dataThese are some notes for my future reference on editorial policies of major scientific publishers on retraction. Most publishers have retraction policies (see: Resnik, D. B., Wager, E., & Kissling, G. E. (2015). Retraction policies of top scientific journals ranked by impact factorJournal of the Medical Library Association: JMLA103(3), 136.).

Many, if not most, publishers rely on guidance provided by the Committee on Publication Ethics (here in PDF), and more generally here. COPE states: “Retraction is a mechanism for correcting the literature and alerting readers to publications that contain such seriously flawed or erroneous data that their findings and conclusions cannot be relied upon. Unreliable data may result from honest error or from research misconduct.”

Here are some guidelines from different publishers, with my present focus on cases of flawed data that underpins published results (there are obviously other reasons for retraction):

  • Springer: “Retractions are published to set the record straight. Where there are severe errors in an article that invalidate its conclusions, the article may need to be retracted. . .  A retraction is necessary when: Unreliable data mean that the results and conclusions cannot be trusted – both genuine mistakes by the authors and scientific misconduct including data fabrication.”
  • Elsevier: “The retraction of an article by its authors or the editor under the advice of members of the scholarly community has long been an occasional feature of the learned world.” Its editorial approach relies on guidance of the International Committee of Medical Journal Editors (ICMJE), which states: “Errors serious enough to invalidate a paper’s results and conclusions may require retraction”
  • Nature: “Notification of invalid results that affect the reliability of a previously published article.”
  • Science: “Science is a member of COPE (Committee on Publication Ethics) and all Science Journals are committed to correcting errors in published papers. . . In cases of irreproducibility of research findings reported in a Science Journal paper, a retraction may be considered if the core conclusions are thereby invalidated.”
  • BMJ Journals: “Retractions are considered by journal editors in cases of evidence of unreliable data or findings, plagiarism, duplicate publication, and unethical research. “

Bottom line: a paper that relies on erroneous data that calls into question the integrity, validity or trustworthiness of reported results should be retracted. There appears to be a strong consensus here, not surprisingly, under the COPE guidelines.