You really should not depend on the overall similarity index. A similarity index in excess of 40 percent is almost always problematic. However, to be of the safe side, try to reword the text in the similar parts as far as possible. A similarity index between 20-40 percent generally means there is a problem unless a large portion of text that should have been skipped was not (e.g., block quotes, reference lists, or appendices of common tables). Having said that, I personally feel that 23 percent is not very high in terms of content similarity. You should be very careful to avoid similarity in the results section. If the results section has similar content, you might be in trouble. If the introduction or discussion has more of similar content, you should try to reword it as far as possible. If most of the similar content is in the methods section, you have nothing to worry about. If you check the detailed report provided by Turnitin, you will find that the parts where similar content has been detected have been marked. Thus, more than the overall percentage of similarity, what actually matters is which section of your manuscript has the maximum similarity. This is particularly true of the Methods section of the manuscript, and generally editors have a more lenient attitude towards content similarity in the methods section.

The acceptable limit for similarity of text from a single source is 5.

The similarity index in such cases is higher than the approved limit for a. The Content Tracking mode ensures that manuscripts with an overall low percentage similarity (but which may have a higher similarity from a single source) are not overlooked. The acceptable limit for similarity of text from a single source is 5. Journal editors are also aware of the fact that in some cases rewording content would be very difficult, and they tend to make some allowances in such cases. The text of every submitted manuscript is checked using the Content Tracking mode in iThenticate. In most cases, editors use their discretion in consultation with the peer reviewers to judge whether a paper should be rejected on account of plagiarism. They understand the limitations of these tools and in most cases do not reject a manuscript outright based only on the results of the plagiarism detection software unless the percentage of plagiarized content is very high, at least above 30 per cent. Journal editors do not rely completely on plagiarism detection software.

Ithenticate acceptable similarity index