Cracking Down on Coercive Citations

Citation manipulation manifests in different ways, some of which have been discussed in previous blog posts here and here. A report published in Nature in September 2019 addresses coercive citation, the practice of reviewers encouraging authors to cite the reviewers’ own research in exchange for positive reviews. While some citation overlap can be expected because authors and reviewers often work in the same field, an abundance of reviewer citations in an author’s paper after peer review tends to raise eyebrows. Elsevier is investigating hundreds of researchers suspected of deliberately manipulating the peer-review process to boost their own citation numbers. Elsevier analytics experts Jeroen Baas and Catriona Fennell initiated their investigation after an investigation at the Elsevier journal Geoderma concluded that an editor had manipulated citations. The editor’s most dramatic offense was suggesting 53 citations be added to 1 manuscript. pic 10-3-2019

Baas and Fennell looked at the peer-review activity of 54,821 academics who review for Elsevier journals to determine how often these researchers’ work was cited in the papers they reviewed. The academics included in the sample met the criteria of having 5 or more authored publications and having reviewed 5 or more publications of which at least 1 cites the reviewer. Of the 54,821 reviewers, 0.79% were cited in more than 50% of the citations added during review, raising questions about citation manipulation. While this percentage seems relatively low, no similar analyses are available for comparison.

Citation manipulation is insidious because authors may be hesitant to argue with their reviewer, even when they find the citation recommendations to be erroneous or irrelevant, because adding superfluous citations is a relatively painless aspect of getting their paper published. Also, in the Elsevier investigation, citations in reviewer reports often lacked author names, obscuring the fact that papers being recommended were authored by the reviewer.

An important limitation of the Elsevier study is that underlying intent of reviewers could not be measured.

Elsevier said it won’t be necessary to retract any studies affected by coercive citation because the authors aren’t responsible for the problem, and citation manipulation doesn’t affect the research, which begs the question: why go through the trouble of an investigation?

Citation manipulation is unethical. Citation metrics, such as impact factor, CiteScore, and h-index, could become meaningless if they are generated in part through citation manipulation tactics including self-citing and coercive citation. Until a concrete way to police this behavior is developed, journals will continue their prevention efforts by warning authors and reviewers against citation manipulation —just as the Ochsner Journal has done with this series of posts.