Promoting Integrity in Science Journals

Integrity begins with a standard, something that's currently lacking, says the Council of Science Editors.

By | January 1, 2007

<figcaption>Disgraced South Korean researcher Woo-suk Hwang. Credit:   AHN YOUNG-JOON / ASSOCIATED PRESS</figcaption>
Disgraced South Korean researcher Woo-suk Hwang. Credit:   AHN YOUNG-JOON / ASSOCIATED PRESS

Ethical controversy and reports of scientific misconduct have received substantial attention recently. Are researchers behaving unethically more than they used to? Are technologies both enabling such misconduct and making it easier to catch? Are there truly more such misconduct and breaches of ethics relative to the amount of research performed and reported?

Answering these questions would require duplicating all past research and subjecting it to today's more-intense examination, a clearly impossible task. What we can say definitively is that research is disseminated more widely and rapidly than in the past, and thus any problems, especially with important discoveries that can greatly impact society, receive much more scrutiny. The stakes seem higher than ever before, both for the career advancement, prestige, and financial support of researchers reporting the discoveries, and for science itself, because unfavorable attention can erode public trust, ultimately decrease funding, and slow research progress.

No common standards of acceptable practice exist for reporting original research that transcend disciplines or cultural and geographical boundaries. It falls to science journal editors to be responsible for promoting practices that minimize errors, taking action when allegations are brought to their attention, and communicating to the readership when corrections or concerns are needed.

The Council of Science Editors aims to provide editors with a resource on which to establish and benchmark a journal's policies and practices. Our White Paper on Promoting Integrity in Scientific Journal Publications (available at covers roles and relationships among editors, authors, reviewers, sponsors, journal owners, and the media; and how research misconduct can be identified and handled. While many aspects discussed may be obvious, they are often not put into practice by journals, and so there is a lot of opportunity to improve ethical publication practices.


Editor responsibilities towards authors, reviewers, readers, and board members should include providing the journal's policies and guidelines. These can include submission requirements, such as protocols for animal and human research, data sharing, identifying funding sources and disclosure of conflict of interest; registration of clinical trials, adherence to guidelines (e.g., CONSORT, MOOSE, QUOROM, Declaration of Helskinki); defining authorship and contributorship; establishing effective, fair, and rapid methods for peer review, editorial decisions, and publication; maintaining the integrity of the literature by quickly publishing errata, retractions, and statements of concern; adhering to an agreed-upon mission, publication practices, and schedule; and operating in a fiscally sound manner.

Authorship or Contributorship models are common elements among journals. Adopting a combination of both models can prove beneficial. Many journals, particularly in medicine, follow the International Committee of Medical Journal Editors (ICMJE) criteria for authorship. The contributorship model is gaining acceptance as an alternative to the defined authorship approach because it is more specific, and sometimes more flexible, in identifying individual roles, responsibilities, and accountability.

Regardless of the model chosen, rules surrounding multiple submissions, compliance with requests of funding agencies, transferring copyright, and authorship disclosures need to be explicit.

Reviewers, who play a key role in publications by critiquing the research reported in the manuscript, are expected to provide timely and unbiased critical comment to the editors in areas of their expertise, comment on any ethical concerns with the research or the publication, and disclose any conflicts of interest. Peer review continues to be a confidential, privileged communication.

Sponsor roles vary in the publication process, but might include financial or in-kind support of the research or the publication. The author's relationship with sponsors (e.g., grantee, consultant, employee) should be disclosed. Some journals ask authors to list the specific roles sponsors played in the research or the creation of the manuscript to grant authors full access to data.

Publishers, sponsoring societies, and journal owners (referred to as owners) should confer and agree upon roles, policies, and practices with their journal editors prior to engaging them. Owners are generally responsible for the financial, business, and other management issues, while "editors must be free to authorize publication of peer-reviewed and other appropriate research reports, as well as society news, appropriate advertising, and other materials." This extends to the right to review and reject advertising and placement. It remains important to keep editors from as many commercial, political, and other influences as possible that could affect editorial decisions, which should help maintain editorial integrity.

Media coverage of published science extends the reach of the findings beyond the journal readers to the public. Journals can aid media outlets through press releases that give background information and study limitations in concise, everyday language, advance materials to allow preparation of article before an embargo date, answering questions, and referring reporters to authors and other experts.


A variety of factors have yielded different definitions of misconduct. However, most definitions include falsification, fabrication, and plagiarism. Negligence and deceit are usually central to the definition, so honest errors typically do not rise to the level of misconduct. Image manipulation is one area of concern where additional awareness is needed about the boundaries of acceptable and unacceptable manipulation. The guidelines of Rockefeller University Press are presented and outline proper handling of image data and how editors can establish policies to identify and handle suspected manipulation. Interdisciplinary or international collaborations may contribute to problems unless there is clear communication and agreement about accepted practices before these collaborations begin.

Not many countries have an established national means to respond to allegations of misconduct: those of the US and Denmark are two of the most highly developed. Within the US, the NIH Office of Research Integrity (ORI) and the National Science Foundation address misconduct of federally funded research most actively, and may impose sanctions (as may many institutions and scientific societies). However, the employing institution generally performs the primary misconduct investigation unless allegations involve many institutions, conflicts of interest exist, or institutional resources to do so are inadequate. National bodies more often serve review or appellate functions. Editors, who are responsible for maintaining the integrity of the literature, should be advised when a misconduct investigation relates to either a pending manuscript or one already published.

Misconduct allegations
Misconduct allegations may come from any source. If the editor concludes the allegations are credible, she/he might notify any or all of the research participants, from authors to funders; the WhitePaper describes preferred approaches.

Correcting the literature
Established policies for literature corrections are not standard. The common choices of erratum, retraction, or statements of concern are all viable options. If a finding of misconduct is made, the offending scientist typically is required to request a correction or withdraw the manuscript if that has not already occurred.

Errata and retractions as a result of misconduct come from various sources. An offending author may refuse to submit an erratum or retraction and the White Paper gives examples of responses when this has occurred. Statements of Concern typically come from editors.

Editors can do a lot to promote good publication practices, but their efforts alone are not enough to make a desirable difference in minimizing misconduct. Education systems at many levels, and in particular those where research is performed, must engage in instruction and oversight of defined acceptable research practices for their institutions and for the collaborations their researchers engage in and report. Other institutions or communities that influence science and its reporting practices (e.g., sponsors, publishers, government and funding sources, professional and scientific associations) must continue to play major roles in establishing responsible and ethical publication practices.

Diane Scott-Lichter is Vice President and former Chair of the Editorial Policy Committee of the Council of Science Editors. She is Vice President and Publisher, Blackwell Publishing.


Avatar of: Hugh Fletcher

Hugh Fletcher

Posts: 44

January 11, 2007

A good way to check erroneous science would be for journals to provide an online discussion forum rather like the one here where anyone could make legitimate comments and ask searching questions which would automatically be appended to the on-line material. The authors could succinctly reply too. This would allow the whole scientific community to share the reservations of particularly expert or observant individuals, and speed the progress of science by promoting debate. Currently, journals generally do not accept correspondence pointing out inconsistencies or alternative interpretations of things they have already published. They hide behind referees who may have been either superficial or excessively picky, and who often have a bias as competitors or colleagues in the same field.\nJournals could take some pressure off potentially honest authors by being more discriminating about reviewers? criticisms. A bad example is being asked to make an irrelevant change to procedure which would involve completely repeating 3 years difficult work, and was impossible in practice. The paper could have been published with the simple statement that this particular thing/method was not done, and allow peer readers to evaluate it. The temptation to say ?Did that, no difference? could be hard to resist when one?s future depends on it.\n
Avatar of: Michael Pyshnov

Michael Pyshnov

Posts: 10

January 16, 2007

To someone who works with policies and procedures, the problem may seem to be in the lack of standards. However, when a concrete case is studied, the problem appears to be the systemic dishonesty in academia at all levels. When the victim of fraud and plagiarism complains, he/she quickly finds that out. While procedures can indeed be deficient, the standards do exist since time immemorial, and, with ANY procedure, it would be easily possible to stop fraud in academia. A much better way to restore integrity of the system is to publicise every case rather than making improvements in the procedures ad infinitum. There is a proof of what I am saying, in the documents, at:\n\n\nMichael Pyshnov.

Popular Now

  1. Running on Empty
    Features Running on Empty

    Regularly taking breaks from eating—for hours or days—can trigger changes both expected, such as in metabolic dynamics and inflammation, and surprising, as in immune system function and cancer progression.

  2. Gut Feeling
    Daily News Gut Feeling

    Sensory cells of the mouse intestine let the brain know if certain compounds are present by speaking directly to gut neurons via serotonin.

  3. Athletes’ Microbiomes Differ from Nonathletes
  4. Government Nixes Teaching Evolution in Turkish Schools