Plagiarism and Similarity Checks in Research Papers

A similarity report comes back with highlighted passages, and you wonder whether your manuscript has a problem with plagiarism in scientific publishing. The answer depends on what was matched and whether it was cited or reworded, not on the percentage alone.

• Quick answer on plagiarism: presenting someone else’s work, wording or ideas as your own is plagiarism, and citing the source does not excuse copied wording.
• Quick answer on the score: journals set their own limits, so read the matched passages and the target journal’s policy before reacting to a number.
• Quick answer on your own text: reusing your earlier writing without citing it can be self-plagiarism, so cite the original and rewrite what you can.

This guide takes those problems in turn. It also fits into the wider set of guidance for bioscience journals.

Defining plagiarism in scientific papers

Plagiarism is presenting someone else’s work as your own, according to a university science writing guide. A broader definition from the Life and Science journal describes it as taking and using another person’s thoughts, writing or inventions as one’s own. That reaches beyond sentences to ideas. The same journal says ideas presented with superficial modifications and no credit to the original source also count.

In scientific writing, the university guide points out, what gets plagiarized is how the information is presented. Facts such as a gene’s function are not the issue. The issue is the wording and structure someone else used to present them. The guide also says that scientific writing generally does not allow direct quotes from reference material, so everything must be paraphrased. That raises the standard for how closely your sentences may follow the source.

Three points shape how you should treat your own drafts:

• Citation does not cure copying. The university guide says a passage that stays too close to the original is still considered plagiarism even if the sources have been cited.
• One sentence is enough. Copying even a single sentence word-for-word counts, even if the surrounding sentences are different.
• Intent is not a defence. The same guide states that the penalties still apply when the plagiarism was inadvertent.

Reading a similarity report

A similarity report compares your manuscript with a database of published and online text and highlights the matches. Journals run this check at submission. Life and Science employs Turnitin to evaluate the similarity index of all submitted manuscripts. Academia Scholarly Journals uses iThenticate to check each manuscript. Editage reports that over 250 publishers use Crosscheck, which is powered by iThenticate.

What percentage is acceptable?

No single figure applies everywhere. Life and Science states that no more than 5% similarity should come from a single source. That is a ceiling for one matching source, not a total score to aim for, and not a rule other journals necessarily share. Academia Scholarly Journals immediately rejects manuscripts with an unacceptable level of similarity to other published works, and it leaves the judgement of “unacceptable” to the journal. Editage notes that journals automatically reject articles when Crosscheck returns a positive result for plagiarism. Find the target journal’s own wording before you submit, and treat the number as a prompt to read the matches.

What to look at in the report

• The matched source. Check whether the highlighted text comes from a paper you cite, from your own earlier work or from a source you never referenced.
• The type of passage. Decide whether the match is a copied explanation or something that needs no rewriting, such as a reference entry.
• The concentration. A long run of matched text from one source deserves more attention than scattered short matches from many.

Plagiarism, self-plagiarism and legitimate overlap

The table below sorts common situations by how the supported definitions classify them and what to do next.

The classifications come from a university science writing guide and the Life and Science plagiarism policy.

Is reusing my own methods text self-plagiarism?

It can be. Life and Science defines self-plagiarism as the verbatim or near-verbatim re-use of significant portions of one’s own copyrighted work without citing the original source. The definition turns on reuse without citation, so the safe course is to cite the earlier paper and reword the passage where you can. Text recycling can also involve a whole manuscript. Editage describes combining earlier articles into a larger article or book without acknowledging them as self-plagiarism. Read the target journal’s ethics page, because journals set their own text recycling rules.

Fragmenting one study is a related problem. Life and Science describes redundant publication as one study split into several parts and submitted to more than one journal. Overlap is legitimate where the earlier work is openly acknowledged and the new paper adds something. Overlap becomes a problem when the earlier work is hidden.

Paraphrasing biology text correctly

Correct paraphrasing means rebuilding the idea in your own words and sentence structure, then citing the source. The university guide warns that changing a few words while leaving large parts of the original sentence’s words and structure intact is still plagiarism. Editage gives the same warning about changing a few words here and there. If a run of sentences keeps the source’s structure with reworded vocabulary, the guide says you risk structural plagiarism.

An illustrative example

The sentences below are invented for illustration and are not taken from any published paper.

• Original: “Bacteria exposed to the antibiotic developed resistance through horizontal transfer of a plasmid.”
• Too close: “Bacteria that were exposed to the antibiotic acquired resistance by horizontal transfer of a plasmid.” Only a few words changed and the structure is intact, so this is still plagiarism.
• Genuinely reworded: “Plasmid exchange between cells explains how the treated population became resistant (Author, Year).” The idea is restated from your own understanding and attributed to its source.

The rewrite works because it starts from what the finding means, not from the original sentence. Read the source, put it aside, state the finding in your own terms and then compare your version against the original. If your sentence follows the original’s clause order, rebuild it again.

Reducing similarity without gaming the check

Legitimate reduction means improving the writing and the attribution, not disguising matches. Swapping a few words to lower a score does not remove the problem, because that practice is still considered plagiarism under the guidance above. These changes are legitimate:

1. Rewrite flagged prose from understanding. Use the paraphrasing approach above for explanatory passages, and cite the source.
2. Cite your own earlier work. If a methods passage or other text comes from your published paper, say so and reword it where you can.
3. Keep citations relevant. Academia Scholarly Journals states that including citations not relevant to the work is strongly discouraged, so do not pad the reference list to look scholarly.
4. Write the study up once. Report one study as a coherent whole, not in fragments.

While revising, give the title the same scrutiny as the text. The guide on how to write a title for a scientific paper covers that step. Venue choice also shapes which policies you work under, and open access publishing for biology researchers covers that side of the decision.

What happens when plagiarism is found

Consequences depend on whether the problem is caught before or after publication. Before publication, the outcome is usually rejection. Academia Scholarly Journals immediately rejects manuscripts containing unacceptable similarity, and it says it follows the guideline of the Committee on Publication Ethics (COPE) in suspected cases of plagiarism.

After publication, the stakes are higher. Life and Science states that if plagiarism is proven the article will be retracted, and the case would be referred immediately to the author’s institution or another regulatory body for investigation. Editage adds that those found plagiarizing could lose tenure or funding and, more importantly, credibility in the academic community. A retraction is the documented consequence for a proven case, so the safer course is to raise a concern with the journal early.

Frequently asked questions