I've finished writing a paper with 2 co-authors. As everyone participated in the writing, I thought there might be some ( maybe unexpected) plagiarism in the paper. So I checked the paper with Cross check software, for some reasons I couldn't supply all of the paper to the cross check program at the same time. Therefore, I divided the paper in 4 parts in which, each part has almost 2,000 words.
The plagiarism level found by the program for part 1,2,3 and 4 are 12%,20%,1%,16%. So my questions are:
1) what is the overall plagiarism level of the paper? In other words, should get the average of these 4 parts or get their summation them? (assuming that each parts has 2000 words.)
2) lets assume the last part only contains the references (16% plagiarism level detected). Do journals take the plagiarism level of the references into account as well?
3) what are the thresholds of the most popular journals? (IEEE,Elsevier,John wily, Taylor & Francis)

  • 6
    There is nothing like a "plagiarism level". Either a paper contains copied material, or it doesn't. Citations and references don't count. Software can help you and flag suspicious sentences, but in the end that is something that a human being must evaluate. – Federico Poloni May 12 '15 at 21:01
  • Good to know. Thanks for the info @FedericoPoloni – SAH May 13 '15 at 5:42
  • For 1) you may want to ask on math.stackexchange.com :) – silvado May 13 '15 at 9:16
  • @silvado Whats your idea ? ;) – SAH May 13 '15 at 16:52

Overall levels of plagiarism are completely irrelevant. What matters is whether there is at least one clear-cut case in which an extended phrase has been plagiarized from one of the sources. If so, the paper is unacceptable. In principle, one could have 1% similarity and clear plagiarism, or 20% similarity and no plagiarism.

Once you look it in this light, it doesn't matter what the program does with respect to the references. Properly formatted references will look the same in any paper and so duplicating the text of any single reference is not considered plagiarism. (You can plagiarize references, however -- if I were to copy a set of important citations from another paper and then repeat them in order as if choosing them were my own work, this would be plagiarism.)

Of course, plagiarism checking software only catches the most obvious instances: direct plagiarism of blocks of text. One need not copy text exactly to have plagiarized.

Finally, if any of the major publishers have "thresholds", they are being foolish. As the software producers plainly admit, we are still at the point where a human needs to look at the instances reported by the program and make a judgement call about whether one is dealing with an actual instance of plagiarism.

Edit: Given the comment below, it appears that some publishers are running submissions through a plagiarism detector upon submission. I suppose it might make sense to have a high cutoff above which everything gets rejected as probably plagiarized and not worth the editors' time -- but if they do that, "beating the plagiarism checker" is not what you need to be doing as a researcher. The last thing you want to do is submit a paper that slips in under the threshold yet contains instances of plagiarism that will come back to haunt you later.

| improve this answer | |
  • Thanks for your reply @Corvus But I know someone's paper has been rejected from Elsivier at the initial steps after submition . Elsevier told them , our cross check software detected 16% of plagiarism in your paper, So you need to edit it. – SAH May 12 '15 at 19:52
  • 2
    @Electricman In recent submissions to IEEE conferences, I had similarity scores ranging from 11 to 24 % - without any plagiarism, and without any questions or comments from the editors to this. I wouldn't revise a paper just to get below any (arbitrary) threshold in this score. Better choose a journal where editors actually check the content. – silvado May 13 '15 at 9:14
  • IEEE is a good boy @silvado . I didn't have problem with IEEE,Transaction journal as well. I had 12% plagiarism level as I checked it with cross-check software and my paper got accepted without any trouble. – SAH May 13 '15 at 17:42
  • 1
    Again, a plea for people to ensure that they are not plagiarizing in the first place, rather than worrying about whether they can slip in under some journal's filter. – Corvus May 13 '15 at 17:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.