Nautilus

Accountability of authors

This week’s Nature addresses how the responsibilities of co-authors for a scientific paper’s integrity could be made more explicit (Nature 450, 1; 2007). The text of the (free access) editorial:

The two most notorious frauds of modern science, by the stem-cell biologist Woo Suk Hwang and the physicist Jan Hendrik Schön, both brought into question the responsibilities of co-authors in the oversight of their colleagues’ work. But despite the concerns raised after these episodes, there remains a need for a clearer understanding, both within a collaboration and by readers of the eventual papers, of the various contributions made by the authors not only to the research but also to safeguarding its integrity.

One welcome development in transparency was pioneered by the medical journals. Authorship of a paper is justified when a researcher has contributed significantly to the work being described and to the writing or approval of the manuscript. But the traditional publication style is entirely opaque as to which co-author contributed what. Concern about ‘honorary authorship’ — in which an author is unacceptably included for reasons other than any scientific contribution — and about this lack of transparency has led to the increasing use of statements in papers that specify authors’ contributions. Some medical journals require them, and others, including the Nature family, strongly encourage their use and may yet make them compulsory.

Such statements delineate contributions to the work but do not underwrite its integrity. Something more is needed.

It is too glib to state that every co-author of a paper shares full responsibility for its content. A researcher who specializes in the radio-active dating of rock strata cannot necessarily be expected to vouch for a palaeontologist’s analysis of fossils within them — especially if the work has been carried out in labs on different continents.

The fact that simple trust may no longer suffice is a sad reflection on recent scientific history, but anything that supports public confidence in research has to be welcomed, provided that its burden is not too great. What follows is a proposal in that direction, on which we invite readers’ comments.

We suggest that journals should require that every manuscript has at least one author per collaborating research group who will go on record in a way that collectively vouches for the paper’s standards. Each would sign a statement with reference to Nature’s publication policies as follows:

“I have ensured that every author in my research group has seen and approved this manuscript. The data that are presented in the figures and tables were reviewed in raw form, the analysis and statistics applied are appropriate and the figures are accurate representations of the data. Any manipulations of images conform to Nature’s guidelines. All journal policies on materials and data sharing, ethical treatment of research subjects, conflicts of interest, biosecurity etc. have been adhered to. I have confidence that all of the conclusions presented are based on accurate extrapolations from the data collected for this study and that my colleagues listed as co-authors have contributed and deserve the designation ‘author’.”

Principal investigators traditionally bask in the glory of a well-received paper. We are proposing now that they willingly open themselves to sanctions that could be brought to bear should the paper turn out to have major problems.

Misconduct investigators go out of their way to spare anyone apart from the direct perpetrators, but they have indicated concerns over the degree of oversight within collaborations. If the damage to reputations were more widespread in the event of fraud, researchers would be even more fastidious about the data emanating from their labs and the due diligence they would impose. The chances of major frauds, with their disproportionate impact on the reputation of science as a whole, would be diminished.

We invite comments from readers on this editorial.

(The Nature journals’ current policies can be seen at Nature‘s Guide to Authors and at the Author and Reviewers’ website.)

Comments

  1. Report this comment

    John Bechhoefer said:

    Bravo! Having the “corresponding author” be the responsible author is a long-overdue suggestion. I would encourage Nature and other publications to adopt such a policy.

  2. Report this comment

    Mark said:

    Sadly, this kind of frauds is not someting new to us at all. I personally have encountered three cases in the past. The lesson I learned is that if you point out the problem you will get into trouble in the end.

  3. Report this comment

    Girish J. Kotwal said:

    In an era of structural biology, systems biology, and translational medicine research, multidisciplinary collaborations are a key to enhanced understanding. To expect authors to police their collaborators unless there is a reason to do so is placing an undue burden in the rising constraints in doing modern science.

    On top of that ethics is in the eyes of the beholder and increasingly it is being exploited to damage the reputation of competitors and to keep them occupied in investigations and a cloud of suspicion. Science should have a way of correcting for and exposing flaws in all aspects of reporting in a collegial manner. This should be carried out as an open dialog between the person who is alleging misconduct and the person who stands accused.

    Fair play is always ethical and providing the facts on both sides of the argument is also the best way to allow people to make their own judgement as to the extent of the error and what consequences should be borne by the accused. This applies to news reporting as much as to accusations made to institutions or funders.

  4. Report this comment

    A. Solovine said:

    The proposed statement is unenforceable. The pressures of being a laboratory head often do not allow the time to check raw data in their own laboratories, let alone others. The creation of scientific achievement is often an intimately personal thing, although to admit this is not presently in vogue. Prying will cause resentment and create a jittery community, increasingly at the beck and call of lawyers who might start by questioning the legality of making publication of publicly-funded projects conditional upon a statement of this type; it smacks of coercion.

    What is needed is to foster a culture in which honesty prevails, not bludgeon integrity by scaremongering and passing the buck. I recently was told by the Director of a well-known institute that “publication is all that counts”. He has a point. But until the manner of science communication and evaluation are changed, the pressures to publish will remain. However many edicts are promulgated to prevent cheating, cheating will occur; the more draconian the rules that seek to stop it, the more sophisticated cheating will become.

  5. Report this comment

    Dr Deepak Modi said:

    The proposal that the principal author should bear the sanctions that could be bought in the event the paper has been found to have problems is very attractive and hopefully would decrease the so called “overlooks”.

    However, with the increasing numbers of multidisciplinary papers where it is virtually impossible for the PI alone to have all the expertise and understanding of the subject. Thus it is not justified to have a single author responsible for the misconduct of others. Perhaps there should be more than one authors that would coordinately share such responsibility. In my opinion decentralizing the mentioned responsibilities to all authors classified based on area of specialization may be a better option. For e.g. as in above, a person responsible for the radioactive dating and another for palaeontologist’s analysis could be an alternative and more justified option to diminish the chances of major frauds.

    Secondly the proposed statement by no means would lay sanctions on issues realted to authorship. While the principal author may take all the responsibility of good lab practices, how would the mentioned undertaking reduce the possibility of a denied or ‘honorary authorship’ granted ? This is particularly relevant in conditions when a junior investigator is a principal author.

    We need better transparent ways, both at the institutional levels and also at the level of the journal, to minimize fraud and issues realted to authorship.

  6. Report this comment

    Attila Kofalvi said:

    1) This is a recurrent problem. I got puzzled by the opening thread as the expression “major fraud” is used. I think there is no minor or major fraud. There is only fraud. Otherwise Woo Suk Hwang’s fraud would be very light and minor for me as I have nothing to do with stem cell research, it indeed did not turn upside down my laboratory’s life.

    2) What if each subheading of an article had the initials of the contributing authors at its end? It would be clear then who participated in what. Seniors and directors may appear only in the Discussion. Assistants would appear in the subsections of Methods. Still, there are seniors (president, director etc of an institute) who raise a lot money for the research groups or provide the daily functionality of the institute itself. Their contribution to the article is metamaterialistic. Without them the article would not be born, they deserve honorary authorship, but not any responsibility for the data in the article.

    (Maxine notes: to specify author contributions is strongly encouraged by Nature and the other Nature journals as part of our existing policy. About half of papers we publish carry these statements.)

  7. Report this comment

    Jean-Claude Bradley said:

    If scientists make their laboratory notebooks public, it would resolve a lot of these issues of responsibility.

    I don’t think the corresponding author can be held responsible for re-analysis of all the raw data of their collaborators. Often the reason you have collaborations is that you require someone’s expertise in another field. If you don’t then why not do it yourself?

    But if the raw data and laboratory notebooks are public, at some point, someone who is an expert in the field will be confronted with trying to repeat the experiments and will have to redo the analysis. At that point, meaningful feedback can be obtained and a discussion can take place in the open.

  8. Report this comment

    Phil Bentley said:

    I know people who are being forced, from the highest levels, to lie on documents and who are threatened with disciplinary procedures if they refuse. I also know of numerous cases of “stolen” data, ideas and projects, and highly questionable order of authorship on papers. Does anyone actually believe that yet another piece of signed paper will prevent dishonesty in the presence of mismanagement, and over-reliance on impact metrics and fashionable keywords?

  9. Report this comment

    Shi V. Liu said:

    I think Nature’s proposal is unnecessary since every author, not just the leader of each research group, is already responsible for telling the truth. “Honorary” authors put their names on papers because they know that this misbehaviour is not “misconduct” and will bring them no real harm. Many institutions protect these grant-earning “principal” investigators so that their financial support to the institutions will be maintained. This protection is often not afforded to junior researchers.

    Many problems of misconduct would be solved by making scientific publishing completely transparent and open for public criticism/comment. For example, journals could publish as many names as the authors of the paper wish, but allow the public to challenge the role of any author of a publication. If a name on the paper is found as “honorary” or irrelevant then the name should be removed. and the fact of the removal recorded.

  10. Report this comment

    Sonia Vasconcelos said:

    As several other research integrity-related issues, publication ethics needs to be encouraged and discussed at the educational level. That journals have a major role in this discussion is a fact, but researchers and professors at universities should share the responsibility of promoting responsible co-authorship.

    Newcomers in scientific research, if not aware of their role in a scientific paper, might simply assume that gift authorship, ghost authorship or even reciprocating favors for previous gift co-authorships are OK. Concerning doctoral students, pressure to publish one, two or more papers before defending the thesis cannot be an excuse to overlook their role as co-authors of a manuscript. They should develop such awareness before getting their academic credentials and becoming established researchers. Included in this educational guidance should be discussions on authorship ordering.

    It is true that different academic traditions may have different authorship policies and interpretations of the meaning of the first and last positions and of the corresponding author. However, supervisors and graduate students should discuss this frankly. To me, the better the research environment the greater the possibilities of developing responsible research and authorship.

    The whole academic community has its share of responsibility, but, in my opinion, editors and educators have a decisive role in this matter, and they should combine their efforts to promote publication ethics.

  11. Report this comment

    V D Ramanathan said:

    One way of avoiding giving honorary authourship and its opposite and more subtle evil of excluding somebody who has contributed enough to warrant authourship is some kind of a guarantee from the head of the institution where the work has been done.

    This will offer a measure of protection.

    However, the only way fraud can be prevented will be by attaching much less importance to publication as a measure of reward.

  12. Report this comment

    Abhay Sharma said:

    Prevention is always better than cure. While there is no doubt that evolving more effective criteria of accountability is needed, this exercise will be most useful to identify the culprit once a paper has been found to contain fraudulent data. Many papers with such data may go unnoticed.

    Further, how much it will prevent attempts of misconduct at the first place itself is doubtful. Here I would like to suggest that scientific community evolves a research equivalent of classical or modern versions of “Hippocratic Oath” traditionally taken by physicians. Although a Hippocratic Oath may not be necessary for conducting ethical practice of medicine (see BMJ 1994;309:952-953 for a discussion), introducing a research oath would certainly constitute an additional preventive measure against scientific misconduct.

  13. Report this comment

    James L. Sherley said:

    I fear that the editors of Nature may be throwing the proverbial baby out with the bath water. What do two high profile cases of published scientific frauds tell us about the state of integrity in submissions of original scientific reports on the whole? Possibly nothing at all. Before further encumbering already taxed authors with more barriers to efficient submission of their findings, the editors should consider a well-designed quantitative evaluation of the determined or estimated cases of publication fraud. In any event, the most effective means to keep fraud to a minimum is fostering a generally high degree of training and practice for responsible and ethical conduct in the scientific community and ensuring complete, rigorous, and ethical review of submissions by referees and editors. Putting more unenforceable paperwork in place will not change unethical behavior; but even more detrimental, it may give editors an unwarranted belief that a high level of reviewer and editorial scrutiny is no longer necessary.

  14. Report this comment

    Qoot Alkhubaizi said:

    When it comes to research in Third World, it is often a common practice to find the principal investigator taking the role of the facilitator of the ‘political’ permits which are required to complete such research. In other words, the main person who needs to complete the research will be obliged to include that person’s name so that s/he can complete the research.

    How can you tackle such a situation when the PI has no clue of what the research is all about? Some scientists have to lie about the amount contributed by their PI so that they can get their research completed and submitted for publication.

  15. Report this comment

    N Periasamy said:

    The present review process of a manuscript does not ensure that the editor’s and referees’ comments are accessible to all authors. Therefore, I propose an addition to the proposed statement by the corresponding author, along the lines of: ‘I promise to ensure that all authors receive copies of the editor’s and referees’ comments, irrespective of the success or failure of the publication of the manuscript in this journal.’

  16. Report this comment

    Lutz Slomianka said:

    Statements like the one that Nature now suggests just add another signature to the ones already required when submitting a manuscript. Most likely it will be one of the signatures, which like some of the others, is being provided without too much thought and with not much more than a glance at what is being signed.

    Ethical standards in the research community have apparently deteriorated to a degree at which it is necessary for journals to consider that scientists could be scoundrels. Yet another signature will not provide anything but another leaky patch on the injury that we, as scientists, journals and politcians have inflicted on science: the creation of a scientific environment in which a patient, honest, conciencious scientists not using his elbows, not striving for power, only taking as many students which he can carefully supervise, without the talents of a used car salesman, dedicated to a problem indepent of it being the flavor of the day and able to say “I have been wrong” would no be out of a decent job very quickly but, today, unlikely to ever obtain one.

    I am so immensely tired of statements and concerns about ethical conduct. For years and years it has been lip service by those who could change things if they wanted to, but who would not have attained this position of power they have if they would not have tacitly accepted or tolerated the princple that causes the problem: the forces of natural selection we have created.

    Now, try to beat natural selection by a signature! Good luck.

  17. Report this comment

    Allan Reese said:

    As a statistician, I’m worried that authors certify “the analysis and statistics applied are appropriate.” My experience is that many papers get published with inadequate or wrong analyses, not picked up by authors, reviewers or editors. There is a severe need for actual statistical reviewers.

    Personal hobbyhorse. Most authors can’t draw graphs and most readers don’t care. Graphs are treated by readers as images, to be taken in with a superficial glance. Authors treat them the same or as quasi-tables, expecting them to be used for looking up individual values. I collect examples showing graphs have not been proofread. Great scope for improvement, but who cares?

    Maxine notes: The Nature journals use statistician referees regularly. We have a statistical checklist which is sent to authors before acceptance for publication and a detailed page of advice about common statistical treatments and errors in our guide to authors – on this page we ask for suggestions and feedback. See http://www.nature.com/nature/authors/gta/index.html

  18. Report this comment

    Allan Reese said:

    On the actual topic: the editorial and comments to date seem to conflate problems of:

    – genuine attribution of work (pulling rank, honorary authorship)

    – honesty of work (certifying results as genuine, counteracting fraud)

    – accuracy of work (competence in techniques, certifying results as accurate)

    – avoiding errors through group-think and committee anonymity.

    The first two appear endogenous to the culture of the particular workplace; the second two might be detected by peer review. Writing to another editor, I suggested reviewers might allocate their opinions into (1) practical science, did the work appear competently and honestly carried out (2) intellectual science, ie novel and interesting (3) data processing, the reporting and analysis of results (4) interpretation and context. Attribution is also obscured by the choice of passive voice. I once read a PhD abstract that was literally “Samples were collected, samples were analysed … Conclusions were drawn.” It was not clear which, if any, of the steps had been carried out personally by the student!

  19. Report this comment

    Torben Clausen said:

    It is difficult to share responsibility, mainly because it often gets diluted. The Danish poet Piet Hein once remarked that when two persons share a responsibility, there may be only one per cent for each of them. Something similar may apply to publications, in particular when their messages turn out to be doubtful. When the number of co-authors is increasing further, the dilution of responsibility gets even worse. Moreover, it gets more difficult for each co-author to ascertain that all contributions to the manuscript justify co-authorship or live up to a high scientific standard.

    The number of co-authors per paper has continued to increase for many years.

    About 10 years ago, less than 5% of the papers in New England of Medicine had one author, whereas 100 years earlier 98% of the papers in the same journal had one author (Shapiro et al., JAMA, vol. 271, pp 438-442, 1994). The same study showed that among the co-authors, the contribution to the papers differed by orders of magnitude. In the Journal of Biological Chemistry, the average number of authors per article increased from 1.5 in 1907 to 5.1 2007. In most journals, this inflationary trend continues, in spite of repeated proposals to limit the increase by defining requirements for co-authorship.

    What drives this increase? It is frequently stated that it is due to uncritical appointment of “honorary authorships” to senior colleagues with reputation. More often, I believe, co-authorships are offered to the more numerous junior scientists in order to recruit them to continue to work in a research group. In either case, co-authors with limited insight into the publication may find themselves formally accountable for the content and the validity of the paper. However, their risk is limited and they get credit for an increased number of publications.

    What keeps it going is the apparently ineradicable tradition to assess an applicant’s scientific merit from simple counts of her/his total number of publications. You may be the co-author of 50 papers, but if the average number of co-authors on these papers is 5, your average production only amounts to 10 papers. This adjustment to the overall evaluation is almost as simple as the sheer counting of papers. Further appropriate adjustments are to count how often the applicant is first or last author and how often the papers have been quoted.

    When taken together, these simple rules of thumb often provide a more realistic evaluation. If commonly used in the evaluation of applications for grants or positions, they may improve the reliability of the research and sharpen the definition of the criteria for co-authorship. Of course, you can only be responsible for your own research input, but to justify co-authorship, your contribution should be of a reasonable size, comparable to that of other co-authors. It should be possible to define this in the paper, just like in the Ph.D. theses submitted to our universities.

    In my opinion, the signing of declarations as proposed in “Who is accountable” provides little or no protection against fraud. A better protection could be obtained by indicating the relative size (in per cent) of the contribution of each co-author at the end of each publication. This would put a limit to the dilution of responsibility.

  20. Report this comment

    Ravi Joshi said:

    Declaration in whatsoever format may not solve the problem of inclusion of senior professor’s/officer’s name as a co-author. One can easily see an inappropriate increase in the number of research papers/articles/reviews by a researcher as soon as he/she becomes administrative/officiating head of a group of researchers, atleast in India. I would like to suggest that such researchers should be identified who get a spurt in the number of research articles/papers/reviews because of some sort of pressure they exert related to honour, funding, carrer assessment, etc. on the junior researchers.

  21. Report this comment

    Dwijendra Singh said:

    To avoid fraudulent publications, the main and each co-authors should sign a letter upon submitting the paper to a journal, and include a one-line statement of how each author has contributed. Journals can insist on both these conditions. A signature of each individual author at submission would exclude authors who have not contributed significantly and are taking undue advantage.

    Second, in some research organizations, each co-author has to sign a prescribed form before submission of the paper. This is a useful way to monitor authorship.

  22. Report this comment

    Francois Eisinger said:

    It is not all about Science, it might be about Justice.

    True genuine science can only leads to unreprodicible results. My high school physics teacher used to say « all things considered equal» while my philosophy teatcher answered « you cannot step twice in the same river» (Heraclite’s law). Inconsistent results might be attributable to many causes among which fraud. Science for that concern might follow an almost universal law, the slippery slope phenomenom. We do observe undisputable misconduct on one side and, on the other extreme position of the range, a crystal clear trustful research.

    Is it possible to draw a red/yellow/white line discriminating it? Could we figure out a process that allows us to give always a clear answer to the issue where this author stands? Or alternatively should we ask for the logic of the courtroom 1 with at least two standards: « more likely than » used in civil trial or « beyond any reasonable doubt » used in penal trial.

    Francois Eisinger MD Sophie Eisinger Ms Law

    1. Root, D. H. Bacon, Boole, the EPA, and scientific standards. Risk Anal 23, 663-8 (2003).

  23. Report this comment

    Kim said:

    I seriously doubt another piece of paper will make a difference:

    From my own experience, fraud is routinely committed by individuals who are so arrogant that they do not care that the data they are publishing is fraudulent, only that their that they get another article under their belt.

    The tragedy is that when the activities of such a person were brought to the attention of the director of the large high profile institute I work, the entire thing was hushed up and the whistleblowers ostracised.

    Only a fundamental shift in attitude that places scientific rigour above egos and reputations will prevent fraud.

  24. Report this comment

    Yongxian Lu said:

    when the 1st response of the people interviewing you is to look for Natures, Sciences in your pub list, nothing will change whatever new policies you adopt. so the really fundamental change should be the way of how to evaluate a scientist’s achievement, not really how to make them more honesty.

Comments are closed.