None, and that is not a technicality, it is Turnitin’s own position: the similarity report does not check for plagiarism, it identifies matching text and leaves the judgment to your instructor. There is no universal threshold where a percentage becomes an offense, no number that is automatically safe, and no number that is automatically guilty. What the score does do is trigger human decisions, and if your score has triggered one against you, the case is not about the percentage. It is about what the matches actually are, which is a question the headline number cannot answer and your defense must.
What Does the Similarity Score Actually Measure?
The score is the percentage of words in your submission that match text in Turnitin’s databases, which can include internet pages, publications, and previously submitted student papers, depending on how the assignment was configured. Matching is mechanical and indiscriminate by default: properly quoted material matches its source because it is supposed to, your bibliography matches because citations are identical everywhere, common technical phrases match because the discipline writes that way, and assignment templates match for every student who used them.
This is why the raw number routinely overstates the problem. Instructors can apply filters excluding quotes, bibliographies, and small matches, and applying them commonly drops scores substantially with no change to the writing, but the unfiltered number is often what appears first, and what panics students and, sometimes, what convinces an instructor to file. Whether your score was read filtered or raw, and whether anyone examined which text matched what, is frequently the difference between a number and a case.
Why Do Honest Papers Get High Scores?
For reasons built into how research writing works. A heavily sourced literature review matches its sources. A qualitative paper full of required quotations matches everything it quotes. A lab report following the course’s mandated structure matches every other correct lab report. A resubmitted draft can match your own earlier submission if repository settings stored it, and a paper on a narrow topic matches the small universe of things written about that topic. In each case, the match is real and the plagiarism is absent, which is the entire distinction the score cannot draw.
The distinction is drawn inside the report, at the match level: which sources, which passages, cited or not, quoted or paraphrased, body text or references. A defensible paper with a scary number usually has its innocence written plainly in that detail. Whether the detail gets read that way depends on who walks the reviewer through it, which is what a defense is.
What Should You Do If a Similarity Score Becomes an Accusation?
First, get the full report, not the summary number, because the source-by-source detail is the actual evidence and it is often more favorable than the percentage implies. Second, resist the instinct to revise the paper or resubmit a cleaned version, since post-accusation edits complicate the record and can read as concessions. Third, understand which matches the accusation actually rests on, because answering the wrong ones wastes your response on ground nobody contested. And underneath all three: how the matched passages are characterized, as citation practice, as permitted overlap, as patchwriting, or as copying, is where these cases are decided, and characterization is advocacy. Students describing their own matches carelessly have converted citation sloppiness into misconduct findings that the evidence never required.
What Mistakes Turn Scores Into Findings?
Arguing the number instead of the matches. Saying “25 percent is normal” concedes the frame. The case lives in what the 25 percent is.
Fixing citations after the flag. Corrections made once the process starts do not undo the submission and can be portrayed as acknowledging the violation. What to touch and when is a strategy question, not a housekeeping one.
Explaining patchwriting as paraphrasing. Close paraphrase that tracks a source’s structure is the hardest category, and describing it inaccurately, in either direction, damages credibility precisely where the case is closest.
Assuming the instructor read the report deeply. Accusations get filed on headline numbers with the match detail unexamined. If nobody has done the source-level reading, your response is where it happens, or it never does.
How Does an Attorney Handle a Similarity Score Case?
Richard Asselta is a student defense attorney who defends students in plagiarism and academic misconduct cases nationwide. In similarity cases, he takes the report down to the match level, separates required and cited overlap from anything genuinely contested, frames the characterization the panel will actually vote on, and holds the school to proving misconduct rather than pointing at a percentage.
Because these cases turn on each school’s own policies and burden of proof, he defends students at colleges and universities across the country.
Frequently Asked Questions
Is a 30 percent Turnitin score plagiarism?
No percentage is plagiarism by itself. Thirty percent of properly quoted and cited material is a research paper; ten percent copied without attribution is a violation. The matches decide, not the number.
My bibliography and quotes are driving my score up. Is that held against me?
They match by definition, and filters exist to exclude them, but whether your reviewer applied those filters is not guaranteed. Part of defending a similarity case is making sure the score being judged is the meaningful one.
Turnitin matched my paper to my own earlier draft. Am I in trouble?
Self-matches from stored submissions are a known artifact, and same-class resubmissions are normally excluded by default. When one surfaces anyway, identifying it as your own prior work, correctly and on the record, usually resolves that match, though how it is raised matters.
My instructor says the score is too high but will not say which parts are the problem.
You are entitled to understand what you are accused of before responding, and a vague accusation is itself worth addressing through the process. Responding blind to an unspecified number is how students concede things nobody had established.
The Number Started This. It Does Not Get to Finish It.
A similarity score is a search result, not a verdict, and Turnitin itself says the judgment belongs to humans. If your score has become an accusation, the match-level truth of your paper is the case, and building it is the work in front of you. Attorney Richard Asselta defends students in similarity and plagiarism cases at colleges nationwide. Call 855-338-5299 before you respond to the report.