Ethics
The Journal of Infection in Developing Countries (JIDC) will be launched in a few months and is already receiving and processing a number of articles. Many of the articles we receive are sent directly for peer review. Others require pre-review mentoring, a unique service that JIDC is committed to provide in order to overcome some of the documented biases against developing country science (Horton 2003). The JIDC editors would like to emphasize our stance on the following important aspects of publication ethics in the face of having received a few articles in violation.
Articles that violate the ethical norms of scientific publishing have so far been very few and have come from well-meaning scientists who are unfamiliar with publishing etiquette and ethics. However even unintentional misrepresentations, duplications and other violations can be viewed as misconduct, and are therefore harmful to journal and author reputation. Below we summarize some of the more common violations associated with scientific publication as a guide for prospective authors, reviewers and editors. The JIDC editorial and mentorship offices are also happy to address any pre-submission questions on publication ethics.
Scientific misconduct
It is generally presumed that data and interpretations of data presented scientific manuscripts is generated, to the best knowledge of all authors, to the highest quality and integrity. Rigorous scientists perform checks on their data to ensure that it is valid, accurate, and robust. Deliberate fabrications or falsification (including data enhancement) amount to scientific misconduct or fraud. Whilst there is evidence to suggest that the vast majority of scientific publications do not contain errors or fraudulent data, the number that do probably exceeds the number that have been documented. Should any author discover that fraud or negligence has resulted in the submission or publication of false data, a manuscript must be withdrawn or retracted, even if such a discovery occurs after publication. Whilst post-publication retractions due to scientific fraud are becoming increasingly common, most retractions arise from unintentional error (Nath et al. 2006). If authors discover that they have made a mistake that alters specific data or statements but does not change the overall conclusions of the paper, the error must be corrected by communication to the editor, or in the case of a published article, by a corrigendum or an erratum published in the same journal.
As indicated in the Instructions to Authors, the use of human and animal research subjects must be subjected to appropriate ethical review. When financial or other compensation might bias an investigator's handling or interpretation of data, readers need to know this in order to read the significance of the findings in appropriate context. Conflicts of interest should be avoided or declared.
Plagiarism
Plagiarism is essentially misrepresentation of a work, or part of a work as one's own (see Benos et al. 2005 for a review). When an idea is taken from another source, it must be appropriately cited or acknowledged. Text as well as ideas can be plagiarized. Vessal and Habibzadeh (2007) have observed that unintentional text plagiarism is more common among non-English speakers who borrow phrases, blocks or even pages of text because of the desire to communicate accurately in the face of difficulty in creating their own prose. While this reasoning can be appreciated, the line between text and idea plagiarism is too fine for the former to be accepted and the consensus in the scientific community is that plagiarism of any kind amounts to misconduct. JIDC is willing to help mentor or copy-edit articles by authors whose first language is not English. As the journal aims to provide a route to publication for all developing country authors, irrespective of first language, we routinely process manuscripts that need mentorship or editing to strengthen English composition, but cannot consider fully or partially plagiarized manuscripts. Electronic sources and software have made plagiarism easier to perpetuate, but also easier to detect, particularly in the case of text plagiarism. JIDC will routinely screen manuscripts for plagiarism. Our mentoring office will insist that, where detected, unintentional plagiarism is remedied before a manuscript can enter the peer-review process. Authors however have the ultimate responsibility in ensuring that their ideas and text are original.
Duplicate publications
Each primary research article is presumed to be an original, unpublished piece of work. Prior communication of research findings at scientific conferences is permissible but findings can only be published once. JIDC cannot accept previously published papers, even if those papers were published in regional or non-indexed journals. Similarly, duplicate submission is unethical and is a frequent cause of duplicate publication. A manuscript can only be under consideration by one journal at a time. If an author elects to submit to a new journal, he or she must withdraw any prior submissions.
Overlapping publications
A research paper manuscript cannot include findings that have been published in a previous paper. If those findings are needed to interpret new data, they should be cited and referenced accordingly, and not reproduced. When part of one paper is reproduced in another, the two papers are said to be overlapping and raise ethical concerns. Review papers may overview data from research papers but this must be appropriately cited and, where applicable (such as for figures) appropriate permissions must be secured.
Publication siblings
A research study should be published in its entirety unless there are valid reasons for not doing so. The excessive fragmentation of data (some times referred to as 'salami publication' in analogy to thin slices of processed luncheon meat (Gollogly and Momen 2006)), is cumbersome to readers in the field and is damaging to the author's reputation. Whilst fragmentation is not an ethical violation per se, it does not enhance a scientist's career in the long term. JIDC receives many manuscripts where, due to funding or other constraints, a slice of data represents the entire study. The publication of such papers does not constitute 'salami publishing'. Authors with papers of this nature should present them as a short communication. In the case of extensive or long term or very large studies, multiple publications are impossible to avoid. Each publication must however acknowledge the existence of the others and there should be no data overlap.
Authorship
Authors have made a substantial intellectual contribution to the work, such as study design, data collection, data analysis and interpretation. Authors also contribute to the presentation of the data and writing the paper. People who provide only lab space, instrumentation, funds, routine technical assistance, managerial services or paid services do not qualify for authorship (Macrina 2005). Where a contribution is technical but not intellectual, it may be acknowledged. Similarly, minor suggestions can also be acknowledged as 'helpful discussions' or ideas. It is unethical to include co-authors who did not contribute to the research (so called 'gift authorship') just as it is unethical to omit collaborators who deserve co-authorship ('ghost authorship') (Gollogly and Momen 2006). All authors and persons acknowledged should approve of the final version of a manuscript before it is submitted. They should be satisfied with their position as authors or in acknowledgement, and, where applicable, with their placement in the author list. Whilst co-author contributions to writing may vary, all take responsibility for a manuscript's contents (ICMJE 2006). Authors are credited for the findings reported in a paper but can be held liable for scientific misconduct, plagiarism and other ethical violations.
Peer review
Collaborators, recent co-authors, and relatives should not peer review one another's papers. Similarly, if a paper is in competition with the reviewer's work, accepting it constitutes a conflict of interest (Macrina 2005). Editors typically watch out for the most obvious conflicts by not selecting reviewers who have been recent co-authors, who are acknowledged by the authors, or who work at the same institution of any of the authors. Editors cannot however identify all potential conflicts. A conflict of interest may also arise if the reviewer is conducting the same research contained in the paper and his or her work would no longer be novel if the paper under review is published. If an editor requests a peer review from someone who has or may have a conflict of interest, the reviewer should decline to review the manuscript. Reviewers should not communicate with authors during the review process. Questions or clarifications should be requested through the editor.
Further reading
The following resources provide important information about scientific integrity and the ethics of authorship: Gollogly and Momen 2006; Benos et al. 2005 & Macrina 2005.
Literature cited
Benos DJ, Fabres J, Farmer J, Gutierrez JP, Hennessy K et al. (2005) Ethics and scientific publication. Adv Physiol Educ 29(2): 59-74.
Gollogly L, Momen H (2006) Ethical dilemmas in scientific publication: pitfalls and solutions for editors. Rev Saude Publica 40 Spec no.: 24-29.
Horton R (2003) Medical journals: evidence of bias against the diseases of poverty. Lancet 361(9359): 712-713.
ICMJE (2006) Uniform requirements for manuscripts submitted to biomedical journals: writing and
editing for biomedical publication. Available: http://www.icmje.org.
Macrina FL (2005) Scientific integrity: text and cases in responsible conduct of research. Washington, D.C.: ASM Press. xxv, 402 p. p.
Nath SB, Marcus SC, Druss BG (2006) Retractions in the research literature: misconduct or mistakes? Med J Aust 185(3): 152-154.
Vessal K, Habibzadeh F (2007) Rules of the game of scientific writing: fair play and plagiarism. Lancet 369 (9562): 641.
Wednesday, May 16, 2007
by Iruka Okeke
