Teaching the Responsible Conduct of Research:
The Why, the What and the How

Panel
Research Credit and Authorship: Graduate Student and Postdoc Case Studies
Vivian Weil, Director, Center for the Study of Ethics in the Professions
Illinois Institute of Technology

Teaching and Learning Research Ethics: Needs and Opportunities Revisited
Judith P. Swazey, President
The Acadia Institute

Stephanie J. Bird, Special Assistant to the Provost
Massachusetts Institute of Technology

Introduction
by: Vivian Weil
Center for the Study of Ethics in the Professions, Illinois Institute of Technology

Here we are, seven years after the last Sigma Xi forum on ethics, values and the promise of science. I was pleased to see the proceedings volume from that forum distributed here, because it provides a kind of benchmark. Since that volume came out, the National Institutes of Health has made it a requirement that grantees who hold training grants have to include ethics in their training, and very recently that requirement has been extended to NIH grants more broadly. So responding to these promptings and to other promptings, as well, the programs, graduate study and graduate schools have undertaken some kind of research ethics training.

We hear reports about lecture series, about graduate student seminars, about survival courses and other kinds of courses that suggest some educators have awakened to the need to incorporate ethics in graduate training. At the same time, we continue to get reports that attention to ethics in graduate programs is thin and sporadic in many places; that the responsibility is handed off to less powerful members of research groups and departments; and that well-meaning people often hesitate to push because they believe they do not know enough about how to proceed. Some may even think that there isn’t enough knowledge about how to proceed.

Now, Judith Swazey on our panel points out that we need more and more current knowledge about the nature and effectiveness of ethics training. And of course, the evidence I mentioned is hardly systematic. She holds that in particular we need to get the views and the perceptions and experience of faculty who train graduate students in science and in engineering. They should be the ones primarily responsible for the training of their students in ethics. And Swazey speaks out of empirical knowledge.

She and colleagues conducted a study of faculty and students in graduate education that was completed in the early 1990s and funded at least in part by the National Science Foundation. Her important finding was that many faculty held the view that the best way to transmit ethical standards was by example, by a process of osmosis. The point she makes is that if many, or most, still hold some form of that view, then we have an urgent need to train the trainers. In explaining how she makes that inference, she is going to address the "why" of teaching research ethics and focus on faculty responsibility.

While we lack empirical studies of the perceptions and experience of faculty during the 1990s, we have access to the views and the experience of some graduate students and postdocs with regard to training and research ethics. In fact, the 1993 Sigma Xi Forum proceedings volume that you found on your chairs at the earlier session contains the reports of some postdocs at that meeting, and one of them introduced a set of ethical obligations for faculty of graduate students and postdocs. She modestly concluded that she wasn’t in the best position to say what the duties of students should be.

In the meantime, we’ve gotten published findings of a number of surveys of graduate students and postdocs of biomedical trainees, and postdoctoral researchers in biomedical programs of research, and also of physics students, and, from my point of view, most relevant this morning, four volumes of research ethics cases that came out of a project funded by the National Science Foundation and carried out by the Association for Practical and Professional Ethics.

That program initiated, in 1996, workshops for graduate students and postdocs in sciences and engineering. The requirement for each attendee and participant was to produce a case study. The case studies of all of the students who participated in the workshops between 1996 and 1999 have been published. There are four volumes, one for each of those years, and the volume for the summer workshop of 2000 is in press right now. Following the workshop, the participants refined their cases and produced commentaries, and the faculty at the workshops also produced commentary on the cases. So those volumes contain the commentaries, as well.

My remarks, which will follow Judith Swazey’s, are drawn from a study of those volumes. I went back and re-read them all recently. There are 63 cases, and of them, 27 deal with conflicts in research groups or departments over data, over ideas, over credit, over the assigning of authorship and over intellectual property. The 2000 volume, which is in press, has its share of such cases.

These are precisely the topics that Judith Swazey picked out in her abstract as in great need of attention. Interestingly, they are the topics that over 40 percent of the participants in our workshops wrote their cases about. The cases include narratives and dialogues, and they point up the need for faculty to assume responsibility for articulating policies for making them explicit, that address the issues in those cases. And the students comment on the need to create opportunities for discussing the policies for considering the rationales.

Stephanie Bird unites three important perspectives in approaching the teaching of research ethics. One is the perspective of a scientist in the neurosciences. The second is the perspective of an administrator in a major research university who is charged with dealing with research ethics, and particularly the teaching of research ethics. And the third is her position as co-editor of a journal Science and Engineering Ethics. That journal, in fact, has published some of the empirical research that I mentioned representing the perceptions of graduate students and postdocs.

Stephanie Bird has extensive experience in teaching research ethics in a variety of settings. We know that part of the challenge is fitting the teaching to research groups, seminars, courses, even company situations. She observes that teaching the responsible conduct of science is primarily a matter of conveying the professional values and ethical standards of the discipline, and she emphasizes how this effort requires making explicit information that is often implicit.

Notice the implication that it is a mistake to rest on unspoken standards. Again, we have a counter to the idea that we can rely on teaching by example or osmosis alone. Stephanie Bird goes on to explain the methods of teaching in a variety of settings within research groups and in courses. This is the "how." So the "why," the "what," and the "how." And I’m now going to turn to Judith Swazey.

Teaching and Learning Research Ethics: Needs and Opportunities Revisited
by: Judith P. Swazey
The Acadia Institute

My comments this morning focus on why teaching and learning research ethics, now termed the responsible conduct of research, is important, and why I continue to believe that the first order task is to train the trainers—that is, the faculty responsible for teaching undergraduate and graduate students in fields of science and technology. In February 1993 I did a background paper for Sigma Xi’s Forum on Ethics, Values, and the Promise of Science, for the session on Teaching Ethics. That paper drew on some of the findings from the Acadia Institute’s study of professional values and ethical issues in the graduate education of scientists and engineers. The study included a survey of graduate school deans and surveys of national samples of 2000 doctoral students and 2000 faculty in 98 of the largest graduate departments in four fields: chemistry, civil engineering, microbiology and sociology. Reviewing our study and my paper for the 1993 Sigma Xi Forum, and some of the subsequent literature on the responsible conduct of research, I must confess I found myself wondering what I have to say that is new on the occasion of this year’s forum. And the answer, to myself, was "not much, in terms of a research-based presentation." There has been no comparable large scale study in the decade since Acadia’s work, which my colleagues and I believe there should be. We would like to know whether the experiences and views of students and faculty have changed, and if so, how and why?

There have been a number of smaller studies of individual departments and groups, such as medical residents; individuals have been assessing their research ethics programs; and the Office of Research Integrity is launching a much needed initiative for research on research integrity. On a larger scale, agencies such as the National Institutes of Health have, I think, missed a major opportunity to learn more about both the content and effects of research ethics education by not having built an evaluation component into their requirements for training grant recipients. This is a defect that I wish they would remedy as those requirements currently are being extended to all investigators supported by NIH funds, and as a linkage is being developed by the Department of Health and Human Services between education and training in the protection of human subjects and in the responsible conduct of research.

Let me next briefly recap a few of the Acadia study’s findings that bear on the focus of this session—authorship/recognition/ownership of ideas—and on research ethics training more generally. In our surveys, we asked graduate students and faculty if they had ever observed or had other "direct evidence of" various types of research misconduct, such as plagiarism, and what a decade ago was termed "questionable research practices." With respect to plagiarism, a cardinal type of misconduct with respect to professional recognition and advancement and ownership of ideas, almost one-third of the faculty reported knowledge of plagiarism by their graduate students, with over 40 percent of faculty in civil engineering and sociology reporting such knowledge. Between 6 percent and 9 percent of both students and faculty reported knowledge of faculty who had plagiarized; in civil engineering 18 percent of faculty said they knew of plagiarism by their colleagues.

Another survey item relevant to today’s session, under the "questionable research practices" category, was inappropriate assignment of authorship credit. About one-third of both faculty and students reported knowledge of such behavior by faculty. Not surprisingly, since students have less control over authorship, a much smaller percentage reported this behavior by students (9 percent of faculty, 12 percent of students). Given both the actual and symbolic importance of authorship credit, issues around authorship can be some of the most contentious and stressful ones in academia. As we heard repeatedly in testimony before the Commission on Research Integrity, and as faculty and students recurrently report, disputes over authorship can be a prime cause of high stress—or worse—for students, ruptures in research groups and the filing of misconduct allegations.

We also asked students about what they judged to be the most important influences in shaping their professional values and preparing them to deal with ethical issues in their field. Of the 10 sources rated by the student respondents, the three top rankings were supportive faculty, other graduate students and family. The bottom three were discussions in other courses, labs and seminars; professional organizations in their field; and, at the bottom, courses dealing with ethical issues. Faculty, in turn, were asked to rate the effectiveness of seven ways that students can learn about professional values and ethical issues and standards in their field. "Interaction with faculty in research work" and "informal discussion of ethical problems when they occur" received "very effective" ratings from 65 percent and 61 percent, respectively. There was then a dramatic plunge in "very effective" ratings for the other items: for example, only 19 percent judged discussion in regular courses and 14 percent special courses as very effective, and only 7 percent viewed codes of ethics and standards of professional organizations as very effective methods.

Three points about these data concerning teaching and learning research ethics seem especially salient for today’s session. First, as other data in our study indicated, most students and their faculty had not been exposed to or engaged in formal instruction in the responsible conduct of research, which poses the question of the bases on which they were judging the effectiveness of these educational methods. It would be informative to have large-scale comparative data now, a decade later, to know whether more students and faculty have been involved in formal teaching and learning, and, if so, how they assess its effectiveness. Secondly, while the students held that supportive faculty members were the most important influence in shaping their professional values and in their "ethical preparedness" training, other data in our study and a profusion of other research and reports show that students receive far too little "mentoring" by their advisors or other faculty with respect to many important aspects of their training.

Third, our faculty respondents clearly believed—and I would bet that most faculty today continue to believe—that informal means are the most effective way to transmit professional values and ethical standards-what we call "the osmosis factor." Many of us who have been involved in research ethics believe that informal means, including latent role modeling or "teaching by example"—assuming, which is not always the case, that the examples are positive ones—are important and necessary, but not sufficient for a number of reasons. For example, as Dr. Bird will discuss, authorship credit involves complex issues and a wide range of variations in disciplinary and sub-disciplinary conventions and practices that are difficult for "novices" to decipher unless they are explicated and discussed. In many fields, furthermore, the large size of laboratory or other types of research groups, and often absentee senior figures, makes reliance on informal transmission processes even more dubious.

Finally, one might ask, "So, so what? Why all the fuss about training in the responsible conduct of research?" Attention to research ethics has been catalyzed by many highly publicized cases of research misconduct, such as data falsification. But to me, misconduct is almost the least important facet of learning about and engaging in the responsible conduct of research. There are more, and more complex, issues in the everyday conduct of research and life of a lab group that arise in developing research proposals; in performing and analyzing research; in reporting findings; and in the professional interpersonal relationships between members of a research group.

Among the many answers to "so what?", one is that the both the public and the academy believe that research integrity is important. When we surveyed graduate school deans, faculty, and students, very large majorities (99, 88, and 82 percent respectively) felt that "ethical preparedness training" should be an important activity of their universities and departments. All three groups, however, reported a substantial difference between "should" and "is" for research ethics training. Remembering that our data were collected a "long time" ago, between 1988-91, 51 percent of the deans reported that their institution was not very or not at all active in this realm, and 25 percent of students and 13 percent of faculty said their departments were not at all active. Hopefully, these percentages would be much lower today—but we just don’t know.

Finally, why do I believe that "first, train the trainers" is a priority for increasing and enhancing teaching and learning research ethics? Since both students and faculty, in our study and in the experience of most of us, hold that faculty are the most important source of students’ acquiring their professional values and learning about ethical standards and issues, it doesn’t take a rocket scientist to reason that the teachers of graduate students can, in principle, do the most effective job of explicitly teaching them about various aspects of the responsible conduct of research. And by "effective" I mean, in part, that the students’ own faculty will have the greatest credibility and impact, compared to those of us who are parachuted in from the outside. If faculty are to accept and assume this task, however, they need to learn both the subject matter and methods of teaching research ethics—and on that note, let me turn to my colleagues, Vivian Weil and Stephanie Bird.

Research Credit and Authorship: Graduate Student and Postdoc Case Studies
by: Vivian Weil
Center for the Study of Ethics in the Professions, Illinois Institute of Technology

My data are the cases produced by the graduate students and postdocs in the summer workshops I mentioned earlier. In many of the 27 cases dealing with conflicts over data, ideas, credit, authorship, intellectual property, the writers speak of ownership of ideas. Now, that’s not, strictly speaking, what intellectual property is. Our patent law doesn’t cover ideas. It is restricted to material embodiment of ideas, and research groups, in fact, have a strong interest in generating ideas and in circulating ideas. I came to the conclusion that the conflicts expressed in terms of ownership of ideas were misleadingly expressed in those terms, and that they have, more importantly, to do with recognition for work, with credit and authorship, rather than property.

The research groups in question almost always lack explicit understanding about form of credit, the bases of credit, and the grounds for assigning authorship. Discussion about such understandings and the rationales for those understandings seem to be rare occurrences, and the cases, of course, show the damaging misunderstandings that occur when criteria for recognition and standards for authorship are unspoken.

Because self-esteem and career advancement are at stake, evaluations within research groups that determine credit and authorship can be uncomfortable to make and uncomfortable to confront and to accept. These very natural reactions, I want to argue, make it all the more necessary for research groups to have policies and to state those policies and explain them, the policies governing credit and authorship.

Some of the cases even yield suggestions for policies. For example, there was one case in the most recent group, which is not yet published, that indicates a need to articulate whether credit can be gained from data that is produced when students rotate through labs and produce data in a lab as a rotating student, rather than a member of that lab or research group.

What I want to do now is go through one of the cases produced in our workshops, and it comes, of course, from a subgroup of the 27, dealing with the topics that we’ve mentioned. As in many of the cases, this one shows the absence of local policies and standards, and failure to use any channel of communication to articulate in advance standards or criteria for authorship. After we go through that case, I want to show what can be accomplished when people give attention to the standards. We had a marvelous demonstration in the lecture that preceded this group of what you can achieve if you really give attention to making explicit the principle that should govern conduct.

I’m going to present some authorship standards that were arrived at by a consortium of journal editors and offer it as a model that groups can consider. And then, in conclusion, I’m going to have some more to say about why research groups and departments should—a strong sense of "should," ethically "should"—undertake the responsibility for defining standards and criteria, and making them explicit, and bringing them forward for discussion, with senior members taking the lead.

The first example is from our first volume, and it’s called "Informal Discussions, Formal Authority." It’s a situation in which a graduate student and his professor have two discussions, the first one quite lengthy, over an hour. Through those discussions, the professor and the graduate student together come up with a hypothesis that seems to give punch and impact to a paper by a postdoc in the same group, which the professor is revising. The graduate student is really very instrumental in putting a robust interpretation into the paper.

When the graduate student sees the revised draft, he recognizes the conclusion. It is the premise of his thesis and something that he considers to be a seminal element of his work. When he mentions this to the professor, the professor replies that if the student thinks he deserves authorship, well, then he can be co-author. The student answers that he is uncomfortable arbitrating his own case, especially when it involves the postdoc in the group. The professor then decides to deal with the situation, which he calls unusual, at a group meeting the following week, and they will let the lab group decide whether the graduate student should be co-author.

The discussions between the graduate student and the professor seem to exemplify the kind of mutual teaching and learning that should go on between competent graduate students and their professors. But the situation leaves it unclear whether the student has made a contribution that is significant enough to merit co-authorship. Don’t know what the standards are and how the professor regards the student’s contribution in criteria. It might well be that the student merits co-authorship if the professor is using a model that was developed by the student. Notice that the professor treats the matter carelessly, leaving it up to the student, who justifiably feels uncomfortable.

I think we can say that the professor acts wrongly in two respects. He says, in effect, that the standards for authorship and credit are not his concern; thereby, he’s refusing to accept responsibility for his own conduct and for deciding this issue. Notice, recalling Judith’s point, that by saying that we don’t want to rely solely on modeling, we don’t want to deny that modeling is important and takes place in any case. In this case, the professor offers a model of treating credit and authorship in a cavalier, you might even say, negligent way.

The case shows how a lack of clarity or of conventions for credit and authorship in a research group can produce damaging misunderstandings. In opting to leave the matter for the lab group to decide collectively at a group meeting, the professor continues to refuse to take the responsibility himself, and he sets up a situation for discussing credit and authorship after the fact, having lost the opportunity to do it in advance, and in a situation that is likely to heighten the discomfort associated with assigning credit and authorship.

What I want to do now is turn to a model for authorship standards, so that we have an example of what we can get if people do give attention to articulating standards. This one, as I mentioned, comes from a consortium of journal editors, and you see it there. It rejects guest, ghost and gift authorship. It also rejects what one editor described as this murky business of deciphering what being first or last means. It’s also attuned to the circumstances of collaborative research and to situations in which there are many contributors to the research.

The idea is that the byline includes only a few names of investigators, who take responsibility for the contents of the article. Think of them as guarantors who vouch for what the report contains. At the end of the article is the listing of contributors, with the contribution of each, as each contributor has described it. The editors provide a checklist of the kinds of contributions that contributors might indicate. Finally, there may be acknowledgments, and the acknowledgments are limited to sources of funding or sponsorship of the research. So you see a move from authorship toward contributorship, which corresponds better to the circumstances under which research is carried out and reported.

My suggestion is that this kind of approach is one that a research group can pick up in looking for, defining and explaining their own standards. What is the function of the names in the byline; and why do we need to know what each of the contributors did; and why do we relegate to acknowledgments the funding sources and include nothing else in the acknowledgments?

I want to conclude with the question of collective responsibility of the research group or department. I think through our questions so far we’ve already brought out good reasons for the responsibility falling onto the research groups, themselves. That’s where the problems arise. That’s where credibility of those who pronounce on standards resides. In addition, I think we can make an ethical argument that research groups and departments have a duty to develop collective action strategies.

There’s a philosopher at the University of Arizona, Alan Buchanan, who has observed that there’s a fundamental ethical principle of common sense that says, "Acting responsibly requires doing what we can to improve the chances of acting responsibly." So if you put this principle together with the perception of laxity in many research settings, laxity which may be understandable in people when you don’t have some mechanism to coordinate the behavior, individuals waiting for others to step forward, or have a sense that they don’t want to be putting forth when others aren’t going to put forth, a lot of reasons why—not evil intent or anything of that sort—laxity persists.

When you combine this principle with the existing laxity, we have a strong basis for saying that departments and research groups have an ethical duty to assume collective responsibility for defining and making explicit and explaining their policies for managing data, for circulating ideas, for allocating research projects to students and postdocs, and for assigning authorship. I think it’s a kind of neat and interesting point that the practice within local research groups complies with the content of ethics teaching, and they also make up the environment for the ethics teaching, or the climate. And investigators and research groups must attend to both.