NEWS FOCUS
Ethical Issues Underlying Responsible Conduct of Science Explored
Forum concluded ethical behavior needs to be communicated and practiced at all levels of research, urged training in ethics for all science students
Pamela S. Zurer, C&EN Washington
Two days isn't much time to wrestle with the pressing ethical issues facing science today.
But some 400 scientists, engineers, and research administrators attending Sigma Xi's forum on "Ethics, Values, and the Promise of Science" last month worked diligently to do so. They joined philosophers, historians of science, and government officials in assessing scientists' obligations both to themselves and to the society that supports their work.
Sigma Xi, the scientific research society composed of chapters and clubs across the nation, sponsored the meeting in San Francisco in an attempt to involve the scientific community in debate on key ethical issues. One goal of the forum was to recommend actions scientists can take to help restore public confidence in the way research is conducted.
Some speakers viewed the challenges facing science as stemming only from external pressures. In his address to the forum, Nobel Laureate J. Michael Bishop restated a theme that has become familiar within the community during the past decade: Science is underappreciated, underfunded, and under attack from a public unschooled in its basics. Bishop is professor of microbiology, immunology, biochemistry, and biophysics at the University of California, San Francisco.
"Fear, bewilderment, disdain: these are all opponents science must best," Bishop said. "And there is one other which is now current: mistrust The public is ignorant of the formula by which science advances, and hence is easy prey for our critics."
Bishop's lament, echoed by a few other speakers, found many sympathetic listeners at the forum. Yet the
Gert: scientists react like sports fans
stir were those that challenged the re-. search community to look within and
examine its own principles and behaviors.
A philosopher confronted researchers with their illogic in holding established scientists to lesser standards than junior members of the community. A medical ethicist countered skepticism that ethics can or should be taught. And a panel of postdoctoral researchers described treatment by supervisors that could be viewed as discouraging at best.
"With 2 million members, it's unlikely that the moral character of scientists and engineers is different from that of other groups of people, such as doctors, lawyers, or philosophers," said Bernard Gert, professor for the study of ethics and human values at
Dartmouth College in Hanover, N.H. Gert was the only philosopher to serve on the National Academy of Sciences (NAS) panel that last year produced the report "Responsible Science: Ensuring the Integrity of the Research Process." Everyone agrees, Gert said, that certain acts such as deceiving, cheating, and neglecting one's duties are immoral un
less one has adequate justification. "Some scientists may claim that . . .
some kinds of misreporting of their experiments, in order to enhance the acceptability of an hypothesis of whose correctness one is very confident, is justified," Gert said. "They may hold that if experienced scientists are very confident of their claims, those claims are usually true, so that such enhancement will result in less time being wasted doing futile research. Thus, they may claim that this kind of deception actually results in more truth being discovered than failure to deceive. It may be, in the words of [the NAS report] a questionable practice, but it is not scientific misconduct.
presentations that caused the greatest Lo: ethical guidelines are insurance "I do not claim that as presently un-
MARCH 15,1993 C&EN 7
NEWS FOCUS
Government should help promote scientific integrity, committee advises The federal government has a role to play in promoting the responsible conduct of research, according to the Department of Health & Human Services' Advisory Committee on Research Integrity. The government should provide information, support scholarly research on ethics, and prod universities to take concrete steps to foster integrity in science, the panel concluded at a meeting last month.
Advising the government to get more involved in the affairs of aca-demia is something of a switch for the group. In the two years of its existence, the committee—formally charged with advising the Secretary of Health & Human Services on issues of research integrity—has for the most part focused on limiting government's role in dealing with scientific misconduct.
For example, the 10-member panel's first efforts focused on narrowing the Public Health Service's (PHS) definition of scientific misconduct (C&EN, March 23,1992, page 14). Its aim was to lessen the anxiety of the scientific community, many of whom fear the current definition as vague and overly broad. PHS intends to formally propose a revised definition based on the
committee's recommendations later this year.
Last fall, the committee asked PHS's Office of Research Integrity (ORI) to abandon its application for an exemption from the federal Privacy Act (C&EN, Nov. 2, 1992, page 19). ORI, which investigates allegations of scientific misconduct, wants the exemption so it can protect the identities of whistleblowers and other confidential informants. In this instance, however, PHS has chosen not to accept the panel's advice and is proceeding with its application.
The committee held its sixth meeting in San Francisco on Feb. 27 and 28, immediately following Sigma Xi's forum on ethics. Although the group in previous sessions has concentrated on reviewing ORI's policies and procedures, this time the panel focused on what committee chairman Nicholas H. Steneck described as "the more difficult part of our charge: providing advice on strategies for promoting integrity in science." Steneck is professor of history and director of the Historical Center for the Health Sciences at the University of Michigan, Ann Arbor.
Research integrity is not the same thing as research fraud or misconduct,
Steneck told the committee, but encompasses behavior the group did not want included in the definition of scientific misconduct. "Integrity means living a complete professional life," he said. "It brings in issues of sloppy science, inadequate mentoring or laboratory supervision, inadequate citations, and unwillingness to cooperate with colleagues."
Steneck said there is ample room for improvement. "We don't know the frequency of scientific misconduct but we do know a fair amount about problems with the integrity of biomedical research. . . . There is bias in reviewing. Researchers are pretty sloppy in some things, such as use of statistics. It may not be misconduct but it is research lacking integrity."
Education can address issues of scientific integrity, he continued. "We can either foster proper values or turn them off in our training of researchers. . . . The goal is to help researchers be more knowledgeable about proper handling of data, sharing information, and the social implications of their research. I'm a believer that education can make a difference."
But Steneck saw little evidence that universities will provide such educa-
derstood [such deception! would count as scientific misconduct; that is, fabrication, falsification, or plagiarism," Gert continued. "But I hold that from the point of view of morality, the only difference between many questionable practices and scientific misconduct is that the former are fairly widespread and the latter fairly rare."
As an example, Gert discussed the attitude of some scientists toward Robert A. Millikan's selective reporting of results of his famous oil-drop experiments. Gert characterizes their view as "consequentialism," the idea that all that counts morally are the consequences of one's actions.
As described in American Scholar [60, 505 (1991)], Milli-kan—after being challenged by a rival on his unit theory of electric charge—published a paper in which he says he presented all his data from 60 consecutive days. But Millikan's notebooks show that he omitted data from experiments that did not turn out the way he expected.
The author of the American Scholar paper, David Goodstein, vice provost of California Institute of Technology, says Millikan's omissions were motivated by the need to convince skeptics of what Millikan perceived to be the scientific truth. Goodstein says that "inestimable damage to science would have been done" had Millikan not succeeded.
Gert disagrees. "How does [the consequentialist] know that if . . . Millikan had been honest, admitting that [his] re
sults were not perfect, that others might not have acted in such a way as to bring about the correct results more quickly? How does he know that these acts of honesty under great temptation to deceive might not have so influenced the scientific tradition that science would be even more successful than it is now?"
The tendency to hero worship great scientists, together with loyalty among members of the profession, may explain the failure to recognize that great scientists sometimes behave immorally, Gert said. "Some scientists seem to react very much like sports fans. If a scientist or player is great enough, then almost nothing they do counts as seriously wrong. . . . I do not want to inhibit any scientist from exercising his scientific judgment; all I want is that he not deceive anyone about what he is doing."
"Imagine teaching a graduate student to exercise his scientific judgment by keeping secret all unwanted results, publishing only those that confirmed his hypothesis, and claiming that he is publishing everything," Gert continued. "If this is the way in which graduate students in science are being taught, it is amazing that scientific fraud is not far more common than it seems to be. But I suspect that the students are made to understand, though not explicitly told, that this kind of behavior is not acceptable for them—it is only acceptable for someone who has already attained stature in the field."
8 MARCH 15,1993 C&EN
tion on their own. Only since the National Institutes of Health (NIH) imposed a requirement that all recipients of research-training grants receive a grounding in ethical issues have some universities set up courses or symposia. "My reluctant conclusion is that if left on their own, [universities] would continue to make general statements about the importance of integrity, but there would be limited concrete action," he said. "There is a role for government in this area."
Surprisingly, Steneck's views met no opposition from the other committee members present. However, Estelle A. Fishbein, vice president and general counsel of Johns Hopkins University, who at earlier meetings argued forcefully against any expansion of government into the university's domain, did not attend the San Francisco meeting. In her absence the committee turned to the question of what the government could reasonably do.
The panel members decided the requirement for education in research ethics should be extended to all PHS-funded research. They advised ORI to publicize successful ethics courses in its new quarterly newsletter and to suggest—but for the time being not re
quire—such training at all institutions receiving PHS funds.
The group also endorsed the idea that principal investigators on grants be required to complete a checklist on matters relating to scientific integrity when applying for grants or submitting reports. Suggested by Eleanor G. Shore, dean for faculty affairs at Harvard Medical School, such a checklist could include assurances that the investigator had formulated an appropriate plan for retaining primary data; all coinvestigators had been consulted in drafting proposals, reports, or publications; trainees had received training in research ethics; and attributions and citations had been verified.
"Faculty time spent on a checklist pales beside the time and effort of an inquiry and investigation" into scientific misconduct, Shore said. "A checklist reaches everyone, from the most junior to the most senior, with the notion there will be accountability if problems arise."
Noting that most current research on ethical issues in science is supported by the National Science Foundation, Steneck posed the question of whether PHS should set aside some funding specifically for research on questions
of scientific integrity. He cited as an example NIH's Center for Human Genome Research, which reserves 3% of its $110 million annual budget for studies of ethical issues raised by the human genome project.
The committee backed the idea. "I would hope ORI would develop a research program with requests for proposals saying what sort of questions it would like to address, such as the prevalence of misconduct," said Paul J. Friedman, dean of academic affairs at the school of medicine at the University of California, San Diego.
Currently, ORI has no funds for underwriting such research, noted Lyle W. Bivens, ORI's acting director, but could incorporate the idea in its plans for next fiscal year's budget. He pointed out that PHS sets aside 1% of its total budget each year for program evaluation, suggesting these monies might prove to be a source for supporting research on scientific integrity.
The panel recommended that ORI include ethics research programs in its next budget. In the meantime, the committee advised that ORI make known its interest in such work in its newsletter and challenge outside groups to support ongoing research.
Gert's presentation seemed to leave many scientists in the audience uncomfortable—including Goodstein, who the day before had spoken persuasively on scientists' responsibility to educate the public. Not surprisingly, when Goodstein asked for a show of hands from all who considered themselves "hero-worshipping consequentialists," no one responded.
Bernard Lo, associate professor of medicine at UCSF, offered rebuttals to the objections of skeptics who doubt ethics can or should be taught. Lo heads a program in medical ethics that holds seminars for trainees on ethical issues in the conduct of biomedical and clinical research. Many of his colleagues, he told the forum, wonder whether ethical issues are worth the time, preferring to stick to their labs.
For example, Lo said he often hears: "Ethics is being a good person, not a system of rules or set of guidelines." He responds, he said, by pointing out that being a good person isn't always enough to act properly. "Scientists may be genuinely perplexed in a given situation; for example, how to respond to an allegation of misconduct. Scientists are human; they make mistakes and work under stress. Ethical guidelines are like insurance, protecting against scientists whose character is flawed, who suffer lapses in character, or who are overwhelmed by a difficult situation."
"Ethics is common sense and experience" is another objection to ethics training Lo regularly faces. He replies that
many ethical dilemmas are so complicated that sensible and experienced scientists may be perplexed or disagree over what to do. Ethical guidelines can also help scientists make their thinking explicit and help them explain their reasoning to students and the public.
Another objection Lo hears is that "Every case is unique, so guidelines are impossible." He counters that general guidelines obviously cannot be applied mechanically. "We'll always need practical knowledge, judgment, and compassion. But certain situations—such as problems with missing primary data—come up repeatedly and ought to be resolved in consistent ways if we are to be fair."
Students learn best about ethics when their interest is captured by examples from real life and they actively take part in thinking issues through for themselves, Lo said. In seminars at UCSF, research trainees discuss realistic cases and participate in role playing that leads to developing specific suggestions for dealing with the problem at hand. Lo also tries to involve senior scientists, who by sharing their own experiences legitimize discussion about ethical issues.
Lo recruited from his institution two of the four postdoctoral researchers who took part in a panel discussion of their uniquely vulnerable place in the academic research system. Two more—both chemists—were drafted by Stanford University chemistry professor Carl Djerassi. The forum's organizers were puzzled by how difficult it was to
MARCH 15,1993 C&EN 9
NEWS FOCUS
Jones and Meyer: postdoctoral scientists occupy precarious positions
get postdocs to come forward and describe their experiences—but the panel members were not surprised.
"It's pretty much thought to be professional suicide to stand up and speak about a bad experience you've had/' said Jan Gurley, a physician participating in the Robert Wood Johnson clinical scholars program at UCSF. She described postdoctoral scientists' dependence on their bosses as at best apprenticeship and at worst indentured servitude or slavery.
Postdocs are often first to find out when something is wrong with a thesis, reagent, or experiment, Gurley pointed out. Disappointed research supervisors' reactions all too often fit the "kill-the-messenger syndrome," with the postdoctoral scientists blamed or shunned.
Lisa Backus opted for medical school at UCSF rather than postdoctoral work after getting her Ph.D. degree in neuropharmacology in large part because of issues of mentoring and gender, she told the forum. "I can't overemphasize how important the issue of mentoring is," she said.
Mentors must teach young researchers to do good science, write papers, and get grants, Backus said, but they also must steer beginners through the politics of modern science. "He—and it almost always is he—must look out for your career. He must help you publish in the right place, meet the right people at conferences, and make calls to get you into the old-boys' network."
"It's difficult to find a principal investigator who will do all of this," she continued, "and issues of gender make it even harder. Men feel more comfortable mentoring other men." The lack of mentors has had noticeable negative effects on women's careers, she said.
Tara Meyer, a postdoctoral researcher in chemistry at UC Berkeley, said her own experience has been positive. Nevertheless, she noted that postdocs occupy a uniquely precarious perch in the typical feudal hierarchy of academic institutions. Having no separate relationship with the university, they are totally dependent on their research supervisors.
Meyer outlined what she sees as the obligations of research advisers toward their postdoctoral students—obliga
tions she believes are largely understood but rarely discussed. Among those she listed are that postdocs must not face discrimination or sexual harassment. Their advisers need to give them a realistic picture of the nature of the project they will be working on and provide sufficient materials and equipment so they can perform the work.
Postdoctoral duties to the research group as a whole—such as supervising graduate students or ordering supplies— should be kept under control so that the postdocs will have time to carry out their own research. Their work should be properly acknowledged, and they should be given encouragement and help in finding jobs, she continued.
Meyer's husband, Garth Jones, a postdoctoral chemist at Stanford, described the difficulties in finding postdoctoral work. "Both academic and industrial institutions want new employees to have the breadth of experience a postdoc provides," he said,
yet there is no systematic way to discover what opportunities exist.
Many institutions, fearing repercussions for failing to comply with equal employment opportunity regulations, never advertise available openings, Jones said. Jobs are found through the old-boy network or through the luck of contacting the right place at the right time. "There are endless possibilities for unfairness in hiring," he said.
After two intense days of thought-provoking speeches and spirited discussion groups, the attendees developed a set of conclusions and recommendations for the scientific community. Addressing issues ranging from teaching ethics to peer review to improving mentoring to the ethics of diversity, many of the recommendations assigned to scientists themselves the responsibility for improving both the practice of science and the public's perceptions.
Among other conclusions, the group decided that appropriate ethical behavior needs to be communicated and practiced at all levels of academic, governmental, industrial, and other research organizations. They also strongly recommended that training in ethics be required for all science students.
Another conclusion held that significant and widespread problems exist in mentoring relationships, while at the same time there is widespread reluctance to address those problems. Sigma Xi was urged to set up a committee to assess the situation and to define an ideal mentor.
And the group agreed that scientific misconduct is often an outgrowth of mismanagement or lack of proper supervision. In responding to allegations, institutions should also seek to ascertain why and how misconduct in research was allowed to occur.
Sigma Xi plans to publish the proceedings of the forum and its conclusions and recommendations by early summer, according to the society's executive director, John F. Ahearne. He described the forum as a first step in a major effort by Sigma Xi in the area of ethics and values. The society will form a steering committee to decide what to take on next. •
10 MARCH 15,1993 C&EN