Direct answer
Plagiarism detection software compares a submitted document against large indexed databases of web pages, academic publications, and previously submitted student papers, then reports the percentage of text that matches those sources. The most widely used system in universities across the United States, the United Kingdom, Canada, Australia, New Zealand, and Ireland is Turnitin, and its output is a similarity report rather than a verdict of guilt [1]. Understanding how that report is built — the similarity index, the match overview, and the inline highlights — is what lets a student decide whether a draft is ready to submit [1]. This guide explains the mechanics, what a normal similarity percentage looks like, and how to preview your own paper before the deadline.
Introduction
Plagiarism detection software compares a submitted document against large indexed databases of web pages, academic publications, and previously submitted student papers, then reports the percentage of text that matches those sources. The most widely used system in universities across the United States, the United Kingdom, Canada, Australia, New Zealand, and Ireland is Turnitin, and its output is a similarity report rather than a verdict of guilt [1]. Understanding how that report is built — the similarity index, the match overview, and the inline highlights — is what lets a student decide whether a draft is ready to submit [1]. This guide explains the mechanics, what a normal similarity percentage looks like, and how to preview your own paper before the deadline.
How Does Plagiarism Detection Software Actually Work?
Plagiarism detection software does not "understand" plagiarism. It converts a document into a digital fingerprint and then searches for overlapping strings of text in its indexed corpora — web pages, journal and book publications, and a repository of student papers previously submitted to the same institution [2]. When the software finds a match, it highlights the passage and attributes it to a specific source, so the reader can judge whether the overlap is a legitimate quotation, a common phrase, or unattributed copying [2].
The report you see is layered. A top-level similarity index gives the overall percentage of matched text, and beneath it a match overview lists every source with its own contribution to that percentage [2]. Instructors can filter the view by source type — internet, publications, or student papers — and can exclude small matches or bibliography sections, which changes the number displayed without changing the underlying text [2].
This is why two people can look at the same paper and see different percentages. The index is a function of the comparison set and the exclusion settings, not an absolute property of the writing [2]. A paper checked against a small institutional repository will score differently from the same paper checked against the full web and publication index [2].
What Is a Good Similarity Percentage in a Turnitin Report?
There is no universal "good" similarity percentage, and Turnitin itself does not publish pass/fail thresholds — the company states that the score must be interpreted in context and that a high percentage can result from correctly quoted and cited material [3]. A literature review heavy with direct quotations may legitimately sit in the twenties or thirties, while a short reflective essay with no quotations should normally sit much lower [3].
What matters more than the headline number is the pattern of matches. A single 15% match to one essay-sharing site is a very different signal from fifteen separate 1% matches to journal articles you cited properly [3]. Read the match overview source by source: if every match is attributed to a reference you included, the report is doing its job, not accusing you [3].
Because interpretation is contextual, the percentage alone should never be treated as a submission decision. Instructors read the report alongside the assignment brief, the citation style required, and their knowledge of the student's writing [3]. If a number looks unexpectedly high, the productive move is to examine the highlighted passages rather than to panic about the total [3].
Can I Check My Own Paper for Plagiarism Before Submitting It?
Most institutions do not give students direct self-submission access to the same repository their instructors use, because a draft added to the student paper database can be retained and later matched against other students' work [4]. That repository behaviour is intentional: it is what stops one cohort's essays from being recycled by the next [4]. It also means that a "free" check that adds your file to a database can create a future match problem for you [4].
What students can do is run a non-repository check — a review of the draft that produces the same style of similarity and AI writing reports without adding the file to Turnitin's student paper database [4]. This lets you see the match overview, the flagged passages, and the AI writing indicator on your own draft, then revise before the official submission [4]. Turnitin's own guidance encourages students to discuss any report they receive with their instructor rather than treating it as a final judgement [4].
The practical sequence is straightforward: finish the draft, run a non-repository check, read the match overview source by source, fix genuine unattributed overlap, and only then submit through the institutional portal [4]. That order keeps your work out of the database while still giving you the same information your instructor will see [4].
If you want to see exactly what your instructor's screen will show — the cover page, the similarity index, the flagged passages, and the source-by-source match overview — turnitin0 lets you preview a real Turnitin AI and similarity report on your own draft without adding it to the student paper database.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does plagiarism detection software prove that someone plagiarised?
No. It reports textual matches and their sources; the judgement about whether a match is plagiarism belongs to a human reader who can see the citations and the assignment brief [1].
Why did my similarity score change between two checks?
The index reflects the comparison set and the exclusion settings applied at the time, so a different repository, a newly indexed web page, or a change to bibliography exclusions can shift the number [2].
Is a 0% similarity score realistic?
It is possible but uncommon for longer academic writing, because common phrases, headings, and reference titles are themselves indexed and can register as small matches [3].
Will checking my own paper get it flagged for self-plagiarism?
Not if the check is non-repository, because the file is not retained in the student paper database and cannot be matched against later submissions [4].
Can I get the same report my instructor sees?
Yes — a Turnitin AI and similarity report preview shows the cover, score, flagged passages, and source-by-source match overview in the same format used in institutional learning management systems [1][4].