»Foundation Text«
Table of Contents
- What is scientific misconduct?
- What policies and procedures are in place to deal with misconduct?
I. What is scientific misconduct?
A. Office of Science and Technology Policy definition
"Advances in science, engineering and all fields of research depend on the reliability of the research record, as do benefits associated with them in areas such as health and national security," begins the December 6, 2000, Notice of Final Policy in the Federal Record from the Office of Science and Technology Policy in the Executive Office of the President. It continues, "Sustained public trust in the research enterprise requires confidence in the research record and in the processes involved in its ongoing development."
To ensure research dependability, to maintain public support, and to unify policies across the federal government, the OSTP on that day issued, after an almost 10-year process, what it calls a unified definition of misconduct. The policy applies to all agencies and all recipients of funds from federal agencies. The OSTP's goal was to have all agencies adopt this policy. The two federal agencies that support most research in academia - the National Science Foundation and the Public Health Service, which includes the National Institutes of Health - have implemented the OSTP policy, for the most part.
What is the OSTP's definition of misconduct? According to the OSTP, "research misconduct is defined as fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results. Fabrication is making up data or results and recording or reporting them. Falsification is manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record. Plagiarism is the appropriation of another person's ideas, processes, results, or words without giving appropriate credit. Research misconduct does not include honest error or differences of opinion."
"Research," according to the OSTP, "includes all basic, applied, and demonstration research in all fields of science, engineering, and mathematics. This includes, but is not limited to, research in economics, education, linguistics, medicine, psychology, social sciences, statistics, and research involving human subjects or animals. The research record is the record of data or results that embody the facts resulting from scientific inquiry, and includes, but is not limited to, research proposals, laboratory records, both physical and electronic, progress reports, abstracts, theses, oral presentations, internal reports, and journal articles."
The policy also addresses how research institutions should manage allegations of misconduct, which this module will address later. At the NSF, the Office of Inspector General has oversight of institutions regarding misconduct.
The Office of Research Integrity (ORI) oversees and directs PHS research integrity activities on behalf of the Secretary of Health and Human Services (HHS), with the exception of the regulatory research integrity activities of the Food and Drug Administration (FDA). The Centers for Disease Control, the Agency for Healthcare Research and Quality, and the Agency for Toxic Substances and Disease Registry all fall under the PHS aegis.
Historical examples of fabrication, falsification, and plagiarism will be discussed next to illustrate examples of misconduct. Today, plagiarism is deemed to be a serious problem facing academia. Some examples of plagiarism include: investigators during the peer- review process taking ideas from others' grant proposals or articles and including them in their own publications; students taking material from the Internet verbatim, without attribution, during write-ups of research; and faculty taking dissertation material from students and including it in publications without giving due credit.
B. What is the history of scientific misconduct regulation?
1. Examples from history
Throughout the history of science, there have been many claims that scientists have fudged data, which today might be deemed misconduct. Michael Kalichman, of the University of California, San Diego, points to a few examples: Isaac Newton may have adjusted calculations to fit observations; Gregor Mendel's results with pea plants were cleaner than what is observed experimentally, indicating that he might have changed the data; and Robert Millikan, in a research paper describing the charge of an electron, failed to mention that he eliminated some data points although he probably should have reported why he removed some data in the publication. Others have noted that Louis Pasteur failed, in his studies, to cite that he used the vaccine against anthrax made by a competitor, saying instead that he used his own vaccine to inactivate the bacilli by oxygen.
2. William Summerlin
More recently, certain scientific misconduct cases have caught the public's attention after being reported in the news media. In 1974, William Summerlin, working with immunologist Robert Good at the Sloan-Kettering Institute for Cancer Research, reported that he could transplant tissue from genetically unrelated animals without rejection by the recipient animal if he kept the tissue from the donor in organ culture for four to six weeks. Such a finding would have been invaluable for transplant medicine. Summerlin demonstrated his claims by showing white mice that had black patches on their backs because their skin had been transplanted from unrelated donor mice. However, after Summerlin made a presentation in Good's office, a technician noticed that these "transplanted patches" were actually drawn on the skin of the mice with a felt-tipped marker. The patches could be removed with alcohol. When confronted, Summerlin admitted the fraud.
3. John Darsee and Robert Slutsky
Two other cases in the 1980s brought awareness of misconduct to the scientific community and to the public. One case involved Dr. John R. Darsee, a young clinical investigator in cardiology at the Brigham and Women's Hospital (a teaching affiliate of Harvard University), who published many research studies and included faculty members as authors on the work although they had minimal or no direct involvement in the papers that bore their names.
In May of 1981, Darsee's associates and supervisors at Harvard caught him fabricating data. Other investigations led to the conclusion that Darsee fabricated research publications beginning when he was a biology student at Notre Dame, continuing through his medical residency and cardiology fellowship at Emory University, and ending at Harvard. More than 10 primary journal articles and more than 45 abstracts were retracted as a result of the investigations. The Darsee case highlighted the problem of using so-called "gift authors" on papers. Gift authors have little or no contact with the person giving them the authorship, but the presence of the prestigious names may influence an editor to consider a paper for publication.
The co-authors in the Darsee case either did not know they were on the papers or had no direct involvement in the papers. Eventually, they were found not to have participated in the fraud, having been unaware of the fabrications. But the case pointed to the need for co-authors to have responsibility for information in papers on which their names appear. The case also revealed that the peer-review process could not be relied upon to find the fabrication.
Soon after Darsee, the case of Robert A. Slutsky, working in radiology at the University of California, San Diego, also brought to light the problem with co- authors. Slutsky apparently was publishing one paper every 10 days for years and including names of many co-authors to mislead editors and cover up for what later was learned to be a false output. His behavior continued until 1985. At the time, he was up for promotion, and a reviewer noted duplicated data in two of Slutsky's articles. Slutsky resigned, but an investigation later found that, of his 137 publications, 77 were valid, 48 were questionable, and 12 were fraudulent.
4. Thereza Imanishi-Kari, David Baltimore, and Margot O'Toole; hearings and regulations
This infamous case pitted a postdoctoral researcher against a Nobel Prize winner. It became a cause célèbre in the mid-1980s, and was the subject of federal investigations and congressional hearings, which ultimately led to the establishment of regulations against research misconduct.
The case took more than 10 years to resolve. It began in 1985, when Margot O'Toole, a postdoctoral research at Tufts University, could not validate the findings of some work soon to be published in Cell by her employer, Thereza Imanishi-Kari. , A co-author on the paper was a Nobel Prize winner, David Baltimore, of the Massachusetts Institute of Technology. O'Toole came to believe that Imanishi-Kari manipulated research data and presented her evidence to committees at Tufts and MIT. The committees ruled that misconduct had not occurred but that Imanishi-Kari had been sloppy in her research. An NIH committee at the time also cleared Imanishi-Kari.
Walter Stewart and Ned Feder, two scientists from the NIH who were involved in exposing misconduct, then became involved in the case, believing that the committees and the NIH did not do an adequate job of revealing problems with the paper. In 1988, congressional hearings were held, headed by Representative John Dingell, who chaired the Energy and Commerce Committee, which oversees the budget for the NIH. All three took up the cause because they believed that problems existed in the way that universities handled allegations of misconduct.
Dingell's committee even brought in the Secret Service, which determined that Imanishi-Kari had tampered with evidence that she had provided to the committee. The hearings were contentious in nature, and many believed that Dingell was particularly hostile to Baltimore, who continued to defend Imanishi- Kari.
In 1991, the NIH's fraud division, at that time called the Office of Scientific Integrity, found that significant portions of the data were fabricated. Thereza Imanishi-Kiri was banned from receiving federal grant money for 10 years. David Baltimore resigned from his presidency of Rockefeller University on December 31, 1991, saying that he could no longer be an effective leader because of the controversy over his role in the case. He began in July 1990. Margot O'Toole was ostracized by the science field, and at one point, for employment, she answered telephones for a moving company.
The OSI became reconstituted in HHS as the Office of Research Integrity, which upheld the OSI decision. The ORI, however, had an appeals mechanism, called the Appeals Board, which, on June 21, 1996, overturned the ORI decision. The board said that it had not been proved that Imanishi-Kari falsified the scientific record but instead that she was guilty of sloppy scientific practices.
No one emerged untarnished from the case. Soon afterward, regulations concerning the management of misconduct allegations were announced by the NIH and the NSF; they included provisions for protecting the rights of those who report misconduct and those who are accused of misconduct, and also a provision for an appeals process.
5. Jan Hendrik Schön
In October 2002, the physics community was shaken when it became known that some groundbreaking electronics research from Bell Labs was based on fraudulent data. An independent committee found that Jan Hendrik Schön, a Bell Labs researcher, had made up or altered data at least 16 times between 1998 and 2001, and the company terminated his employment.
Thirteen of Schön's disputed findings had been published in Science and Nature, the New Scientist reported. According to physicsweb.org, he produced one paper every eight days, with a total of 20 collaborators, who were cleared of any misconduct by a Bell Labs committee. Bell Labs had formed the committee in May 2002 after some concern was raised about the validity of the research performed by Schön and his collaborators. The experiments in question were in the areas of superconductivity, molecular crystals, and molecular electronics.
"The evidence that manipulation and misrepresentation of data occurred is compelling," the committee wrote, linking all misconduct to one researcher, who committed falsification or fabrication of data on at least 16 occasions, some interrelated. According to the committee's report, the researcher "did this intentionally or recklessly and without the knowledge of any of his co-authors." But an editorial in the New Scientist asked if the collaborators should have been held accountable for Schön's fabrications, since only one of them actually witnessed him doing the research for his so-called groundbreaking experiments.
Also called into question by the editorial was the peer-review system, which failed to note some overt errors in at least one of the manuscripts, such as the use of the same curve to represent the behavior of different materials. But even though the committee said that the collaborators did not commit misconduct, the report discussed the issue of the potential problem of a lack of oversight for co-authors. The committee acknowledged that each co-author may have a different responsibility in a paper, but it raised questions about Bertram Battlogg, who led the research team. The committee said that he met his responsibilities as a co-author but questioned whether Battlogg asked for sufficient validation of the research results, which he, as a senior investigator, knew would ultimately come under scrutiny.
Based on the committee's report and language, the American Physical Society changed its professional guidelines in response to the Schön case. The guidelines now include the following: "While not all coauthors may be familiar with all aspects of the research presented in their paper, all collaborations should have in place an appropriate process for reviewing and ensuring the accuracy and validity of the reported results, and all coauthors should be aware of this process."
C. How common is scientific misconduct?
Given these high-profile cases, one question that is often asked is whether they were anomalies or representative of a real problem. It is difficult to assess whether there has been an increase in misconduct cases during, say, the past 50 years, because data on misconduct was not collected until the 1990s. The rate of overall research misconduct in the United States has been estimated to be one case per 100,000 researchers, given two million active investigators in the United States.
Between 1990 and 2002, the Office of Inspector General at the NSF investigated 800 allegations of misconduct in 600 cases, according to Peggy Fischer, the Associate Inspector General for Scientific Integrity, and others writing in the Fall/Winter 2002 Journal of Public Inquiry. The investigations resulted in 60, or 10%, of these cases being deemed misconduct. Actions ranged from debarment to reprimand, with some recovery of funds. Fischer wrote that she and her colleagues believed that the number of misconduct cases reported to the NSF was lower than the actual number of misconduct cases occurring each year at institutions funded by the NSF. The 50 misconduct cases reported represented .14% of all grant recipients, an estimate that was 100 times lower than the number arrived at from other data. The authors attribute the low number to the institutional practice of resolving allegations without reporting the result of the resolution to the NSF's OIG, even though the institution is required to do so.
In 2002, the ORI reported that 99 institutions had 83 cases of misconduct, with 71 institutions reporting a new allegation. In 1998, the ORI reported that 67 institutions had 54 cases of misconduct, with 41 institutions reporting a new allegation. As at the NSF, concerns also have been raised at the ORI about settlements of cases by institutions without them being reported to the federal government, which thus creates a low estimate of the total number of misconduct cases.
II. What policies and procedures are in place to deal with misconduct?
A. What are the issues surrounding the reporting of a misconduct case?
In 1995, the National Academy of Sciences said that scientists are advised that "someone who has witnessed misconduct has an unmistaken obligation to act" (Columbia University has the same policy). Government regulations require institutions to have systems in place so that individuals are able to report misconduct confidentially and without retaliation. But even though people who report misrepresentation about research may be turned into heroes by Hollywood - as in "The Insider," in which Jeffrey Wigand disclosed nicotine's addictive properties - in reality, whistle-blowers may suffer.
According to a 1995 ORI study on the consequences of whistle-blowing in the scientific community, "the willingness of individuals to allege misconduct is likely to depend on how the system deals with and protects them when they come forth with their allegations. Potential whistleblowers must consider whether the allegation will be taken seriously and the report treated confidentially and whether reporting will provoke retaliation not only from those accused but also from the larger academic and scientific community."
The ORI study found that more than two-thirds of all whistle-blowers reported experiencing at least one negative outcome as a direct result of their whistle-blowing. Those negative experiences included pressure to drop allegations, facing counter- allegations, being ostracized by colleagues, reductions in research support levels, threats of being sued, and actually being sued. Some said that they were fired or were denied promotions. For the whistle-blowers who did not experience a negative consequence as a result of coming forward, more than 90% said that they would do it again. Surprisingly, though, 75% of those who suffered a negative outcome from reporting an allegation of misconduct also said that they would do it again.
Since that time, better systems have been put in place at institutions to protect the rights of those who report misconduct.
B. What are some guidelines for those who report misconduct?
People who report allegations of misconduct are protected by rulings from state governments and the federal government. The First Amendment protects the freedom of speech of someone who reports an allegation of misconduct. Also, the False Claims Act of 1986 protects whistle-blowers. The act, which was originally developed during the Civil War to protect the government from fraudulent contractors, awards a whistle- blower 15% to 30% of the resulting settlement in a case of misconduct. The act also provides for remedies if it can be shown that a whistle-blower suffered a discriminatory action in retaliation for the allegation brought under the legislation. There are also new federal guidelines proposed by HHS to protect whistle-blowers; to date, these have not been finalized.
Even with these regulations and rules, how should a whistle-blower proceed with an allegation? Michael Kalichman, of the University of California, San Diego, provides some guidelines for those who report allegations of misconduct:
- Documentation: When making an allegation of misconduct, clear documentation of who did what, and when they did it, will provide the best chance for a fair and timely resolution of the allegation.
- Rules and procedures: Even though the federal government has minimal requirements about how institutions should handle misconduct, institutions have some leeway in applying the regulations. As soon as someone is involved in an allegation, he or she should review institutional procedures on the issue. A whistle-blower needs to know who should be apprised of the allegation, what constitutes evidence for or against an allegation, how the evidence should be obtained, who will review the allegation, what the whistle-blower's role will be, and how much time the process is expected to take.
- Perspective: People with little experience in research should seek guidance before making allegations of misconduct. What might appear to be a serious action could be a misunderstanding. It might be appropriate to talk to peers, senior researchers in a team, an ombudsperson, or the individual in question.
- Dispute resolution: Some allegations of research misconduct might be resolved through other means, such as conflict resolution. This involves dealing with a problem as soon as possible; striving for an agreement rather than disagreement; emphasizing the problem, not the people involved; and using a third party, such as an ombudsperson, to clarify issues if necessary.
- Motivation of a whistle-blower: Whistle-blowers should be aware that they may suffer retribution for their actions and that institutions are responsible for a misconduct inquiry and investigation. They should also distinguish between facts and speculation and not guess at the motives of others. Whistle-blowers should ask questions rather than draw conclusions.
C. What are the inquiry, investigation, and adjudication processes?
According to the OSTP's 2000 regulations, "federal agencies have ultimate oversight authority for federally funded research, but research institutions bear primary responsibility for the prevention and detection of research misconduct and for the inquiry, investigation and adjudication of research misconduct alleged to have occurred in association with their own institution." At Columbia University's Health Sciences campus, individuals bring their concerns to the Committee on the Conduct of Science. At Columbia University's Morningside campus, individuals bring their concerns to their departments.
1. Definitions of inquiry, investigation, and adjudication
Inquiry is the assessment of whether the allegation has substance and whether an investigation is warranted. Investigation is the formal development of the factual record and the examination of the record leading to dismissal of a case or to a recommendation for a finding of research misconduct or to other remedies. During the adjudication phase, recommendations are reviewed and corrective actions, such as sanctions, are determined.
To determine whether misconduct has occurred, the action must have been committed intentionally or knowingly or in reckless disregard of known practices. The allegation must be proved by a preponderance of the evidence, which means determining whether the claim or fact is more probably true than not true. The standards of "clear and convincing evidence" and "beyond a reasonable doubt" require a much higher burden of proof.
Agencies rely on institutions to make the initial response to allegations of misconduct. Agencies also generally refer to the institutions any allegations of misconduct that are made to them. Occasionally, agencies will perform their own inquiries into or investigations of allegations. Under certain circumstances, agencies may undertake investigations or act quickly to protect the public interest, such as when public health and safety are at stake.
2. Requirements to report to a federal agency
Research institutions have to notify an agency or agencies if the inquiry into an allegation of misconduct involving federally funded research leads to sufficient evidence to proceed to an investigation. When an investigation is complete, the institution also has to forward a copy of the evidence, the investigative report, recommendations made to the institution's adjudicating official, and the subject's written response to the recommendations, if there is one. Institutions also have to inform the agency about the decision of the adjudicating official and about any corrective action.
If, during an inquiry or investigation, there is any immediate risk to public health or safety, research activities should be suspended. If there may be violations of criminal or civil law, or if allegations are made public prematurely, the institution must notify the federal agency.
3. Safeguards for whistle-blowers and subjects of allegations
Institutions must provide appropriate safeguards to those reporting misconduct and those who are subjects of misconduct allegations. People who report allegations of misconduct in good faith must feel that they can report allegations confidentially and without fear of retribution. Subjects of allegations should feel that the filing of an allegation will not stop their research or bring other undue adverse action against them. Subjects also should have their rights protected, with allegations kept confidential, and have timely written notification of the allegations made against them, a description of the allegations, reasonable access to data and other evidence supporting the allegations, and the opportunity to respond to allegations and findings of research misconduct. Appeals processes should also be part of the system.
4. Sanctions
Sanctions against those found guilty of misconduct can range from appropriate steps to correct the research record to letters of reprimand; imposition of certification requirements to ensure compliance with the terms of a grant; suspension or termination of a grant; and/or personal suspension or debarment. If criminal or civil fraud violations have occurred, the agency will refer the findings to the appropriate agencies. The ORI releases the names of people guilty of misconduct on its Web site. Agencies may also issue additional sanctions beyond those of the institution.
Continue to the next section: → Resources