WASHINGTON, D.C.--After you've spent months toiling away in the laboratory, preparing your latest research findings for publication should stir up feelings of satisfaction and accomplishment. But instead, many young scientists find themselves thrust into contentious authorship disputes, controversial decisions regarding data manipulation and arguments over review procedures. If there were better oversight of ethical and research responsibilities--by the scientific societies that solicit manuscripts for the journals they publish--many disagreements would not occur, argue speakers at a conference held earlier this week.

The American Association for the Advancement of Science (AAAS) together with the U.S. Office of Research Integrity (ORI) convened a meeting here on 10 to 11 April to discuss how U.S. scientific societies could better create and promote research integrity and publication ethics policies. The conclusion of conference speakers: During the past 50 years, scientific societies appear to have made little progress in imposing standards for research integrity, despite repeated calls to embrace new ethical recommendations. Many societies seem hesitant to enforce guidelines across their respective communities because they believe ethical issues are best dealt with on a case-by-case basis. Panelists contend, however, that societies and scientific mentors must invest greater effort to develop community-wide integrity codes so to better moderate research communications. Despite skepticism that few communities will rush to adopt new policies, some societies have taken steps toward rectifying certain problem areas: A few medical journals, for example, are considering alternative ways of assigning credit and may ditch traditional means of naming authors in order to combat authorship disputes.

Mentoring Ethics Without Policies

Societies need to educate investigators so they can better enforce rules in the research environment, panelists say, but in today's climate, even if researchers are aware that guidelines exist, they seldom refer to them. Consequently many ethical principles remain unheeded in experimental design and execution and in manuscript preparation. "Mentors should impress upon student trainees the ethical challenges involved in every phase of research," says Mark Frankel, director of AAAS's Scientific Freedom, Responsibility, and Law Program, in the opening session of the conference. "Senior scientists should use the laboratory setting to ensure that those whom they supervise understand the values, ethical prescriptions, and institutional guidelines governing research," he states.

Such tutoring extends to the practice of writing manuscripts, and if the apparent lack of ethical guidance is anything to go by, principle investigators and mentors must boost efforts to better convey research-writing skills to their students and staff. Anne Hudson Jones, a professor at the Institute for the Medical Humanities of the University of Texas who spoke at the meeting, confirms an absence of ethical control in the publication of medical studies. Relating the findings of a U.S. medical school survey, she reported that 65% of respondents (77 medical schools) had either given up attempts to create their own authorship policies or had never broached the subject, and a further six schools "reported that they did not know whether there were authorship policies, or whether there had ever been discussions of such policies."

A preliminary analysis of a similar survey of scientific societies by Frankel revealed equivalent findings: 17 of 46 society respondents did not have general statements on publication ethics, and over 90% did not have their own codes regarding authorship placement in a manuscript. This particular finding, however, does not rule out that such guidelines actually do appear in the society's journals. So while society publications may describe ethics-related statements with regard to authorship, societies themselves may have no overriding codes or guidelines.

Publication Ethics: High Crimes and Misdemeanors

"Mentors should give appropriate credit to co-authorship between the trainees and should publicly acknowledge students and trainees that contributed to the research and preparation of their work," reiterates Frankel. Because this is often not the case, scientists end up clashing among themselves as well as with journal editors. Michael Zigmond, professor of neurobiology at the University of Pittsburgh, considers authorship disagreements, duplicate publications, and misleading graphics to be "misdemeanors." He contends, however, that these practices occur much more frequently and do more harm than do the "high crime" behaviors of fabrication, falsification, and plagiarism. Nevertheless, these serious behaviors continue to occur, says Zigmond, who claims that demonstrating such practices to your students and postdocs is "the worst thing you can do."

To temper arguments over misdemeanors--such as authorship order and determination of principle author--it has been proposed that the traditional way of listing authors on manuscripts be scrapped. Would you rather be referred to as "first author" or "contributor"? It was this aspect of proactive misconduct resolution that Jones described. The byline of a manuscript, she says, should "name only those that contributed most substantially" to the research. How authors are listed can be confusing: postdocs who have seen authorship statements had "considerable difficulty" interpreting those policies, relates Jones.

First Author, or First Contributor?

Under a model proposed in 1997 by a group at the Institute for Health Policy Studies at the University of California, San Francisco, investigators would be referred to as "contributors." These researchers would not have to contribute significantly, nor write, revise, or approve the manuscript or the research, and they would not be held accountable. Jones describes an additional assignation of "guarantor" which would identify persons who invest substantial intellectual and practical effort in the research, and who are publicly responsible for ensuring the integrity of the entire project and the final publication. In their original 1997 Journal of the American Medical Association paper, Drummond Rennie and his colleagues argue that such arrangements will "eliminate the artificial distinction between authors and acknowledges and will enhance the integrity of publication."

Leading medical journals such as the Lancet and the British Medical Journal (BMJ) have already incorporated a "contributor" section in their publication policies. Following the Lancet's lead, the BMJ made the change roughly 2 years ago because, they say, the traditional definition of authorship "has some serious flaws." Jane Smith, deputy editor for BMJ, tells Next Wave that in addition to the on-going disputes of who did what and who should go where, it is "unrealistic" to expect researchers in multidisciplinary projects to fully understand and defend the contributions of collaborators. "You're asking the statistician to understand pathology," explains Smith. The idea is akin to movie credits, where you list the part each person plays. "It works quite well and it makes researchers stop and think," and that may help them better assign projects and clarify experiments in future collaborations, she says.

Reserve Caution: Electronic Ethics

Faith McLellan, a medical writer who is editor-in-chief of Second Opinion and who co-edited the recently released Ethical Issues in Biomedical Publication with Jones, informed the audience of the ethics and pitfalls of publishing in the electronic media. The immediacy of results posted online, for example, could be both "very attractive and very unattractive," she says. If it is online, your paper can be read globally and will therefore be more accessible to the public and lay person, so conclusions and findings must be clear and unambiguous. Inline advertising--links to scientific companies within methods sections--are also ethical considerations because conflicts of interest may be perceived.

Researchers should also be wary that attempting to publish research data that has been posted on their personal, lab, or departmental Web pages could be difficult if the journal to which they wish to submit their work insists that such postings constitute "prior publication." Because there are no standards governing publication of information that has been made available online prior to peer-review, some journals will not publish papers that include such data. "I wish somebody could explain this to me, because I just don't understand this," requests Zigmond.

An inferred message from the discussion on promoting responsible conduct and ethical oversight is that it is in researchers' best interests to make sure they are fully aware of publication guidelines and research ethics. And to achieve this level of understanding, supervisors and mentors need to convey to up-and-coming researchers the ethical skills they will need throughout their scientific careers, both on and off the bench.