An applicant reads a dozen sample personal statements to understand what a strong one looks like, then writes their own from scratch. The opening has a familiar rhythm, one paragraph about a lifelong interest and another about a reading list. Weeks later, an admissions service tells them their statement shows similarity to existing text, and they are baffled, because they never copied anything.
Admissions essays are one of the few places where a similarity check compares a document against a library built specifically for that purpose, and the way those libraries work explains a lot about why honest applicants sometimes get flagged.
A Library of Past Statements, Not Just the Web
In the United Kingdom, UCAS runs every undergraduate personal statement through its Similarity Detection Service, which uses software called Copycatch. According to guidance published by UCAS and by universities that follow its process, each statement is compared against a library of personal statements previously submitted to UCAS and a library of sample statements collected from websites and other sources.
Each new statement is added to that library once it has been processed. A statement an applicant writes this year therefore becomes part of the comparison set for applicants next year, which is one reason published guidance tells applicants not to post their statement online or share it with others.
Why Honest Applicants Still Get Matches
Personal statements are short and follow a predictable arc: a spark of interest, relevant experience, evidence of reading, and a link to the course. Applicants working from the same advice and the same sample statements often reach for the same openings and the same transitions, so overlap can appear even when no one copied a sentence.
The risk is highest when a sample statement has influenced the structure and wording of a draft more than the writer realizes. A paragraph that follows a sample’s order of ideas, uses its linking phrases, and swaps in the applicant’s own subject can look very similar to its model even though every word was typed independently.
Advisers who help applicants often see this pattern. A teacher reading a draft may recognize the borrowed shape of a well-known sample, and the most useful feedback is usually to replace the model’s order of ideas with the applicant’s own, starting from the experience that mattered most to them rather than from the structure a sample suggested.
What Happens After a Flag
UCAS-related procedures published by universities describe a human step. Statements showing a potential similarity level of 30 percent or greater are reviewed by members of the UCAS Verification Team, and applicants are contacted and made aware of the similarities. Universities and colleges then decide the significance of the results and any action to take.
Responses vary by institution. One university’s published procedure, for example, notes that it may ask an applicant to submit a new statement if similarity is over 45 percent. Because each provider sets its own response, the same flag can lead to different outcomes depending on who receives it. An applicant who receives a notice should read it carefully, keep their drafts, and respond with a clear, honest account of how the statement was written.
Why the Library Keeps Growing
Because every processed statement is added to the library, and because sample statements are collected from websites, the comparison set grows each cycle. A statement that an earlier applicant posted publicly can later be matched against someone who read it, remembered a line, and reused the idea without meaning to.
The Uni Guide, an advice site for UK applicants, makes the same point in practical terms: the similarity checker includes statements it has found on the internet, and applicants are advised to wait until they have a university place before posting their statement anywhere.
Similarity Checks and AI Rules Are Separate Questions
A similarity check compares a statement with existing text. It is not the same thing as the rules about AI assistance, which UCAS addresses separately through its own-work declaration and its guidance on acceptable and unacceptable uses. Applicants should read both, since passing one question says nothing about the other.
UCAS is a documented example rather than a universal model. Other countries and institutions run their own checks, with different libraries, thresholds, and policies, so applicants should read the instructions attached to each application they submit.
A Pre-Submission Routine for Applicants
The most reliable protection is writing from specific experience that no sample could contain. A detail about a particular book, a specific problem solved, or a real conversation is hard to match, and it makes a statement more persuasive besides.
A short routine helps catch problems early:
- Read samples for understanding, then close them before writing the first draft
- Draft from your own experiences, not from a sample’s order of ideas
- Do not post or share the finished statement, since published statements can later be matched
- Check the draft for matches before submitting, and rewrite any passage that follows a model too closely
- Keep your drafts and notes in case a question ever arises
A plagiarism checker can show which sentences resemble existing text and link to the matching source, which turns a vague worry into a specific list of lines to revise. Results will not mirror any admissions service’s private library, so a clean report is reassuring but not a guarantee.
Specific Beats Familiar
Admissions readers see hundreds of statements that sound alike. The sentence an applicant can only write about their own life is both the safest from a similarity check and the most memorable to a reader. Choosing specific over familiar protects the application on both counts, and it is a habit that serves an applicant well in every later piece of writing that has to stand out from a crowd.
For more on how similarity checks treat short, formulaic writing, further reading on the Phrasly blog covers the underlying research for applicants and the advisers who support them.
