The History and Significance of the U.S. Federal Definition of Misconduct in Science. Concern over misconduct in science grew in the United States in the late 1980s due to several highly publicized cases. Thanks in part to federal pressure, most research universities have since adopted explicit definitions of misconduct in science2 and procedures for handling allegations of misconduct. The early history of investigations is one of very shoddy work, investigations that looked like cover-ups, and tremendous denial on the part of the scientific community. We have come a long way since 1989 in that regard. This is not to say that all of the procedures now in place are perfect, but they are certainly much better than they were initially. We have learned a great deal from our mistakes. There has also been increased attention to teaching the responsible conduct of research, which I think is a commendable development.
The federal definitions of misconduct in science proposed and implemented since 1989 have led to ongoing controversy. I will go over some of the history of that controversy and speculate on what has made it so difficult to find a definition of misconduct in science that is not controversial.
But first, why should anyone care? From a philosophical point of view, federal definitions of misconduct are about as interesting as a military court's definition of "mercy" would be to a theologian. The federal definition of misconduct does not shed a good deal of light on the nuances of ethical behavior, but it is of tremendous importance for practical purposes. Obviously administrators who have to enforce federal regulations and American scientists who might be faced with charges of misconduct have a great stake in the definition of misconduct. It is also important for the public image of science in the United States, which has been seriously damaged by many scandals. Finally, it seems clear that the first line of defense against misconduct in science is the research university, which is charged with investigating internally allegations of misconduct; and it seems very likely that all major American universities will simply adopt the federal definition of misconduct. To adopt a less stringent defintion would be to effectively give up all federal funding; to adopt a more stringent definition would be to run a tremendous risk of lawsuits from scientists found guilty of misconduct, who would charge that the university's definition, being more stringent than the federal definition, was not legitimate.3
Definitions of Two Major Federal Funding Agencies. The Public Health Service (PHS) definition:
"Misconduct" or "Misconduct in Science" means fabrication, falsification, plagiarism, or other practices that seriously deviate from those that are commonly accepted within the scientific community for proposing, conducting, or reporting research. It does not include honest error or honest differences in interpretations or judgments of data. [PHS 1989:32449]The National Science Foundation (NSF) definition:
"Misconduct" means (1) fabrication, falsification, plagiarism, or other serious deviation from accepted practices in proposing, carrying out, or reporting results from activities funded by NSF; or (2) retaliation of any kind against a person who reported or provided information about suspected or alleged misconduct and who has not acted in bad faith. [NSF 1991:22287]There are some interesting differences between these two definitions.
The National Academy of Science's Response. "In 1989, the National Academy of Sciences (NAS), the National Academy of Engineering, and the Institute of Medicine (IOM) initiated a major study to examine issues related to scientific responsibility and the conduct of research. [The study was undertaken by] the Committee on Science, Engineering, and Public Policy." [COSEPUP 1992:ix]
COSEPUP proposed three categories of behaviors: "(1) misconduct in science, (2) questionable research practices, and (3) other misconduct" (COSEPUP 1992:25).
Misconduct in science is defined as fabrication, falsification, or plagiarism, in proposing, performing, or reporting research. Misconduct in science does not include errors of judgment; errors in the recording, selection, or analysis of data; differences in opinions involving the interpretation of data; or misconduct unrelated to the research process."Other misconduct" includes, e.g., "sexual and other forms of harassment of individuals; misuse of funds; gross negligence by persons in their professional activities," etc. (COSEPUP 1992:29).Fabrication is making up data or results, falsification is changing data or results, and plagiarism is using the ideas or words of another person without giving appropriate credit. [COSEPUP 1992:27, emphasis in original]
Questionable research practices are actions that violate traditional values of the research enterprise and that may be detrimental to the research process. However, there is at present neither broad agreement as to the seriousness of these actions nor any consensus on standards for behavior in such matters. Questionable research practices do not directly damage the integrity of the research process and thus do not meet the panel's criteria for inclusion in the definition of misconduct in science. However, they deserve attention because they can erode confidence in the integrity of the research process, violate traditions associated with science, affect scientific conclusions, waste time and resources, and weaken the education of new scientists.
Questionable research practices include activities such as the following:
- Failing to retain significant research data for a reasonable period.
- Maintaining inadequate research records, especially for results that are published or are relied on by others;
- Conferring or requesting authorship on the basis of a specialized service or contribution that is not significantly related to the research reported in the paper;
- Refusing to give peers reasonable access to unique research materials or data that support published papers;
- Using inappropriate statistical or other methods of measurement to enhance the significance of research findings;
- Inadequately supervising research subordinates or exploiting them; and
- Misrepresenting speculations as fact or releasing preliminary research results, especially in the public media, without providing sufficient data to allow peers to judge the validity of the results or to reproduce the experiments. [COSEPUP 1992:28, emphasis in original]4
Note that the COSEPUP definition separates the "other serious deviations" clause into a completely separate sub-definition. But even this was not acceptable to some, as the "Minority Statement" to the COSEPUP report showed:
The report is equivocal in defining misconduct in science and is inadequate in stating explicitly the problems inherent in alternative definitions. The "other misconduct" category introduces ambiguities into the definition, and blurs the boundaries between misconduct in science and questionable practice. Misconduct in science requires rigorous adjudicatory machinery and governmental oversight, protection of whistle-blowers, due process, strong sanctions, and full disclosure. In contrast, questionable practices raise issues about the value system and culture of science, and underscore the need for explicit dialogue and education. Governmental intervention is inappropriate for concerns regarding errors in collecting and interpreting data, incompetence, sloppiness, selection of data, authorship practices, multiple publications, and the like. [Schachman and Yamamoto 1992:180]Howard Schachman is the ombudsman of the National Intititutes of Health, and he has been a continuing critic of federal efforts to regulate and define misconduct in science.
The NSF Definition Defended. In the face of such criticisms, Donald E. Buzzelli of the Office of the Inspector General at NSF pointed out that the "other serious deviations" clause is not a third category of misconduct that could be excised (as "deception" had been excised from the PHS definition). Rather, the NSF definition of misconduct in science is "serious deviation from accepted practices" and fabrication, falsification, and plagiarism are examples of serious deviations (Buzzelli 1993:222). Buzzelli argues that the NSF has to have an open-ended definition because the standards are set by the scientific community, not by the government. "The agency cannot put out a regulation telling the community what the community's own standards are" (Buzzelli 1993:223).
Buzzelli acknowledges the fear that a wide definition could lead to abuse. "The scientific community has a right to be assured that federal agencies are not going to create misconduct in science cases based on some scientist's perfectly innocent behavior. . . . [T]his assurance should be sought in the procedures that agencies use, more than in definitions." (Buzzelli 1993:224)
But he also muddied the waters: "Misconduct in science comprises all those actions, and only those actions, that the scientific community would regard as making someone a bad scientist" (Buzzelli 1993:224). This glosses over the important difference between a bad scientist and an unethical or immoral scientist.5
The Commission on Research Integrity (the Ryan Commission). Ongoing dissatisfaction with the state of affairs led in 1993 to Congress forming the Commission on Research Integrity (CRI, sometimes called the Ryan Commission after its chair, Kenneth J. Ryan) to "advise the Secretary of Health and Human Serices and Congress about ways to improve the Public Health Service (PHS) response to misconduct in biomedical and behavioral research receiving PHS funding" (CRI 1995:vii).
One of five "principal issues" addressed by the Commission was:
The definition of research misconduct: How narrow or broad should the federal definition be? Specifically, should it include other misconduct beyond fabrication, falsification, and plagiarism? How should questions about intent and honest differences in interpretation of data be addressed? How should the line be drawn between serious and less-serious offenses? [CRI 1995:3, emphasis in original]The definition proposed by CRI:
Research misconduct is significant misbehavior that improperly appropriates the intellectual property or contributions of others, that intentionally impedes the progress of research, or that risks corrupting the scientific record6 or compromising the integrity of scientific practices. Such behaviors are unethical and unacceptable in proposing, conducting, or reporting research, or in reviewing the proposals or research reports of others.The CRI proposal also identifies as "other forms of professional misconduct" "obstruction of investigations of research misconduct" and "noncompliance with research regulations" (CRI 1995:14).Examples of research misconduct include but are not limited to the following:
Misappropriation: An investigator or reviewer shall not intentionally or recklessly
a. plagiarize, which shall be understood to mean the presentation of the documented words or ideas of another as his or her own, without attribution appropriate for the medium of presentation; or
b. make use of any information in breach of any duty of confidentiality associated with the review of any manuscript or grant application.
Interference: An inverstigator or reviewer shall not intentionally and without authorization take or sequester or materially damage any research-related property of another, including without limitation the apparatus, reagents, biological materials, writings, data, hardware, software, or any other substance or device used or produced in the conduct of research.
Misrepresentation: An investigator or reviewer shall not with intent to deceive, or in reckless disregard for the truth,
a. state or present a material or significant falsehood; or
b. omit a fact so that what is stated or presented as a whole states or presents a material or significant falsehood.
Free scientific inquiry naturally includes proposing hypotheses that may ultimately prove to be false, offering interpretations of data that conflict with other interpretations, and making scientific observations and analyses that may prove to be in error. The Commission's recommendations pose no threat to such inquiry, which is essential to the advancement of science. [CRI 1995:13-14, emphases in original]
NAS, the Federation of American Societies for Experimental Biology (FASEB), Howard Schachman, and others quickly and vehemently condemned the CRI definition.
The CRI definition is on hold. President Clinton charged the National Science and Technology Council with preparing a common federal definition of misconduct in science (as CRI proposed and IGRIM endorsed), but that definition has not, to my knowledge, been made public.
Reflections on the Ongoing Controversy. I see four related, and sometimes overlapping, reasons for the ongoing controversy: (a) an overly simplistic assumption that FF&P are clear-cut; (b) a (usually unstated) disagreement over what it means to have a federal definition of misconduct; (c) ongoing denial and defensiveness on the parts of some leaders in the scientific community; and (d) distrust of authority.
An overly simplistic belief that FF&P are clear-cut. I cannot think of an example of "fabrication" in science, as defined, that would not be a case of serious misconduct, but there are certainly cases of "falsification" and "plagiarism" that few people would consider serious misconduct. Arguably, any time a scientist throws out data points ("outliers"), she is falsifying her data (defined by COSEPUP as "changing data or results").
Similarly, it is generally agreed that granting or accepting honorary authorship is not acceptable, but it also is not generally considered serious misconduct warranting federal investigation. But is not "plagiarism" to put your name on a publication to which you made no substantial contribution? (COSEPUP: "plagiarism is using the ideas or words of another person without giving appropriate credit.")
The fact is that all definitions and terms require interpretation; no definition of misconduct can possibly be crafted that will not ultimately depend on human judgment. The "other serious deviations" clause would certainly be vague to someone who does not know the standards of the field, or could be construed as vague by someone maliciously trying to frame a scientist. But clearly people who are ignorant about the standards of science or who are on a vendetta could not be relied upon to render just decisions in misconduct cases no matter what definition were used.
It is also interesting to note again that a fourth category of misconduct -- "deception" -- was excised from the original PHS definition. In most kinds of scientific research, deception would be considered serious misconduct (indeed, FF&P are all sub-categories of deception). I wonder whether the objection -- that some kinds of "deception" are acceptable research practices -- stands up to scrutiny. Placebo studies are often referred to as deceptive, but subjects in placebo research are always told that they may get the drug being tested or the placebo; they are not deceived, they are merely kept ignorant. Likewise, in psychological studies involving deception, the deception is temporary -- the true purpose of the research is revealed after the experiment. I cannot think of any research in which deliberate and uncorrected deception is ethical (I would be most grateful to be illuminated on this point if I am incorrect), but I can think of lots of examples of research in which deception is certainly unethical.
It seems as if FF&P are actually used as proxies for "serious." There are certainly examples of FF&P that are not serious -- e.g., failing to properly cite one quotation in a publication when the same source is cited elsewhere in the publication. (Although technically plagiarism, this kind of thing can easily come about due to mistakes in editing and would generally not be considered serious misconduct.) There are also clearly behaviors other than FF&P that can be serious (including deception, authorship abuse, etc.).
A (usually unstated) disagreement over what it means to have a federal definition of misconduct. Do we want the definition of misconduct to be a statement of what the scientific community will not tolerate? If so, things like abuse of authorship, bad data management practices, and sexual harrassment probably should be included. Or do we want the definition only to serve as the trigger for federal investigation for the purposes of protecting investments of governmental funds? In that case, serious fabrication and falsification should clearly be covered, but abuse of authorship should not.
In any case, why should plagiarism be covered? In what way is plagiarism a waste of research funds? In what significant way does plagiarism threaten the integrity of the research record? After all, one of the hallmarks of science is supposedly the fact that it does not matter who makes a discovery; what matters is that the discovery is true. Plagiarism is important because of the politics of science (it is important to get the credit that is your due, and no more), not because it results in bad data.
Arguably, poor data management is also significant misconduct; it is probably a bad investment of scarce research dollars to fund a scientist who is so sloppy that no one else can work from her notes, or whose experimental results are questionable.
There is only a hazy boundary between law, regulation, and ethics. Strictly speaking, the federal definition of misconduct in science belongs firmly under the category of "regulation." But clearly anyone who is accused of misconduct will take it as a slur upon his moral character -- on his ethics -- as will most of the people involved in the investigation. And as we have seen, violations of this regulatory definition often lead to civil litigation. This slipperiness makes defining misconduct in science very problematic.
Ongoing denial and defensiveness on the part of some leaders in the scientific community. Everyone agrees that serious misconduct is rare, but some seem to think that it is nonexistent. This is especially evident in the statements by the FASEB. Recall the statement by FASEB President Ralph Bradshaw in his criticism of the proposed CRI definition: "When a journal declines to publish a scientist's paper, does that constitute interference? . . . It certainly does interfere with a scientist's career" (Goodman 1996). But the CRI definition of interference cannot possibly be construed to cover an instance like this. (Perhaps the legacy of FF&P has reared its ugly head here; Bradshaw is apparently one of the scientists who thinks that FF&P have a clear-cut, common-sense, intuitive meaning, and he generalizes this mistake to "interference.")
Distrust of authority. Americans have an uneasy relationship with authority and government, an unease clearly and famously enshrined in our Constitution. If we have to have a government, we want the smallest possible one, and the one that can do the least harm. Associated with this are a fear of unwarranted intrusion and a fear that the federal goverment will stifle creative science.
To my knowledge, none of the findings of misconduct challenged to date has relied on the "other serious deviations" clause -- and I feel sure that if one had, it would have made the headlines. FASEB would have seen to it. This does not mean that such fears are utterly unfounded, of course, but Buzzelli's suggestion that the procedures are really more important than the definition rings true to me. Bad procedures cannot be salvaged by a good definition, and an imperfect definition (which is all we can hope for) can be salvaged by well-designed procedures. At times all of the fuss over the definition seems like a smokescreen.
It is ironic that so many leading American scientists do not trust the American government -- the very source of their livelihood. At times one gets the sense that some of these scientists feel that federal funding is a birthright, not a privilege; and at times it seems that some vocal critics of federal policy have not come to terms with the sometimes difficult relationship that now exists between science and the government. It is a relationship that creates responsibilities on both sides, with the government accountable for the efficient use of taxpayers' money and the scientific community responsible for addressing concerns raised by the public and the government.
Finally, from my point of view, much of the distrust of governmental intervention seems contrived. As mentioned above, FASEB's Ralph Bradshaw seemed to willfully misinterpret the proposed CRI definition; Howard Schachman's expressed belief that simple errors "risk corrupting the scientific record," but are not the result of misconduct (Goodman 1996) similarly ignores the passage in the CRI report noting that "scientific inquiry naturally includes . . . making scientific observations and analyses that may prove to be in error" (CRI 1995:14).
I do not intend to defend the CRI definition; the test of a good definition of misconduct in science would be the test of time, as the CRI report notes. Certainly there might be legitimate concerns with the proposed definition. But the ones quoted above are not legitimate because they are not concerns with the actual definition, but with strange and creative interpretations of the definition concocted by Bradshaw and Schachman.
Conclusion. Not all scientists have participated in the debate over the definition of misconduct, and many who have have supported efforts to devise a better definition than FF&P. Most of the scientists I know are deeply concerned about the responsible conduct of research and eager to train their students accordingly. (Of course, this is the kind of scientist I am most likely to meet, given my work on the TRE project.) I find the kind of broadsides fired by FASEB and Schachman deeply regrettable. Leaders in the scientific community should work with the government to improve any faults in the system, not caricature attempts at improvement. But I think that the issue is not one of personalities; I think that the controversy over the federal definition of misconduct in science reflects some systematic challenges facing science in the United States today.
As the stakes in science continue to rise, with increasing competition between labs for fewer research dollars, scientists will continue to have greater temptations to cut corners, and the scientific community will tend to be more protective of its diminishing autonomy. "Outsiders," like federal regulators and university administrators, who try to ensure the integrity of the research process will be looked upon with suspicion and hostility. We will continue to hear claims from some scientists that incidents of misconduct are vanishingly rare and that science is self-policing. In the end, the controversy over the definition of misconduct in science is not, at its heart, a controversy about ethics or about science; it is about politics and power. I doubt that constant wrangling for grant money is good for science; it is, however, the world in which we now live.
2. Early in this process, the phrase "misconduct in science" or "scientific misconduct" replaced "scientific fraud" because "most legal interpretations of the term "fraud" require evidence not only of intentional deception but also of injury or damage to victims. Proof of fraud in common law requires documentation of damage incurred by victims who relied on fabricated or falsified research results. Because this evidentiary standard seemed poorly suited to the methods of scientific research, "misconduct in science" has become the common term of reference in both institutional and regulatory policy definitions" (COSEPUP 1992:25).
3. Steneck 1994 is a good overview of the history of the efforts to define misconduct and implement policies regarding the investigation of misconduct in science.
4. "Questionable research practices are not equivalent to misconduct in science, and they are not appropriate subjects for investigations directed to misconduct." (COSEPUP 1992:29).
5. Buzzelli identifies "some of the major questions that must be answered if one tries to define misconduct in science" as follows: (1) "What makes any action misconduct in science?" (2) "What is the relation between a formal definition of misconduct in science and the standards of conduct held by [the] scientific community?" (3) How far does "the government's responsibility with regard to misconduct in science" extend? (4) Is the "other serious deviation from accepted practices" clause overly vague? (Buzzelli 1993:220-221).
6. The record encompases any documentation or presentation of research, oral or written, published or unpublished. [note in original -- KDP]
COSEPUP (Committee on Science, Engineering, and Public Policy). National Academy of Sciences. 1992. Responsible Science: Ensuring the Integrity of the Research Process, Vol. 1. Washington, DC: National Academy Press.
Integrity and Misconduct in Research. 1995. Washington, DC: U.S. Department of Health and Human Services. Available from the Division of Policy and Education, The Office of Research Integrity, 5515 Security Lane, Suite 700, Rockville MD 20852; (301) 443-5300; FAX (301) 443-5351; requests@osophs.dhhs.gov.
Goodman, Billy. 1996. "Scientists are Split Over Findings of Research Integrity Commission." The Scientist 10(2):1, 8-9 (http://www.the-scientist.library.upenn.edu/yr1996/jan/conduct_960122.html).
IGRIM (Implementation Group on Research Integrity and Misconduct). 1996. "Implementation Proposals on Recommendations by the Commission on Research Integrity." Unpublished report available from the Office of Reseach Integrity (see Integrity and Misconduct in Research, above).
NSF (National Science Foundation). 1991. "Misconduct in Science and Engineering." Federal Register 56 (May 14):22286-22290.
PHS (Public Health Service). 1989. "Responsibilities of Awardee and Applicant Institutions for Dealing with and Reporting Possible Misconduct in Science." Federal Register 54 (August 8):32446-32451.
Schachman, Howard K., and Keith R. Yamamoto. 1992. "Minority Statement." In Responsible Science: Ensuring the Integrity of the Research Process, Vol. 1, pp. 180-181. Washington, DC: National Academy Press.
Steneck, Nicholas H. 1994. "Research Universities and Scientific Misconduct:
History, Policies, and the Future." Journal of Higher Education
65 (3):54-69.