Case Summary
The author in this case was a journal editor-in-chief, who submitted a manuscript to another journal in the field. After peer review, the author was asked by the target journal’s editor-in-chief to revise the manuscript and cite two references from that journal. However, the references were only marginally related to the manuscript’s topic and their inclusion would have helped increase the target journal’s impact factor.
Question(s) for the COPE Forum
• Is using the manuscript review process to increase citations of your journal ethical?
• In a related question, is writing editorials which cite large percentages of manuscripts from your journal ethical?
• In summary, is it ethical to use these tools to influence the impact factor metric?
Forum Advice
The Forum replied No to all three questions. Using the review process to gain journal self-citations, especially by irrelevant citation, was deemed unethical. It was unclear if in this case the additions were suggested by the reviewers along with other references, or coercively required by the journal editor. Still, the author could decline or find better references. In addition, editorials should not be used to manipulate impact factors through journal self-citation, but using an editorial to cite and introduce articles in an issue would not be citation manipulation.
The Forum noted that Clarivate Analytics takes action against journals for citation manipulation. However, because up to 15% of citations are allowed to be journal self-citations, editors could purposefully promote journal self-citation without exceeding that level. A link to the Citation Manipulation section of a Council of Science Editors white paper was provided
(https://www.councilscienceeditors.org/resource-library/editorial-policies/white-paper-on-publication-ethics/2-1-editor-roles-and-responsibilities/#215).
Case Discussion
This case is categorised under COPE’s Core Practice of Allegations of Misconduct, which states: “Journals should have a clearly described process for handling allegations, however they are brought to the journal's or publisher’s attention. Journals must take seriously allegations of misconduct pre-publication and post-publication. Policies should include how to handle allegations from whistleblowers”. This would include allegations of citation manipulation.
The case is also categorised under COPE’s Core Practice of Journal Management: “A well-described and implemented infrastructure is essential, including the business model, policies, processes and software for efficient running of an editorially independent journal, as well as the efficient management and training of editorial boards and editorial and publishing staff.” This would include policies and training related to citation practices and how to make suggestions of changes to the references.
Knowledge creation is influenced by, builds on, or refutes what has been discovered before. This process is reflected in knowledge documentation, in the form of peer-reviewed publications, by way of professional argumentation and appropriate use of and referral to past literature, known as academic or scholarly intertextuality. Some of COPE’s Core Practices deal with related areas of authorship, intellectual property including plagiarism and redundant/overlapping publication, and ethical peer review.
Another area of scholarly intertextuality is ethical citation (indicating sources within a research paper’s text) and referencing (giving full bibliographic details in the reference list at the end). The case provides a reminder that everyone involved in the scholarly enterprise is expected to know how to present work ethically, cite sources appropriately, and reference those sources to allow others to verify information and read further. A degree of self-citation is expected, but adding or suggesting the addition of self-citations, journal self-citations, or indeed any other citations, to a research paper for non-academic reasons is unethical. Conversely, deleting or suggesting deletions of citations and references for non-academic reasons is unethical.
COPE has now released a Discussion Document on citation manipulation to guide journals in formulating relevant policies, guidelines, and training, partly based on past COPE Forum discussion documents on citation manipulation and self-citation. It would be unrealistic and impractical to impose universal caps to self-citation rates, and the amount allowable may be discipline-, field-, or topic-specific and also depend on each paper/study. Imposed caps for author self-citation, citations from reviewers/editors, and journal self-citation could actually encourage those practices. The advice in the new COPE discussion document is for journals to formulate policies “even in broad terms about acceptable self-citation”, appropriately suggest any references, and have procedures to handle (attempted) citation manipulation.
The author in the described case could have explained why the two references were not going to be added, given their low relevance. Whether the recommendation came from the editor or the reviewers, the author could quote COPE’s peer review guidelines, which require reviewers not to recommend citation of their or their colleagues’ work to boost citation counts. If the journal editor insisted or made the additions a condition of acceptance, the author could use his/her experience or influence as an editor-in-chief of a COPE-member journal to explain why coercive citation and citation manipulation are unethical and to request a second opinion or an impartial investigation. Ultimately, editors or reviewers who are found to systematically engage in coerced citation can be replaced (eg, as in cases 18-03 and 19-01).
There are legitimate reasons for authors to cite and reference their past works, for reviewers and editors to suggest their own references to authors, and for reviewers and editors to suggest references from their journal. However, authors, reviewers, and editors should make sure that all citations are relevant, used correctly (eg, not misquoted), and add to the academic argument and scholarly value of a paper. As stated in an August 2018 COPE Digest guest post by Clarivate Analytics (Marie E. McVeigh and Nandita A. Quaderi, Citations: Link, Locate, Discover, Connect), “Review of the purpose or academic value of citations within an article is the work of reviewers and editors to ensure the integrity of the scholarly content they publish.”
Unfortunately, citation manipulation has become a pervasive problem because a citation (ie, reference) count has become a currency in itself. It is often forgotten, however, that a source may be cited multiple times in the text (but appear only once in the reference list). Furthermore, qualitative aspects are ignored. Mentions of another source can be negative, positive, or neutral, and a citation could be presented as background or be of critical importance to the interpretation or conclusion of a study.
Still, journal-level citation metrics are widely quoted as an indicator of journal influence, but also widely misused in research evaluation systems. They are known to be prone to manipulation, such that Clarivate Analytics suppresses titles from its Journal Citation Report for excessive journal self-citation or citation stacking to help inflate another journal’s impact factor. Similarly, researcher-level metrics such as the h-index have raised the importance of article-level citations, but have also promoted self-citations. Efforts to raise awareness of the misuse of such metrics in research evaluation include the 2015 Metric Tide report and associated website, Leiden Manifesto, Declaration on Research Assessment (DORA), and, most recently, the Hong Kong Manifesto.
Finally, it is known that journals can take advantage of the impact factor calculation by routinely reviewing and/or listing their own articles in editorials, commentaries, or reviews. Current-issue self-mentions, such as an editorial introducing an issue’s contents, would not affect the latest score because only the previous year’s issues are inspected (eg, the 2018 impact factor was calculated in 2019 and counted citations made to 2016 and 2017 articles). However, articles that review the previous year’s or two years of articles would contribute to the score in the following year. As illustrated in case 04-36, it is doubly bad practice to pretend to review the field but mention only the same journal’s content from recent years so as to raise the impact factor. This practice can be easily detected and should be discouraged.
Trevor Lane on behalf of the COPE Education Subcommittee