A publishing service provided by
The University Library at
The similarity test – things to consider
“Matching text, by itself, is not an indication of plagiarism. To determine if matching text is plagiarism a human needs to evaluate that the text was copied inappropriately and not cited correctly.” – Plagiarism.org
Each case should be considered separately
Is the overall score high? (≥10%)
- Did you exclude the bibliography, references and quotes?
- Could there be formatting issues making iThenticate include these?
- Are there other sources you can exclude? (e.g. several small matches of frequently used terms or phrases, matches of fewer than e.g. 20 words, or technical language that is difficult to rephrase?)
- Are the individual scores consisting of one or a few larger matches (≥5%) and several smaller ones?
- Are the individual scores all small matches (<5%)?
- Depending on the length of the text, the smaller matches could also be of interest
- Are there huge chunks of text from the same source (several sentences per paragraph), or are the paragraphs made up of matches from different sources?
- Would it be possible to rewrite, or cite matches as direct quotations?
Which manuscript section are the matches in?
- The methods, and to some degree introduction, sections can contain a higher degree of matches than other sections. Excluding frequently used terms and phrases, there should be no matches in the results and conclusions sections, and none to very few in the discussion section.
Are the matches instances of text-duplication (“self-plagiarism”)?
- For the methods section and introduction, is it possible to paraphrase the text? Alternatively, can a line referencing where the background/methods have previously been described substitute the duplicated text
- Even though it may be the same author(s), it could still be a case of copyright infringement if large amounts of text are duplicated from a previous publication, as the journal usually holds the copyright to the text
- For Open Access journals under Creative Commons licenses, reuse of parts of the introduction and/or methods from these articles could be acceptable, but it should still always include a citation to the original publication
- Re-use of data without clear scientific justification and transparency should be dealt with according to COPE's guidelines for a suspected redundant (duplicate) submission
For more, see: