Page 21

CIM36

SPECIAL REPORT 21 WINTER 2015 N° 36 What is scientific “fraud?” An international consensus defines fraud as a “serious and intentional violation in the conduct of research and the dissemination of results,” with the exclusion of “errors made in good faith or honest differences of opinion.”3 The international scientific community identifies three main types of fraud, known by the acronym FFP: fabrication, falsification, and plagiarism (see box, right). Fabrication consists in making up research data from scratch; falsification means manipulating it intentionally to make it match the desired hypotheses; and plagiarism is the use of another researcher’s findings or ideas without their knowledge and without giving them proper credit. But these serious breaches of scientific ethics, which some would like to be made liable to criminal prosecution, 4 must not completely overshadow what a notable article published in Nature 5 in 2005 called “misbehavior” and “carelessness”—in other words, the more or less conscious violation of good scientific practice. Albeit less obvious, at least in terms of their immediate effect, and thus more difficult to detect, such examples of unethical behavior nonetheless amount to scientific misconduct. As Michèle Leduc and Lucienne Letellier of the COMETS explain,6 “there is a continuity between outright fraud and manipulated results. Data is ‘cooked’ so as to keep only the desired findings; images are altered with Photoshop; result s are has tily published without proven reproducibility; researchers remain evasive on their experimental protocols so that they cannot be verified or copied; results are covered up; single experiment data is spread out over several articles, making them incomprehensible individually, etc.” Such breaches of good practice never make headlines, but they create bad habits which, in the long run, have lasting consequences on the overall quality of accepted scientific knowledge. Malpractice on the rise The general media is definitely devoting more coverage to scientific scandals, but is fraud, large or small, really such a common phenomenon? And more importantly, is it really getting worse? In his 1992 report, Larivée and co-author Maria Baruffaldi compiled a list of more than 200 known frauds committed between 1800 and 1992 (before the widespread adoption of the Internet) by researchers in five fields covering 40 different disciplines. They noted that 73% of cases took place after 1950, and somewhat more surprisingly, that 58.9% of them were in medical research. The development of online bibliographic databases that track published scientific articles and retractions, like PubMed or Web of Science, now makes it possible to monitor this type of statistics in real time.7 These sources seem to confirm that the incidence of fraud has increased and is even accelerating: the proportion of retractions Crossing Borders One of the main obstacles to drafting a common international policy against scientific misconduct is the wide-ranging definitions (if any) that individual countries may have adopted over the years—depending on their culture, or on the very nature of their respective languages. As an illustration, Daniele Fanelli,8 an authority on the issue, explains that if English distinguishes between “ethics” and “integrity,” equivalence for the latter does not exist in Finnish, which hinders the proper translation of official documents.9 Finland, through its National Advisory Board on Research Integrity was in fact the first country to have created—in 2002—two distinct categories separating “fraud,” the equivalent of fabrication, falsification, or plagiarism (FFP), from “misconduct,” which involves gross negligence and irresponsibility in the conduct of research. In China on the other hand, scientific misconduct covers FFP, but also the submission of false curriculum vitae and violations of rules pertaining to human or animal research. In Australia, it includes the failure to declare a serious conflict of interest. Some countries have debated whether scientific misconduct should go as far as addressing issues of social responsibility. Fanelli argues that a possible explanation for this wide range of definitions may be their intended objective: if it is to hold researchers accountable, then a narrow and closed definition is preferred (such as strictly FFP), but if it is to promote responsible research or foster research integrity, then definitions must enter a broader “grey” area with lower threshlods of intent. To further complicate matters, some countries entirely lack documented and formalized processes for dealing with scientific misconduct. But nation-bound The US Congress was the first—as early as 1989—to address the issue by creating the Office of Scientific Integrity, now Office of Research Integrity (ORI), funded by the Department of Health and Human Services. Like the National Science Foundation (NSF), the ORI can push for legal sanctions when fraud is identified in publicly-funded medical research, with results accessible on its website. In Europe, the Scandinavian countries took the initiative in the 1990s, with Denmark leading the way through the creation of a Committee on Scientific Dishonesty, able to conduct investigations in both private and publiclyfunded research. Yet models vary throughout Europe, from Danish-like centralized models (Norway) to a mix of decentralized and centralized structures (Sweden), to fully decentralized systems (Finland), or ones solely monitored by national research organizations and funding bodies (France and Germany).10 In France, the medical and health-related research agency INSERM was the first to set up a Committee on Scientific Integrity (DIS)11 in 1999, followed by the Institut Pasteur (2004) and more recently, by the ANR.12 The CNRS has also put in place safeguards and disciplinary measures, and can call upon external experts to assess the nature of misconduct should it not be properly handled by the laboratory. Yet many countries solely rely on independent bodies, mediators, and advisors with little legal authority over laboratories, research institutions, or universities (India’s Society for Scientific Values, for example). …


CIM36
To see the actual publication please follow the link above