeven Gregor Mendel, the Swiss monk credited as the father of genetics, is thought to have tossed out data that didn't confirm the general trends in his breeding experiments with pea plants in the mid-19th century. But Martinson and his coauthors suggest that in an era of increased competition for research grants, faculty positions, and space in prestigious journals, scientists face greater temptations than ever to bend the strict rules of research conduct.
Occasional arrogance aside, scientists are indeed only human, and as in any human endeavor, there will always be some individuals who choose to cut corners and even lie, cheat, and steal to get ahead. The worst offenders are almost always found out. But in an increasingly politicized culture, where research results can be buried to boost corporate bottom lines or spun to fit a predetermined set of political goals, the apparent pervasiveness of even low-grade misbehavior is troubling indeed.