Can an “Anonymous” Quote Still Identify a Participant?

Can readers identify someone from a quotation that contains no name?

A participant's real name never appears in your manuscript. You attribute their words to "Participant 8," a pseudonym, or perhaps no speaker at all. On the page, the quotation looks anonymous.

But imagine the participant's colleague reading it.

The quotation mentions a distinctive position, an unusual event, a recognizable complaint, or a sequence of experiences that only a few people share. Suddenly, the missing name does not provide much protection. The reader may know exactly who spoke.

An unnamed quotation can still reveal its speaker

Anonymity depends on what remains, not merely on what was removed

Removing the name solves only the most obvious identification problem

Changing "Dr. Reyes" to "Participant 12" removes a direct identifier. It does not necessarily remove the information that makes Dr. Reyes recognizable.

Qualitative data are often rich in precisely the details that make a person's experience meaningful: occupation, relationships, location, chronology, institutional role, unusual events, personal history, and social context. Those details can also function as indirect identifiers.

UK Data Service guidance on anonymising qualitative text recommends considering names alongside geographic details, organizations, occupations, family relationships, rare characteristics, and other information that may permit identification. The problem therefore extends well beyond names.

Deductive disclosure means working out the identity from clues

In qualitative research, the term deductive disclosure is commonly used for situations in which a person's identity can be inferred from characteristics or contextual information even though direct identifiers have been removed.

Kaiser describes this as a particular challenge for qualitative researchers because detailed narratives can make individuals recognizable to people familiar with them or the research setting. A participant may be effectively anonymous to most readers while being obvious to a small but consequential group of insiders.

The people most likely to identify the participant may already know them

Researchers sometimes test anonymity by asking whether an ordinary journal reader could identify the participant. That can underestimate the actual risk.

A colleague may recognize a workplace story. A family member may recognize a personal event. A supervisor may recognize an employee's complaint. Another participant may remember who made a particular statement during a shared experience.

In Kaiser's discussion of qualitative confidentiality, the difficulty is precisely that someone within the local setting may possess enough background information to connect an otherwise unnamed account with a particular person.

This means the relevant question is often not "Would a stranger know who this is?" but "Would someone who knows this setting, event, or participant be able to work it out?"

A rare role can function almost like a name

Suppose a quotation is attributed to "a senior administrator." That may provide little identifying information in a national study involving hundreds of organizations.

Now suppose the methods section says the study took place in one small university and the quotation is attributed to "the university's only female vice president." The label may narrow the possibilities to one person even though her name never appears.

Roles, job titles, specialties, institutional positions, rare qualifications, and leadership responsibilities should therefore be assessed in relation to the study population rather than treated as inherently safe demographic descriptors.

Distinctive events can reveal the speaker

Life events may be even more revealing than demographic information. A participant might mention surviving a particular disaster, filing a well-known complaint, receiving a rare award, leading a controversial project, undergoing an unusual treatment, or being involved in a publicly reported incident.

People who know about that event may identify the participant immediately. Publicly searchable information can make the connection easier still.

A quotation can therefore contain no conventional identifier and still provide a remarkably efficient biographical search query.

Several weak clues can become one strong identifier

Consider the following characteristics separately: 44 years old, secondary-school teacher, lives in a particular province, teaches physics, recently completed a doctorate.

None necessarily identifies one person. Add the name of a small municipality, an unusual research specialty, and a distinctive professional achievement, and the combination may become highly revealing.

This is the same broader problem in which several apparently harmless details become identifying when combined. Anonymisation should therefore assess combinations rather than approving each detail separately.

The attribution can identify someone even when the quotation does not

Researchers often attach demographic information to quotations to help readers interpret them:

"Participant 14, female, 38, department chair, private university."

The quotation itself may be innocuous, while the attribution provides enough information to identify the participant within the research setting.

Morse and Coulehan caution against unnecessary demographic detail in qualitative publications and argue that confidentiality problems can arise when participant characteristics allow readers to infer identity. Attribution should therefore contain information because it is analytically useful, not merely because it is available.

Repeated quotations can accumulate identifying information

A participant code can help readers follow one person's perspective across an article. That can be analytically valuable, but it can also create what amounts to a composite profile.

One quotation reveals occupation. Another mentions family structure. A third gives approximate age. A fourth describes a distinctive incident. No single quotation identifies the participant, but readers can combine them because each is labeled "P07."

Researchers should therefore consider identification risk across the entire manuscript, thesis, report, or dataset rather than reviewing quotations one at a time.

The surrounding manuscript can provide the missing clues

A quotation cannot be evaluated independently of the paper around it. The methods section may reveal the institution. A participant table may give demographic profiles. The results may identify subgroups. Another quotation may supply the final missing clue.

Research teams should therefore perform a manuscript-level confidentiality review. Someone reviewing only a spreadsheet of quotations may miss identifying information elsewhere in the document.

Searching an exact quotation can create another route to identification

Some qualitative research draws from public online material, social media, discussion forums, blogs, or other searchable sources. In these contexts, reproducing text verbatim can sometimes allow readers to paste the quotation into a search engine and locate its original source.

Research ethics scholarship on Internet research has repeatedly identified this problem. A username may be removed while the exact words remain searchable, effectively reconstructing the connection to the original speaker.

The appropriate treatment depends on the research context, source, consent arrangements, ethical review, and disciplinary norms. Researchers working with searchable online material should not assume that deleting a username makes a verbatim quotation anonymous.

A quotation identifies its speaker without ever naming her

The risk emerges from the interaction between the quotation and its context. Neither can be assessed adequately on its own.

Why apparently anonymous quotations still expose participants

Test quotations from the perspective of an informed insider

The most useful confidentiality check is not simply a search for names. Read the quotation as someone who knows the setting and ask what clues they could connect.

This does not require stripping qualitative research of every meaningful detail. The goal is to retain the context necessary for interpretation while avoiding contextual detail whose principal contribution is telling knowledgeable readers who the participant must be.

Could someone work out who said the quotation?

Questions about identification through anonymous quotations

Anonymous on the page does not always mean anonymous to the reader

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.