A Turnitin percentage tells you how much text matches material in the selected databases; it does not decide whether plagiarism has occurred or show which passage should be revised first. To read a Turnitin Similarity Report correctly, inspect the source, length, context and attribution of every meaningful match. A correctly quoted sentence, a reference entry and an uncited copied paragraph may all be highlighted, but they do not carry the same academic risk. This 2026 guide gives PhD scholars, postgraduate students and research authors a practical eight-category audit for separating legitimate overlap from close paraphrasing, missing citations, patchwriting and possible self-overlap.
Quick answer: Open Match Groups and Sources, check the report settings, and review long, distinctive or repeated matches first. Classify each passage by citation, quotation, source dependence and academic risk. Correct unsupported borrowing, overly close paraphrasing and weak synthesis; verify legitimate quotations, references and required text instead of editing only to force a lower percentage.
Need a second opinion? If you already have the report and source file, request a confidential report review from ThesisLikho. Support is limited to ethical analysis, academic editing and citation correction; no score or institutional acceptance is guaranteed.
Table of Contents
- What the similarity score actually means
- The eight match types
- A five-step report analysis process
- How to prioritise corrections
- Worked examples
- Common mistakes
- Final checklist
- Frequently asked questions
What the Similarity Score Actually Means {#what-the-similarity-score-actually-means}
Turnitin calculates overall similarity by comparing the words in a submission with sources available under the assignment’s selected databases and settings. Highlighted text indicates similarity, not an automatic finding of plagiarism. Turnitin’s current report can group matches according to citation and quotation signals, while its Sources view lists sources by their contribution to the document’s similarity.
This distinction matters because two documents with the same percentage can require completely different responses:
- Document A may contain correctly marked quotations and a long reference list.
- Document B may reproduce the wording and structure of several sources without adequate attribution.
- Document C may match an earlier draft submitted to an institutional repository.
The number alone cannot distinguish these situations. The report must be read alongside the manuscript, its citations, the original sources and the university’s policy.
Important distinction: Turnitin's enhanced report officially uses four Match Groups: Not Cited or Quoted, Missing Quotations, Missing Citation, and Cited and Quoted. The eight types below are a practical editorial framework for deeper human review; they are not eight official Turnitin labels.
Percentage is a screening signal, not a verdict.
There is no universal “safe” percentage that applies to every university, course or document type. A literature review, legal thesis, technical specification and original empirical chapter naturally contain different kinds of repeated language. Institutional rules may also differ on whether references, quotations, templates and small matches are excluded.
For that reason, the correct first question is not “How do I make the score low?” It is: Which matches affect originality, attribution or compliance?
The Eight Match Types in a Turnitin Report {#the-eight-match-types}
Use this classification before changing any sentence. It prevents needless rewriting and helps direct attention to the matches that matter.
Match typeWhat you may seeTypical riskAppropriate response
1. Cited and quoted text Exact words inside quotation marks with a citation Low when accurate and permitted Verify quotation, locator and required style; shorten if overused
2. Cited but too close A citation is present, but wording or structure closely follows the source Medium to high Rewrite from understanding, change the structure and retain the citation
3. Quoted but uncited Quotation marks appear without a traceable source High Add the verified citation and reference, or remove the unsupported quotation
4. Uncited source language Distinctive wording matches a source without attribution High Reconstruct the explanation independently and cite the source
5. Common or technical wording Standard terms, instrument names, formula labels or unavoidable phrases Usually low Confirm it is conventional; do not distort technical meaning
6. Bibliography or template text References, declarations, headings or required institutional wording Usually low Check report settings and policy; do not disguise required text
7. Self-overlap Text matches the author's article, proposal, conference paper or earlier submission Context-dependent Follow reuse policy; cite or disclose prior work and revise where required
8. Patchwriting or source mosaics Short borrowed fragments are stitched together, even when citations appear High Build a new synthesis around an original analytical claim
1. Properly cited and quoted matches
An exact quotation is expected to resemble its source. Check whether the quotation marks or block-quote formatting are correct, the citation points to the right source, and the quote is necessary. A correctly attributed quotation may be academically acceptable even when it contributes to the displayed score. Excessive quotation, however, can weaken the author’s own analysis.
2. Cited text that remains too close to the source
Adding a citation does not make near-verbatim paraphrasing original. Look for the same sequence of ideas, the same clause order, distinctive phrases or merely substituted synonyms. The repair is conceptual: close the source, identify the idea you need, explain it in a structure suited to your argument, and then verify accuracy against the source. Keep the citation because the idea still came from that work.
3. Quoted text with a missing citation
Quotation marks show that wording is borrowed, but the reader must still be able to identify its source. Verify the author, year and—where the referencing style requires it—page or paragraph number. If you cannot locate the source, do not invent a citation. Remove or replace the unsupported passage after checking the evidence.
4. Uncited matching language
This category deserves immediate attention when the matched wording is substantial or distinctive. Inspect the full source rather than relying only on the coloured highlight. Determine whether both the language and underlying idea originated there. Then rewrite through independent explanation and add appropriate attribution. Changing a handful of words or rearranging clauses is not enough.
5. Common and technical expressions
Some matches arise from discipline-specific terminology, standard instrument names, statistical tests, legislation titles or conventional phrases. These should not be rewritten into inaccurate language just to change the colour of a report. The reviewer should assess whether the expression is truly standard or whether a longer surrounding passage has followed a source too closely.
6. References, declarations and templates
Reference lists naturally contain publication titles, author names and journal details that also appear elsewhere. Institution-mandated declarations and approved templates may repeat across submissions. Turnitin provides filters for bibliographies, quoted material, cited text and small matches, although availability can depend on the report version and the user’s permissions. Students should not treat a temporary personal filter view as proof that the official institutional score has changed.
7. Self-overlap and previous submissions
A thesis may match a proposal, published article, conference paper or earlier draft. Authorship does not automatically make unrestricted reuse acceptable. Check the university’s rules, publication agreement and disclosure requirements. Where reuse is allowed, cite or acknowledge the earlier work and clarify what is new. If an earlier draft has been stored in a repository, ask the authorised instructor or administrator about the correct handling; do not attempt to conceal it.
8. Patchwriting and source-by-source synthesis
Patchwriting often survives superficial synonym replacement because the source controls the sentence structure and logic. It is common in literature reviews that summarise one paper per paragraph. A stronger revision groups studies by finding, method, disagreement or limitation. The paragraph should begin with the researcher’s analytical claim, compare evidence from several sources and end by explaining its relevance to the research gap.
A Five-Step Report Analysis Process {#a-five-step-report-analysis-process}
Step 1: Record the report conditions
Before reviewing matches, note the file version, report date, total word count, databases searched and visible filter settings. A percentage without these conditions is difficult to compare with a later report. Save the original report as the baseline; do not overwrite your only copy.
Step 2: Review Match Groups, then Sources
In the newer report experience, Match Groups can separate text that appears not cited or quoted, missing quotations, missing citations, or cited and quoted. Use these as review prompts, not final academic judgments. Then open Sources to identify whether one source contributes a large connected passage or many scattered fragments.
Step 3: Inspect the passage in context
Read at least the full paragraph around each important match. Ask:
- What claim is this paragraph making?
- Is the matched language necessary terminology or author-specific expression?
- Is the source credited at the point where its idea is used?
- Does the paragraph add comparison, interpretation or application?
- Would a reader understand which ideas are mine and which came from sources?
Step 4: Check the source and reference record
Confirm that the cited source actually supports the claim. Citation correction is not simply inserting a reference after highlighted text. Check author details, year, title, page information where needed, and consistency between the in-text citation and reference list. Never cite a source you have not verified.
Step 5: Revise by priority and document the decision
Correct high-risk passages first, then repeated medium-risk patterns. Keep a simple audit log so the same issue does not return during supervisor revisions.
Location Source or matchClassificationAction takenCitation verified?
Chapter 2, paragraph 14 Journal article, long match Cited but too close Rebuilt comparison using three studies Yes
Chapter 3, instrument description Manufacturer or manual Required technical wording Retained terms; added source attribution Yes
Appendix A University form Approved template No text change; flagged for institutional review Not applicable
How to Prioritise Corrections {#how-to-prioritise-corrections}
Do not mechanically start with the largest percentage and rewrite every highlight. Use a two-factor test: academic risk and revision value.
Priority 1: Correct immediately
- Long uncited passages or distinctive language
- Exact wording with missing quotation marks
- Citations that do not support the claim
- Reused tables, figures or data without attribution/permission
- Self-overlap that may violate submission or publication rules
Priority 2: Improve carefully
- Paraphrases that retain the source’s sequence and syntax
- Literature review paragraphs dominated by one source
- Repetitive definitions that can be replaced by analysis
- Several small matches that collectively reveal patchwriting
Priority 3: Verify rather than disguise
- Reference entries
- Approved templates and declarations
- Standard statistical or technical labels
- Properly presented quotations
- Common phrases with no meaningful ownership
This order protects meaning. It also avoids a common failure: damaging a correct methodology description simply because a standard phrase is highlighted while leaving a poorly synthesised literature review untouched.
Worked Examples {#worked-examples}
Example A: Citation present, paraphrase still dependent.
Weak approach: The researcher replaces a few words but keeps the source’s sentence order and argument sequence.
Better revision method: Identify the study’s relevant result, compare it with another study, state the point of agreement or conflict, and explain why that relationship matters to the thesis. Cite both sources at the relevant claims. The revision becomes an act of synthesis rather than vocabulary substitution.
Example B: Methodology match
A paragraph naming a standard scale, software package or test may contain unavoidable terminology. Retain precise names, cite the instrument or method’s authoritative source, and independently explain why it fits this study. The technical label can remain; the study-specific rationale should be original.
Example C: Previous draft match
If the final thesis matches a stored proposal or earlier chapter, first verify the source category and institutional policy. Do not rewrite facts into awkward language merely to escape a repository match. Ask the authorised university contact whether the earlier submission can or should be excluded, while ensuring any reused published work is disclosed and cited appropriately.
Common Mistakes {#common-mistakes}
Chasing a universal target percentage
Institutional rules, disciplines and report settings vary. A target copied from another university or online forum may not apply to your work. Use the policy supplied by your supervisor, department or examination office.
Treating every highlight as plagiarism
Similarity can include quotations, references, common language and templates. A human review is necessary to decide whether attribution and originality are adequate.
Using synonym replacement or automated spinning
Word substitution can preserve the source’s logic, introduce errors and make academic writing unnatural. It does not solve weak synthesis or missing attribution.
Deleting citations to make the report look cleaner
Removing a legitimate citation can turn attributed material into unattributed borrowing. Citations exist for scholarly traceability, not cosmetic score control.
Excluding sources without an academic reason
Filters and exclusions help reviewers focus on meaningful matches; they should not be used to hide problematic overlap. Turnitin notes that exclusions refine the report but do not replace human judgment.
Uploading repeated drafts without checking repository settings
Depending on assignment settings, a previous submission may later match the final document. Use only the official submission workflow and ask the institution how drafts are stored.
Confusing this report-reading task with similarity reduction
Reading the report answers what matched, where it matched and why it may matter. Correcting similarity is a later task that may involve citation repair, independent rewriting, synthesis or an authorised exclusion. If you need that broader process, use the separate guide on how to reduce the similarity index in Turnitin ethically. Keeping these two intents separate prevents this page from duplicating ThesisLikho's existing reduction content.
Ethical Support From ThesisLikho
If the audit reveals complex citation, synthesis or self-overlap issues, ThesisLikho’s Turnitin Similarity Reduction Services and Turnitin Report Correction Services can be used for report interpretation, chapter-level editing, citation verification and original-language revision. The work should improve the manuscript—not manipulate Turnitin, fabricate sources, hide copied material or promise a particular score.
Before requesting support, prepare:
- The latest editable manuscript
- The complete similarity report, not only a screenshot of the percentage
- Required referencing style
- University or journal similarity policy, if available
- Submission deadline and chapters needing review
- Supervisor comments relevant to originality or citation
Share the report for a confidential assessment. Scope, timeline and quotation should be confirmed after the document and report are reviewed.
Final Checklist {#final-checklist}
Before resubmitting, confirm that you have:
- Reviewed the actual matches, not only the overall percentage
- Checked long, distinctive and repeated matches first
- Verified every added citation against the source
- Used quotation marks and page details where required
- Rewritten ideas from understanding rather than replacing synonyms
- Synthesised sources instead of summarising them one by one
- Preserved correct technical terminology
- Checked self-overlap and previous-submission rules
- Followed the institution’s current policy and authorised settings
- Proofread the revised passages for meaning, evidence and consistency
Frequently Asked Questions {#frequently-asked-questions}
Does a high Turnitin score always mean plagiarism?
No. The score measures matching text, while plagiarism is an academic judgment involving source use, attribution and context. References, quotations, templates and previous submissions can also create matches.
Which Turnitin matches should I check first?
Start with long uncited passages, exact wording without quotation marks, repeated matches from one source, and matched text central to your argument. Then review weaker paraphrases and self-overlap.
Should I rewrite properly quoted material?
Not automatically. Verify the quotation, citation and required page information. If quotations dominate the discussion, reduce or synthesise them to make room for your own analysis.
Can references be excluded from a Similarity Report?
Turnitin offers bibliography and other exclusion filters in supported report versions, but access and official settings depend on the institution. A student’s temporary filter view may not change the instructor’s official score.
Why does my final thesis match an earlier draft?
The earlier file may have been stored in a submission repository. Confirm the matched source and contact the authorised instructor or administrator about repository handling. Do not conceal legitimate self-overlap.
Is synonym replacement enough to correct similarity?
No. It often leaves the source’s structure and reasoning unchanged. A sound revision explains the idea independently, integrates it into your argument and retains accurate attribution.
Can anyone guarantee a specific Turnitin percentage?
No responsible editor can guarantee one. The outcome depends on the final document, database coverage, assignment settings, exclusions and institutional rules.
What files are needed for a professional report review?
Usually, the editable manuscript, complete similarity report, required citation style, institutional policy and relevant supervisor comments are needed. A screenshot of the score alone is insufficient.
Conclusion
Knowing how to analyse a Turnitin Similarity Report is more valuable than reacting to a colour or chasing an arbitrary number. Review the source and context of each meaningful match, separate legitimate overlap from weak attribution, and prioritise passages that affect academic originality. Proper correction may involve verified citations, accurate quotations, independent explanation, stronger source synthesis or disclosure of earlier work—not cosmetic word replacement. Keep required terminology intact and follow the university’s own policy because no universal percentage applies to every thesis. If the report is extensive or the matched passages are difficult to classify, share the full report and editable document with ThesisLikho for a confidential assessment of the required scope and next steps.
📞 Call/WhatsApp: +91 96438 02216
Website: https://thesislikho.com/