Self-citation: why (or why not), how, and when?
Citations form an integral part of scientific research. By referring to another source, the author builds on earlier knowledge thereby furthering the development of one’s field. Citing the sources used in a paper is essential as it allows one to provide an opportunity for interested readers to check and gain a more in-depth understanding of the original works, as well as to give credit to the original authors of the cited works thereby staying away from plagiarism.
A good indicator of the relevance and quality of scientific results is their impact on follow-up research. Obviously, this can be measured by the frequency they get cited. This naturally led to using citation numbers as a main tool to evaluate scientific papers and even researchers themselves. It should not be surprising that the – often exclusive – application of these numbers may motivate their manipulation. However, any citation not justified by honest scientific reasons is a seriously unethical misconduct.
When an author refers to one’s own ideas or data featured in one’s earlier publication, it is a self-citation. They are not very different from any other citation, but they are the easiest to manipulate. While they can be regarded as an acceptable form of citation whenever they are scientifically well-founded, self-citation practices can pose several ethical questions and undermine an individual’s academic reputation. In this blog post, we do not only look at the pitfalls associated with self-citation, but also give tips on how to avoid them.
There can be various valid reasons behind self-citation. The most obvious case is author self-citation, that is, when the citing paper contains expansion of an author’s own, formerly published research. This practice is inevitable in the case of a larger research project or when a researcher deals with a research problem similar to another one studied before. Some authors may falsely believe that it is OK to use their own ideas or texts in multiple publications without self-citing; however, this counts as self-plagiarism, and thus, should be avoided.
There are instances of legitimate self-citation that can occur once an author has submitted one’s manuscript to a journal. An editor or a reviewer often points out that some relevant publications are missing from the cited sources. Since their work is typically closely connected to the submitted article, it may happen that the given parties find that some of their own papers are missing from the references. In the same way, they might think that some papers from the journal in question need to be referenced as well. These cases are known as editor, reviewer, and journal self-citation, respectively. It is important to stress that, in order to be considered legitimate, these additions need to be well-founded, and suggested to, rather than forced upon the author.
Self-citation is a relatively common practice among researchers. A study by Ioannidis et al. (2019, p. 2) found that, based on a cohort of 100,000 top researchers from across many fields, “the median percentage of self-citations is 12.7%, but it varies a lot across scientists.” Furthermore, they identified more than 1,000 scientists with a >40% self-citation rate. Honest explanations of such a high ratio can be that the author is indeed the most influential researcher of the subject or perhaps his research is somewhat isolated from others’. However, it also may indicate excessive self-citation practices, which makes it highly likely that the given authors are associated with so-called citation farms, namely, small groups of researchers that cite each other’s works extensively (Ioannidis et al., 2019).
You can self-cite both published and unpublished works of yours. The formatting rules for the self-citation can vary according to the referencing guide used. It is highly likely though that they are the same as in the case of regular citations of published and unpublished sources.
Is the practice ethical?
While, normally, self-citation is an acceptable practice, an important question to ask is when it becomes non-ethical. Illegitimate self-citation or citation manipulation can occur when the act of referencing is driven solely by self-promotion. This can refer to a desire to artificially increase a researcher’s or a journal’s citation count, h-index, or, in the case of a journal, its impact factor as well. Engaging in such a practice means the violation of publication ethics: the given party can risk one’s removal from a journal and the journal can lose its existing membership or chance of having a membership in a citation index.
To avoid being accused of self-citation buildup, we recommend checking your citation records occasionally. In the citation reports of indexing databases, such as the Web of Science and Scopus, information is provided as to the total number of an author’s citations and the number of independent citations (i.e., excluding self-citations). In Google Scholar, you can only track these citation data at the article level.
Affecting the academic reputation
Self-citation can be a double-edged sword when it comes to a person’s academic reputation. On one hand, it can raise a researcher’s citation metrics, such as the h-index and i-10 index, which will potentially lead to more chances of winning grants, being hired, or promoted. On the other hand, excessive self-citation can also undermine a researcher's reputation in the sphere of higher education as the academic community is becoming increasingly aware of this practice done for the sake of self-promotion.
In fact, there have been multiple measures introduced in recent years to detect excessive self-citation. Flatt (2017) created a new citation metric to complement Hirsch H-index (analogously computed from all citations): s-index, in other words, self-citation index, which can be defined in the following way: “A scientist has an index of s if he or she has published s articles each of which has received at least s self-citations.” Also, the Swiss Academies of Arts and Sciences (2021, p. 25) updated their Code of conduct for scientific integrity to expand their list of academic misconduct, which includes “unjustified and/or selective citation or self-citation” (note the occurrence of self-citation on equal footing).
Self-citation: How to do it right?
Perhaps the most obvious question to arise in relation to legitimate self-citation practices is: how much is too much? Well, as there is no rule on the total number of citations in a full-size paper (depending on the topic, it may vary from a handful to a few hundreds), there is no strict limit as to the ratio of self-citations. The only rule to keep in mind is that all citations – thus, self-citations, too – must be founded by honest scientific reasons.
Similarly to the above mentioned study by Ioannidis et al. (2019), Szomszor et al. (2020) found the mean value of self-citation rate to be close to 10%, based on a sample of 3,517 highly cited researchers. Nevertheless, care should be exercised before making neglectful generalizations. This is because there may be field-specific expectations as to what counts as a legitimate self-citation rate when narrowing the focus of inquiry to just one research domain. For example, in Mathematics, the average self-citation rate can be >30% (Szomszor et al., 2020)!
While self-citation levels can be field-specific, there are some general questions to bear in mind as to how a researcher should practice self-citation.
- Does the cited piece of writing have its own originality? In order to give fair credits, the very first occurrence of an idea or result should always be quoted.
- Do you have the right to reproduce the original work? / Who is its copyright holder? Even if you feel that repeating some information from a published work would make your paper more self-contained and easier to read, the copyright rules may prevent you from doing so. Copyright-protected material is only allowed to be quoted but not repeated.
- What are the expectations of the academic community, publishers, and readers? Consider what details about the history and evolution of the discussed research are welcome and choose the number of references accordingly.
- What are the conventions of the genre you are writing in (e.g., journal article, blog, thesis)? Obviously, more detailed reference lists are needed in review articles and theses, while more concise ones in brief reports or letters. Consult the given journal’s Guide for authors for instructions.
- What are the specifications of the style guide that you are using (e.g., APA, MLA)? All citations must be presented in the required style.
AKJournals to help with citation dilemmas
AKJournals stands up against any unethical practice in scientific publishing, including excessive self-citations, self-plagiarism, and involvement in citation farms. We adhere to the principles stated in the COPE Code of conduct for journal publishers, and require our editors to follow the COPE Code of conduct for journal editors, and our peer reviewers the COPE Ethical guidelines for peer reviewers. Our company also has its own Ethical Guide for authors, which all the authors of our journals are expected to comply with.
When in doubt concerning citation practices, we recommend consulting the earlier mentioned documents. Checking our ever-growing collection of scientific articles on the topic at hand might also benefit our interested readers/authors and provide them with additional insights.
Flatt, J. (2017). The case for tracking self-citations. Physics Today. https://doi.org/10.1063/PT.6.3.20170919a
Ioannidis, J. P. A., Baas, J., Klavans, R., & Boyack, K. W. (2019). A standardized citation metrics author database annotated for scientific field. PLoS Biology, 17(8), e3000384. https://doi.org/10.1371/journal.pbio.3000384
Szomszor, M., Pendlebury, D. A., & Adams, J. (2020). How much is too much? The difference between research influence and self-citation excess. Scientometrics, 123, 1119–1147. https://doi.org/10.1007/s11192-020-03417-5
Swiss Academies of Arts and Sciences. (2021). Code of conduct for scientific integrity. http://doi.org/10.5281/zenodo.4707560