Direct answer
A plagiarism check compares your document against three bodies of text — the live and archived web, academic publications, and previously submitted student papers — and returns a similarity report showing every matching passage with its source. The percentage it produces is a measure of textual overlap, not a verdict on intent, which is why the same score can be innocent in one paper and serious in another. What matters most before you submit is not just the number, but whether the check you used stored your draft in a student paper database, because that can create problems for you later [1].
How Does a Plagiarism Check Actually Work, and What Is Being Compared?
A plagiarism check is a text-matching operation, not a judgement. When a document is submitted, the system breaks it into sentences and phrases, then runs those fragments against its comparison indexes looking for sequences that appear elsewhere in near-identical form. The output is a set of matches, each one tied to a specific source and a specific location in your text, which is why a modern similarity report shows colour-coded highlights rather than a single verdict [2].
The comparison set is the part students most often misunderstand. Turnitin's similarity check draws on three distinct collections: the current and archived internet, a publications index covering journals and periodicals, and a repository of student papers previously submitted to the system. Each of those three can produce a match, and the report distinguishes between them so you can see whether you have been flagged against a website, a journal article, or another student's essay [2].
Institutional settings change what actually gets compared. Instructors can exclude bibliographies, exclude quoted material, and exclude small matches below a word threshold, and each of those toggles alters the resulting percentage without altering your writing at all. This is the single most useful thing to understand about the process: the score is a function of both your text and the configuration of the check, so two students submitting identical work can receive different percentages [2].
What Does a Similarity Score Really Mean — and What Score Is Too High?
There is no universal "good" similarity score, and any source that gives you a fixed cut-off is oversimplifying. Turnitin's own guidance is that the percentage has to be read in context, because a score reflects the proportion of matched text rather than the seriousness of the matching. A paper that quotes heavily from primary sources in a literature review may legitimately run higher than a paper that quotes nothing [3].
The distribution of matches matters more than the total. Twenty percent concentrated in one uninterrupted paragraph is a very different signal from twenty percent spread across forty short phrases scattered through a methodology section. The first suggests a block of text was lifted; the second usually reflects standard academic phrasing, discipline-specific terminology, or a reference list that was not excluded from the check. Reading the report match by match is the only way to tell which situation you are in [3].
Institutions do set their own thresholds, and those thresholds are internal policy rather than a property of the software. Some departments treat anything above a low figure as requiring a conversation; others review the report manually regardless of the number. Because the interpretation is local, the practical move is to look at what your own department's handbook says, then inspect your matches against that standard rather than against a number you found online [3].
Does Running a Plagiarism Check Before Submission Put My Paper in a Database?
This is the question that decides whether a pre-submission check is safe, and the answer depends entirely on how the check is configured. In an institutional Turnitin setup, submissions are tied to a class and an assignment that an instructor created, and the assignment settings determine whether the paper is retained in the student paper repository. Papers are not added automatically by the act of checking; they are added when the assignment is set to store them [4].
That detail matters because a retained draft can come back to you later. If your own pre-submission draft enters the repository, a subsequent official submission of the same or a revised version can match against your earlier upload, producing a self-match that looks like collusion or recycling even though you wrote every word yourself. Students who check informally through an institutional account sometimes create exactly this problem without realising it [4].
A non-repository check avoids the issue by design: the document is compared against the full index, the report is generated, and the file is not added to the student paper database or shared onward. That is the configuration to look for if your goal is simply to see your similarity report and AI writing report before you submit. It gives you the same comparison against the same indexes without the side effect of permanently storing your draft [4].
Reading a similarity report is straightforward once you have one in front of you — the difficulty is getting an accurate preview without your draft being retained. That is precisely what turnitin0 was built for: a non-repository check that returns the same Turnitin AI detection report and similarity report your professor sees, with every match highlighted and every source listed, typically in under fifteen minutes and without adding your file to the student paper database.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is a plagiarism check the same thing as a plagiarism verdict?
No. A check produces a similarity report listing matched sources; a verdict is a human judgement made by an instructor who reads that report alongside your citations. Turnitin states directly that a high similarity score is not proof of plagiarism [1].
What percentage of similarity is acceptable at university?
There is no universal figure — institutions set their own thresholds and interpret reports manually. Turnitin's guidance is to evaluate context and match distribution rather than chasing a target number [3].
Will checking my own draft flag me for self-plagiarism later?
Only if the draft is retained in the student paper repository. A non-repository check compares your text without storing it, so a later official submission cannot match against your own preview [4].
Why does my similarity score change between two checks of the same document?
Because the comparison indexes update continuously and because settings such as bibliography exclusion and small-match thresholds alter the calculation. The same text checked under different configurations will not produce the same percentage [2].
Do I need a plagiarism check if I wrote everything myself?
It is still useful, because correct citation practice can still produce matches against the sources you cited, and those matches are what a marker sees first. Reviewing them beforehand lets you confirm your quoting and referencing are formatted the way your department expects [1].