Addressing a Report Indicating High Similarity

Amy Clark, Associate Editor

September 2020

You receive notice from a journal that your submitted manuscript has been run through a plagiarism detection tool and the reported similarity score is too high for the journal to proceed with review of the manuscript. What should you do? First, don’t panic. A high similarity score does not necessarily mean that your manuscript has plagiarism issues. Let’s go over some basic information about what a similarity score is and how to interpret it.

What is a similarity score?

A similarity score is an indicator of the percentage of text in a manuscript that is flagged as overlapping with published material. Many journals screen submitted manuscripts with the plagiarism detection software iThenticate, which uses algorithms to crawl indexed content and generate a report that highlights text that matches with text from published sources.

The similarity report generates a percentage index of the amount of text matched to each source as well as an overall score that combines these percentages. For example, an overall score could be 25% but the highest single-source score could be only 2%. A report like this isn’t likely to ring alarm bells. However, an overall score of 25% with a highest single-source score of 20% more likely indicates a case of plagiarism or text-recycling (duplication of one’s own published material).

The human factor is key in the interpretation of similarity reports. Editors will assess the context of matched content. Text matches in the methods section of a manuscript may not be as concerning to editors as matches found in the results section, for example. Editors must analyze the reports and use their expertise and judgment to determine if there is cause for concern.

After reviewing the report with a high similarity score, what should you do next?

There are several ways to take care of duplication problems in a manuscript that has received a high similarity score.

Cite the work of others correctly.

If you determine that referencing previously published material and quoting the work of others are issues in your paper, you can address these problems by adding citations where needed. Consult the following Learning Nexus articles for help with citing sources and using an appropriate reference style:

Rewrite flagged text.

Determine which part of the paper is most problematic. Is it your introduction, methods, or discussion section? Often you will be able to rewrite text using synonyms and alternate phrasing to avoid duplications. Adding new text to the paper such as in the introduction and conclusion can also help lower the similarity score for the entire manuscript and enable you to reach the targeted percentages so that your manuscript can advance to the review stage.

Write to the journal editor. If field-specific terminology is flagged in the similarity score report, you will have less room for variation. The methods section of your manuscript will often include more instances of exact, field-specific language compared with other sections of your manuscript. If this is the case, you can write to the editor of the journal and provide examples of language that cannot be changed. After all, there are only so many ways to explain techniques and you certainly cannot change the names of proprietary products.

While many authors are unsure about contacting the editors of journals that reject a manuscript, many editors are very willing to hear what the authors have to say, and they may accommodate an author’s request. Remember that editors are also authors and they will likely sympathize with you as they have gone through the process themselves. The key is to be sure that your response to the editor is professionally written and does not have an accusatory tone.

For more tips about communicating with a journal after a manuscript is rejected, read our Learning Nexus article Rejection Is Not (Always) Final.

Write to the journal editor.

Keep in mind that a similarity score is generated by technology, and as with any software tool, it has advantages and limitations. Journal editors use the similarity score in the manuscript screening process, but it is only one piece of their assessment.

Previous Article Next Article