Direct answer
To see whether your paper is plagiarized, run it through a similarity checker that compares your text against student papers, publications, and web sources, then read the resulting report line by line instead of trusting the headline percentage. Turnitin's Similarity Report highlights every matched passage and numbers each source so you can judge whether the match is a citation, a common phrase, or genuinely unoriginal text [1]. The reliable sequence is: check the draft before submission, open the report, inspect each match in context, and fix only the passages that are actually unoriginal.
What is the most reliable way to check whether my paper is plagiarized before I submit it?
The most reliable check is one that uses the same comparison database your institution uses, because a checker that only scans the open web will miss matches against previously submitted student papers — the single most common source of accidental overlap. Turnitin compares a submission against its repository of student papers, publications, and internet content, which is why its results track what an instructor will actually see [1]. Free tools that crawl only public web pages give you a false sense of safety: your paragraph can match an essay submitted to another university two years ago and still return a clean result.
Access is the practical constraint. Students usually cannot generate an official institutional report on demand, because preview and draft-check options depend on whether the instructor has enabled them in the assignment settings [2]. Where self-check is enabled, the report you get is the same report type the instructor sees, and it contains two separate signals: a similarity score for matched text and a separate AI-writing indicator [2]. Treat them as two different questions — "did I borrow text?" and "does this read as machine-generated?" — because one can be high while the other is zero.
The order of operations matters more than the tool. Check before submission, not after, and check the final version rather than an early outline, since late edits, added quotations, and restructured paragraphs all change the match set. If you use a draft-check feature that files your paper into a student repository, be aware that this can create later matches against your own work in a subsequent submission [2]. A non-repository check avoids that side effect entirely.
Finally, keep a short record: the date you checked, the score, and the specific passages you revised. If a department later questions the draft, that record shows a good-faith verification process rather than a last-minute scramble.
What does a similarity score actually mean, and which matches are real plagiarism?
A similarity score is a proportion, not a verdict. It tells you what percentage of your document's text matched something in the database; it does not tell you whether that matching is legitimate quotation, a properly cited paraphrase, a reference-list entry, or copied prose [1]. A paper at 22% built entirely from correctly attributed sources can be perfectly fine, while a paper at 9% containing one unattributed copied paragraph is a real integrity problem. Read the report, not the number.
Turnitin's report distinguishes signals that students often conflate. The similarity score covers matched text, while the AI Writing Report is a separate estimate of how much of the document appears machine-generated [3]. When Turnitin's AI detection confidence falls below its threshold of 20%, the report shows an asterisk (*%) instead of an exact figure — that asterisk means a low-confidence signal, not a confirmed pass and not a confirmed failure [3]. Reading *% as "0% AI" is one of the most common misinterpretations.
Filters change what the percentage counts. Quoted material, bibliography entries, and small matches can be excluded from the similarity total, which is why the same paper can show different numbers depending on which filters are active [3]. Before you panic about a figure, check whether quotes and references are included; a reference-heavy literature review can look alarming until the bibliography is excluded.
Real plagiarism, in report terms, is a match that is substantial, unattributed, and not a common expression. Inspect each numbered source, open the highlighted passage, and ask three questions: is this quoted and cited, is it a phrase anyone would write, or is it someone else's sentence wearing my formatting? Only the third category needs rewriting.
Can I see the exact Turnitin similarity and AI report my professor will see, before submitting?
You can see the same type of report, but exact equivalence depends on settings. Institutional reports are generated inside the LMS assignment, and what your professor opens includes the similarity percentage, the numbered source list, and the AI-writing indicator attached to that same submission [4]. Because the report is tied to assignment configuration — repository inclusion, exclusion filters, and match-size thresholds — a preview run under different settings can legitimately produce slightly different numbers than the final institutional run [4].
That settings dependency is worth understanding rather than fearing. If your preview excludes quotes and bibliography and your department's assignment does not, expect your institutional figure to be higher, and plan for it. The reverse is also true. What stays constant across both runs is the substance: the passages that match will match, and the AI indicator will respond to the same textual features.
Turnitin itself frames the report as a discussion tool rather than a disciplinary instrument, noting that matches frequently reflect citation practice and that interpretation belongs in an open conversation between student and instructor [4]. That framing is useful when you review your own draft: your goal is not to force the percentage to zero — an impossible and often damaging target, since over-paraphrasing strips out the citations that make academic writing credible — but to make sure every match is one you can defend.
The practical takeaway: preview with the same file type and the same filters your assignment uses, keep the report, and fix only defensible problems. A clean, well-cited 18% is a stronger submission than a mangled 2%.
If you want certainty rather than an estimate, turnitin0 lets you run your draft through a real Turnitin check and download the AI detection report and the similarity report together, so you can inspect the same match highlights and score bands your professor will open — before the deadline, not after.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is a high similarity score the same as plagiarism?
No. The similarity score measures matched text, and matches include correctly cited quotations, bibliography entries, and common phrasing [1]. Plagiarism is a judgement about attribution, which is why the report has to be read passage by passage.
Why does my report show *% instead of a number?
Turnitin displays an asterisk when its AI detection confidence is below its 20% threshold, meaning the signal is low-confidence rather than an exact percentage [3]. Treat it as inconclusive and rely on the similarity matches and your own citation review.
Can checking my paper get me flagged later?
It can, if the check adds your file to a student paper repository, because a later submission may then match your own earlier draft [2]. Use a non-repository check if you want to avoid that.
Should I aim for a 0% similarity score?
No. Chasing zero usually means removing citations and paraphrasing sources into unrecognisable text, which weakens the paper. Aim for every remaining match being one you can defend as quoted or cited [4].
Does my professor see the same report I preview?
The report type is the same, but the numbers can differ slightly because institutional reports use the assignment's own repository and filter settings [4]. The matched passages themselves will be consistent.