Last Summer, Royal Society Open Science started using iThenticate – the text comparison tool – on all new submissions. Here, we reflect on our experience and provide some guidance for authors submitting to RSOS.
What is iThenticate?
Royal Society Publishing is a participant of Crossref Similarity Check, an initiative between the scholarly publishing organisation Crossref, and the plagiarism detection software, iThenticate. The tool enables us to compare a newly submitted manuscript with the world’s largest comparison database of scientific publications1 with the aim of identifying any potential cases of plagiarism or self-plagiarism.
Each manuscript check in iThenticate generates a percentage overlap score. The score normally consists of several different sources, each with their own score that contributes to the total percentage overlap. iThenticate displays the manuscript with colour-coded highlighted text that corresponds to the specific overlapping sources. Each source can be investigated individually and can be viewed side-by-side with the newly submitted text.
Why and how do we use it?
We introduced routine iThenticate checking in 2018 to respond to the large growth in manuscript numbers the journal has seen since launch. We were relying heavily on Editors and reviewers to spot overlap, and this was no longer sustainable given the volume of manuscripts. In introducing these checks, the journal is not trying to trip authors up, but help them and ensure that we do our bit to support ethical publishing.
When investigating areas of overlap, we consider the following points: Where is the overlap (Introduction, Methodology, Results, Discussion)? How much overlap is there (snippets of text or chunks of verbatim overlap)? Is it the same authors? How long ago was the paper published?
Overlap in the Introduction or Methodology may be permissible to an extent, but some cases may require rewriting. Overlap in the Results and Discussion, on the other hand, is a red flag and will likely require a closer inspection of the overlapping text. In this case, we may contact the authors for an explanation before we can consider proceeding with the paper.
Instances of significant overlap may result in an immediate reject decision. When dealing with a case of possible plagiarism, we would follow COPE’s guidelines.
Plagiarism vs. self-plagiarism
Although the obvious form of plagiarism is the copying of someone else’s work, self-plagiarism is equally an issue for publication ethics. Self-plagiarism is the recycling of text from a previous publication by the same authors. This may be seen as a more benign form of plagiarism, however, it still has implications for copyright, and wider implications in the form of redundant publications diluting the publishing landscape.
Text recycling by an author, especially in the Results and Discussion, may be a sign of an author ‘salami-slicing’ their research in to more papers than is necessary or appropriate. In this scenario, we would contact the authors and ask them to clarify the difference between the submitted and published work.
Whether or not a paper makes a meaningful advancement on previous work – enough to warrant publication – may depend on several factors, such as the publisher’s policies, the journal’s criteria, the editor’s judgement, as well as on the field of research. Although authors are responsible for submitting original work, the journal should also be mindful of the publishing habits of different subject areas, where the amount of overlap with previous work may vary in acceptability.
Tips for authors:
- If you are submitting a paper that resembles a previous publication, it might be worth making sure that the previous paper is referenced and mentioned in the text, where appropriate. Where work is expanding on a similar previous publication, make sure this is clear, with emphasis on how the work being submitted is new. When in doubt, include a note in the cover letter to the Editor.
- Reusing methodology – some overlap with previously published methodology is fine; however, in cases where the overlap is extensive, it may be advisable to instead refer to the methodology in your previously published work.
- Don’t worry if your paper (or a version of it) is already available from a pre-print repository (e.g. bioRxiv) or from your institutional repository – this will be noted and not considered as self-plagiarism. Cornell University’s preprint server arXiv in fact perform their own text comparison for new submissions, but this is only a comparison with their own arXiv content.
Images: Crossref similarity check logo, COPE logo