Similarity tools for reviewing a thesis (2026)

How to compare reports without misleading rankings

Instead of declaring one tool “definitive,” compare these verifiable criteria:

Criterion	Question to answer
Access	Is the license institutional, individual or temporary?
Coverage	Which source types does the provider say it compares?
Privacy	Does it retain the file or add it to a collection?
Settings	Can quotations, references or short matches be excluded?
Evidence	Does it show the source and matching passage?
Review	Who interprets the report and under which policy?

Turnitin, iThenticate, Compilatio and other platforms change their features, limits and plans. Check the provider's current documentation and your program's instructions at the time of use. We do not publish prices or database sizes as if they were permanent.

A step-by-step review method

1. Keep an identifiable copy

Save the submitted version with a clear date and filename. A report is interpretable only when you know exactly which file and settings produced it.

2. Read the policy before the score

Confirm whether an official rubric or instruction exists. Do not transfer a threshold from another university, faculty or course. If no public rule exists, ask the program coordinator or adviser.

3. Review the main sources

Begin with long or repeated matches to one source. Open the original and classify the case: direct quotation, paraphrase, unavoidable term, reference or unattributed text.

4. Correct the actual problem

• If it is a direct quotation, apply quotation marks or block format, author, date and locator under the required style.
• If the idea comes from another source, attribute it even when the wording is your own.
• If the passage reproduces structure and vocabulary, return to the source, understand the idea and write a genuine synthesis.
• If it is your own finding or data, retain the evidence that supports it.

Do not swap words merely to “beat” a detector. That process can damage meaning and does not resolve attribution.

5. Document decisions

Record which matches you reviewed, which source you consulted and what changed. If the program allows quotation or bibliography exclusions, retain the settings used. You can then explain the result without depending on an isolated screenshot.

Matches that require attention

A correctly attributed quotation may still appear in a report. That does not make it plagiarism, although it may affect the score. These cases deserve priority review:

• paragraphs or tables reproduced without a source;
• translations of another text presented as original work;
• paraphrases that preserve the original structure;
• copied methods, instruments or definitions without attribution;
• reuse of earlier work without disclosure when policy requires it;
• references that do not support the claim.

Also inspect appendices, consent forms, instruments and templates. Their treatment may differ from the thesis body and should follow the program's criteria.

Privacy practices

Before uploading a file to an external tool, remove personal data that is not necessary for the check. Do not upload participant datasets, case files, signed consent forms, clinical information or trade secrets. A similarity tool does not replace ethics review or data-protection rules.

If you only need a citation and writing review, share the minimum required section. Store results securely and remove temporary copies when they no longer serve a purpose.

What to record with a report

When your program asks for it, document the tool name, date, file version, settings and reviewer. A score alone cannot reproduce the result. If a second review was performed, distinguish both versions clearly.

Folium Labs' educational review focuses on attribution, citations, legitimate paraphrasing and traceability. We do not alter text to evade detectors or promise a score. Review the service scope or send the document and current guide for a quote.