These days, the term “credit crisis” invokes thoughts of Wall Street and financial debt. But in the scientific enterprise, structural changes are affecting core issues of intellectual credit and indebtedness as well as taking responsibility when problems arise. Following is an in-depth discussion and proposed solution from distinguished scientist and author Michael Domjan, who has experienced these issues as a researcher, editor, and administrator.
Throughout much of the 20th century, the major figures in psychology were scientific loners, battling for dominance of the field. They typically developed independent lines of research and pretty much ignored everyone else’s work. Skinner’s Science and Human Behavior (1953), which was used as an introductory psychology text at Harvard, did not have a single citation to anyone else’s writings. Many were happy to have graduate students help them conduct their research and disseminate their ideas, but collaborations rarely extended beyond that. A given project rarely had more than one or two contributors, and it was easy tell who did what and who deserved the credit (or discredit).
We now operate in a very different world. The number of authors on publications is increasing, and often the collaborators are not graduate students but scientists from other laboratories and disciplines. This trend is evident in all fields of science (Wuchty, Jones, & Uzzi, 2007) and is an inevitable consequence of interdisciplinary research.
Acting on Principals
– Catherine West
“Interdisciplinary” and “transdisciplinary” have almost become buzz words, but they signify much more than just fashionable labels. Their commonplace use reflects the fact that deep understanding of a problem often requires the coordinated efforts of a range of scientists who are expert at different levels of analysis and employ different specialized techniques. The importance of these issues has commanded the attention of APS President John Cacioppo in several of his recent columns in the Observer (Cacioppo, 2007a, b). As Cacioppo points out, psychology has much to contribute to interdisciplinary efforts, and interdisciplinary and collaborative research will play an increasingly important if not dominant role in psychological science. This shift will require changes in how we deal with issues of scientific credit and responsibility.
I have spent most of my professional life doing single-laboratory science, as is typical of psychologists who started in the field in the late 1960s. I had several research collaborators along the way, and I enjoyed worked with them, but most of my research started and ended in my own lab, where it was easy to identify who was responsible for, and should receive credit for, various aspects of the work. Determining credit and responsibility became more difficult when I had to evaluate research produced in other laboratories during my service as a journal editor and later as chair of the department of psychology at The University of Texas at Austin. But issues of scientific credit and responsibility have been even more challenging in my current position as director of the Imaging Research Center (IRC) at The University of Texas. The primary research modality at the IRC is magnetic resonance imaging (MRI), which was developed through the concerted efforts of physicists, electrical engineers, biomedical engineers, neurobiologists, computer scientists, and others. I had to hire an in-house staff (none of whom are psychologists) and work on hiring faculty in psychology, neurobiology, physics, and biomedical engineering. All of these tasks required assessments of scientific accomplishment. Unfortunately, determining who is responsible for, and deserves credit for, a scientific finding is becoming increasingly difficult.
My first encounter with the credit-assignment problem occurred when I was a journal editor. A serious problem arose with a multiauthored paper that was published in the journal. The problem was of sufficient gravity to warrant publishing a retraction. The authors agreed on what the problem was and which of them had committed the error. The only remaining issue was whose name should appear on the retraction. The original publication did not spell out which author was responsible for which aspect of the research. I argued that all of the original authors needed to sign the retraction, which they eventually did. Authorship is both credit and responsibility. I wonder how often graduate students eager for authorship consider that if you accept credit you also have to accept whatever discredit may befall the publication down the road.
Another memorable encounter with the credit-assignment problem occurred during a discussion I had as department chair with a faculty member who was applying for tenure. The tenure process required specifying the percentage of effort that was contributed by the candidate to each of their publications. In this case, several of the candidate’s publications included numerous authors. I pointed to one of these and asked how much of the work in the publication was done by the candidate. The reply was “80 percent.” This claim immediately raised another question: if the candidate did 80 percent of the work, why were there four other authors? Assuming that those four other authors contributed equally, does a 5 percent contribution warrant authorship? Not in my view. Furthermore, how is a reader to discern that the relative contributions of the authors ranged from 5 percent to 80 percent? By convention, the order of authorship is supposed to convey some information about relative contributions, but I doubt that anyone expects that the variation in credit and responsibility would range from 5 percent to 80 percent.
On another occasion I served as a judge in a poster contest open to graduate students and postdocs. The award was to go to a single individual. One of the projects was outstanding and clearly worthy of the award. It described an interdisciplinary project that involved behavioral testing, genetics, and electrophysiology. There were nine authors on the poster. The presenter was eloquent, erudite, and convincing in describing the results, but said nothing about who was responsible for what aspect of the work. When queried more closely about the assignment of responsibility, it became clear that the presenter had no involvement with a couple of critical aspects of the project. Needless to say, the award went to someone else.
|Publish or Perish
Determining issues of credit for published papers can be tricky, but sometimes it is also a challenge to decide who should be included on the author list in the first place, especially in a project with multiple labs and large teams. In 1993, psychologists Mark Fine, University of Dayton, and APS Charter Member Lawrence Kurdek, Wright State University, laid out guidelines for determining authorship order on collaborative publications with students (“Reflections on Determining Authorship Credit and Authorship Order on Faculty–Student Collaborations,” American Psychologist, Volume 48 (11), 1141-1147).
Many of these principles can be applied to transdisciplinary research projects as well. Fine and Kurdek recommend that, early in the collaborative endeavor, the research team should discuss the guidelines along which authorship is determined, the meaning of authorship credit and order, and what will be expected of each author. They also suggest that each party assesses their specific abilities and the tasks required to complete the publication. On the basis of that assessment, it should be easier for the collaborators to divvy up the tasks in a reasonable manner. Certainly, issues may arise as the project progresses; however, these preemptive discussions may reduce the likelihood of conflict when the paper goes to press.
– Catherine West
The above examples illustrate some of the areas in which the uninformative nature of authorship lists creates significant problems in making informed professional decisions. Another important area is in hiring decisions. When you are trying to hire someone, you are looking for a specific set of skills. However, just because someone is an author on a paper that uses a special technique does not mean that he actually know how to use that technique. Authorship no longer ensures that the person was responsible for formulating the research question, that he was responsible for the literature review, that he did or could have performed the data analyses, or that he wrote any portions of the final manuscript. At some level we all know this, and for that reason we do not make hiring decisions based entirely on someone’s curriculum vitae. However, we do make significant decisions based on CVs, namely which candidates to put on a “short list” and examine more closely.
Interdisciplinary research projects involve bringing people with different skills together to work on a common problem. Each member of the team presumably contributes a different area of expertise. With well functioning research teams, each member also learns some of the language and concepts of others at the table. All of that is to be applauded. Problems can arise, however, when members of interdisciplinary teams begin to claim expertise that is actually contributed by others on the team. I have heard colleagues talk about how they are now doing imaging research, when I know that they have no idea what is involved in collecting MRI data. I have also heard colleagues talk about “doing genetics” or “doing hormone assays,” when in fact what they are doing is sending out samples for someone else to analyze and interpret.
The credit-assignment problem is also often evident when someone describes the results of interdisciplinary research in conference presentations, colloquia, and other scientific gatherings. It has become common to provide a list of collaborators, usually at the end of the talk. Such lists are typically organized by level of professional accomplishment (graduate students, postdocs, faculty colleagues) or by institution. Rarely are we told who did what. I applaud speakers who identify their colleagues during their presentations, as they discuss various aspects of the work. It seems that senior scientists are more likely to do that than junior colleagues. I have been particularly frustrated by job candidates who describe the full range of work that is being done in a highly productive laboratory, without making clear what part of the work is actually their contribution.
Why does all this matter? After all, isn’t the science what is most important rather than who does it? If we truly believed that, we should be satisfied with research publications that do not list any authors or only listed them in alphabetical order. Obviously, that is not a world order that anyone would subscribe to. Authors appear on papers because we need to know who was responsible for the work and who should receive credit (or discredit) for it. This assignment of credit and responsibility is then used in all facets of scientific life. Prospective graduate students use it in deciding which lab to join. University administrators use the information in hiring and promotion decisions, and in deciding how much lab space or graduate student support a faculty member should receive. Granting agencies use the information to decide which proposals to fund. The one area where this information is sometimes not used is in the anonymous submission process that is available with some journals. But, even with anonymous submissions, it is often pretty easy to guess whose lab was responsible for the research.
A New Approach
Given the importance of the proper assignment of scientific credit and responsibility, we should work to improve the current system. We cannot legislate against colleagues who overstate their competence or award authorships based on an exchange of favors. However, we can work to make the process of assigning scientific credit and responsibility more transparent and open to public scrutiny. A good place to start would be to require a statement in all multi-authored publications that spells out what each author contributed to the final product. Authors may disagree among themselves as to who did what. The requirement of a statement of contributions would not circumvent such disputes but it would at least force a discussion among the authors and encourage them to arrive at a statement that they are prepared to defend in public. Such statements would also allow directors of laboratories to see each other’s reasons for inclusion in the authorship list and may help to create more uniform standards for authorship credit.
The suggestion that authors disclose their individual contributions to a collaborative project has been discussed in the medical literature for at least 20 years (Huth, 1986). In a recent reaffirmation of this suggestion, the International Committee of Medical Journal Editors (2007) stated that in dealing with multiple authorship “editors are strongly encouraged to develop and implement a contributorship policy” to clarify what each author did on the project. It may be time to develop such a policy for psychological science as well. ♦
Cacioppo, J.T. (2007). The rise of collaborative psychological science. Observer, 20 (9), 3 and 52.
Cacioppo, J.T. (2007). Better interdisciplinary research through psychological science. Observer, 20 (10), 3 and 48-49.
Huth, E.J. (1986). Irresponsible authorship and wasteful publication. Annals of Internal Medicine, 104, 257-259.
International Committee on Medical Journal Editors. (2007). Uniform requirements for manuscripts submitted to biomedical journals: Writing and editing for biomedical publication. Retrieved from www.icmje.org.
Skinner, B.F. (1953). Science and human behavior. New York: Macmillan.
Wuchty, S., Jones, B.F., & Uzzi, B. (2007). The increasing dominance of teams in production of knowledge. Science, 316 (5827), 1036-1039.
The Mentor’s Dilemma*
One central aspect of the task of a faculty member who is training a graduate student is to turn that student into a researcher capable of designing incisive, conceptually clean experiments that help to reveal the underlying processes that are operating to produce the phenomenon of interest. (Earlier in this chapter we noted that it is precisely the task of designing and doing these sorts of experiments with graduate students that has given us so much joy during our careers.) To do this successfully, we often use pedagogical techniques. One prominent technique involves framing a problem in such a way that the student is led to discover on his or her own just how the experiment should be designed and operationalized. A variant of this ploy is to present competing hypotheses in such a way that leads the student to have an insight that pits the two hypotheses against each other in a fruitful way. There is a bit of licensed pedagogical deception practiced in this instance: The faculty member leads the student to an insight (or set of insights) that the faculty member already has had by framing the problem in a way (or, more generally, providing the background information) that led the faculty member to the insight in the first place. The faculty member then gasps in wonder at the student’s brilliance! The rationale behind this sort of technique is that the student makes the resulting experiment his or her own, and carries it out with the dedication and tenacity that this identification has produced-and that is often necessary for the successful completion of a research project.
There is a wonderful description of this kind of benign deception in the obituary Charles Harris (2001) wrote about Hans Watlach. Harris was explaining why Wallach had been the source of so many students who went on to successful research careers.
With students, Wallach often conveyed the impression that he was simply thinking aloud, engaging with them in a spontaneous search for insights. He not only encouraged them to think incisively, he also made them feel that they were doing it on his level, collaborating with him in his analysis of a problem. A former student offered this description: “He would become utterly involved in whatever question we were focusing on. … If any of us were ever able to contribute some glimmer of possible understanding, his face became radiant with delight — even when what we had thought of went directly against his favorite theory.”
Hartis continued, “In retrospect, I surmise that he put considerable effort and planning — and even some playacting — into his teaching.” For example, on one occasion, at least, Wallach left the room, paced the hall, and returned to say, “I’ve been struggling to solve a problem. I know it’s not our topic for today, but I hope you can help me.”
Harris, a distinguished perceptual researcher in his own right, goes on: “Some years later I told him I was perplexed about that incident because I had since learned that he had solved the puzzle in an article he had published a year earlier. With abroad grin he replied, ‘Yes , I know.’”
In that charming case, the issue of publication credit did not arise. But suppose the faculty member had led a student to an insight on which he or she actually went on to do research? Who ought to be the first author on the resulting brilliant paper? We already know (cf. Ross & Sicoly, 1979) that each participant in a joint enterprise, even a student and a faculty advisor collaborating on research, has a tendency to overestimate the degree, or the value, of his or her individual contribution. So this background tendency creates an authorship dilemma in any case. But the interesting irony is that the mentor, as good pedagogue, has greatly exacerbated the dilemma of authorship allocation by making the student believe that he or she has contributed the original idea, in addition to all the hard work that the student actually contributed to the execution of the project. So the student, benignly tricked at the beginning of the experiment, will carry it out with the induced hope that he or she will be the lead author. As Keith Spiegel and Koocher (1985, p. 352) have noted, one of the most frequent sources of conflict, and of accusations of violations of ethics codes, arises from the authorship dilemma between student and faculty member. We suspect that many of these acrimonious disputes have their origins in the unintended consequences of the initial adoption of the mentoring stance by the faculty member. Irony indeed!
What to do about this dilemma is not easy to say. For senior faculty members, for whom the question of evaluation based on authorship is in the past, the situation is often resolved by giving the student first authorship. This is not so easy to do, however, for a younger faculty member struggling to establish an academic reputation — and to gain tenure. Our suggestion for younger faculty is to have a conversation very early in the faculty–student relationship that generally alludes to the faculty member’s lead role in the general area, perhaps one that has been staked out by the faculty member’s grant proposal. This conversation can be reinstated after the student’s “discovery” of the proper design or procedure of a specific experiment. As the data collection is completed, the faculty member might also take on some of the paper drafting tasks, such as writing the introduction, to demonstrate a continuing involvement in the project.
One thing emerges clearly. Collaborations in which participants of unequal power are involved are fraught with this problem. They will be best avoided by early and clear understandings of the division of labor and the division of authorship credit for the project, and these understandings can and should be revisited if the division of labor shifts during the conduct of the project. Letting each participant form egocentric and unspoken “understandings” virtually guarantees corrosive misunderstandings. We all know this. What the concept of the mentor’s dilemma introduces into the discussion is an understanding of how it is that the faculty member, originally functioning in the mentor role, induced the student to regard him- or herself as the conceptual origin of the study, a fact that the mentor now forgets.
Having realized the force of the mentor’s dilemma, we do not think that our advice provides a complete solution about how it ought to be handled. We only can hope that by articulating the dilemma both student and faculty member are more likely to recognize the complexities of what is, in fact, a highly complex and emotion-laden issue.
*Copyright © 2004 by the American Psychological Association. Reproduced with permission. The official citation that should be used in referencing this material is Zanna, M.P. & Darley J.M. (2004). Mentoring: Managing the faculty-graduate student relationship. In J. M. Darley, M. P. Zanna, & H. L. Roediger III (Eds.), The compleat academic: A career guide (2nd ed., pp. 122-124). Washington, DC: American Psychological Association. The use of APA information does not imply endorsement by APA.