Page 23

CIM36

SPECIAL REPORT Misconduct admitted by researchers in the past three years In 2002 23 WINTER 2015 N° 36 Failure to mention data that contradicts the researcher’s own results, or the intentional use of doubtful data Changing an experimental plan or the results of a study under pressure from a funding source Publishing the same results in several publications Omitting or falsely including an author credit in a published article US laboratories between 1986 and 2005,15 1.97% of respondents confess that they have personally falsified their experimental data at least once, and 14.12% report having witnessed this type of misconduct among their colleagues. Moreover, while 33.7% of researchers admit to other types of unethical behavior, as many as 72% state that they have seen others indulge in such practices. In short, although evaluation methods are approximate by nature, the frequency of fraud, big and small, seems to have significantly increased in recent years. Some scientific misconduct is committed for personal gain (professional advancement, a prize, or peer recognition) or for ideological reasons (to justify or promote a policy) and amounts to simple fraud and/or manipulation. Such incidents, once discovered, have given rise to the most widely publicized scandals of the past decade. In addition to the stigma cast on the laboratories concerned, they have destroyed the careers and reputations of the perpetrators. But the damage caused by scientific fraud goes far beyond the fate of those more or less directly involved. It can lead other researchers astray and invalidate entire theses based on false data. It also tarnishes the image of research as a whole and jeopardizes its funding. As Larivée notes, “it may not take much, especially during an economic recession, for taxpayers, influenced by the sensationalist media coverage of a few notorious cases, to start questioning the government budget allocated to scientific research in a given field or even in its entirety.”2 Could science as a whole fall into disrepute? From a strictly scientific point of view, the worst consequence of fraud is the suspicion that it casts on the body of knowledge acquired through research. The repercussions of this uncertainty are not just epistemological, but also have far-reaching social effects. “If researchers no longer trust the validity of their colleagues’ findings, they are faced with a choice of either working full time reproducing their experiments—or live with constant, disquieting doubt,” warns Larivée. “Given the preponderant role of science in our modern-day societies, the political … Because of their importance to public health and the availability of virtually exhaustive databases, the biomedical sciences have the greatest number of objective indicators for evaluating fraud. This has made it possible to observe a sharp increase in the number and proportion of retractions over the past 30 years. It has also shown that, of the 2047 biomedical articles retracted between 1977 and 2012, only 21.3% were discredited due to simple error, while 53.2% were found to be fraudulent or plagiarized. In 2002, 3247 US- and UK-based researchers in the early to middle years of their careers agreed to answer a questionnaire on the types of misconduct that they had committed or witnessed. The graph on the right shows that, while serious fraud is relatively rare, incidents of misconduct are quite common among both beginning and experienced researchers. Experienced researchers Beginning researchers Misconduct toward colleagues, students, or human subjects Concealing methodological details or results in an article Following an unsuitable experimental plan Excluding certain observations deemed “aberrant” without predefining the criteria for exclusion 13 % 9 % 19 % 14 % 21 % 9 % 6 % 3 % 12 % 7 % 12 % 9 % 12 % 14 % 16 % 27 % 15 % Failing to maintain proper records or archiving procedures 28 % SOURCE: MARTINSON ET AL. (2005) NATURE © INFOGRAPHIES: S. LANDEL FOR CNRS LE JOURNAL


CIM36
To see the actual publication please follow the link above