Can You Quote a Participant Without Revealing Who They Are?

Can you preserve a participant's words without exposing their identity?

Direct quotations are one of the strengths of qualitative research. They let readers see the evidence behind an interpretation and preserve something of the participant's own language rather than reducing every account to the researcher's paraphrase.

But quotations create a confidentiality problem. A participant's name may be replaced with a pseudonym or code while the quotation itself reveals their job, location, family circumstances, distinctive experience, or another detail that makes the speaker recognizable.

So yes, researchers can often quote participants while protecting identity, but confidentiality depends on more than changing the name attached to the quotation.

Protect the identity inside the quotation as well as the label beside it

The quotation itself can contain the identifier

A pseudonym protects the name, not necessarily the person

Replacing "Maria Santos" with "Participant 7," "Teacher B," or a pseudonym prevents the publication from directly naming the participant. That is useful, but it addresses only one route to identification.

Consider a quotation such as: "When I became the only female principal in the district in 2024, everyone knew the new policy was mine."

Changing the attribution to "Participant 7" does little if readers familiar with the district know exactly who held that position.

This is why qualitative confidentiality cannot be reduced to pseudonym assignment. Researchers need to consider what the quotation says and what readers can infer from it.

Deductive disclosure occurs when readers can work out who the participant is

Qualitative research often involves small samples, specific settings, and rich descriptions. Those features are methodologically valuable, but they can also create deductive disclosure: a reader identifies a participant by combining characteristics reported in the study rather than by encountering an explicit name.

Research on qualitative confidentiality has long documented this problem. Kaiser, for example, describes situations in which a participant could be recognizable to local clinicians even after her name was changed because her circumstances were sufficiently unusual within the study setting.

The most relevant audience may therefore be people who know the participant rather than strangers reading the article. Coworkers, supervisors, relatives, classmates, community members, clinicians, or other participants may possess contextual knowledge that the general public lacks.

This is closely related to the problem that an apparently anonymous quotation can still identify its speaker.

Distinctive roles can reveal the speaker

Some roles sound innocuous until the study setting is considered. "Department chair" may describe hundreds of people across a national dataset. "The department chair" in a study of one small academic program may describe exactly one.

A methodological discussion of quotations in qualitative research illustrates this problem with a quotation attributed to a competency-committee chair. Even when a participant code is used, the person's institutional role and the surrounding study description may allow readers familiar with the setting to identify the speaker.

Before publishing a quotation, ask whether occupational titles, organizational positions, seniority, specialization, or institutional affiliations unnecessarily narrow the possible speaker.

Distinctive experiences can be as identifying as demographic variables

A participant may describe an unusual illness, workplace dispute, award, accident, migration history, disciplinary case, public controversy, family event, or other experience. Even if names and locations are removed, that story may be recognizable to people who already know what happened.

The risk increases when the event has been publicly reported or when it occurred in a small community. Searchable online information can also make contextual clues easier to connect with a real person.

Researchers should therefore consider how several individually harmless details can become identifying when combined.

Attribution labels can add unnecessary clues

The words of the quotation are not the only source of disclosure. Researchers often append descriptors to help readers interpret quotations:

"Female professor, 42, private university, Department Chair."

Each descriptor may be analytically relevant, but together they may substantially narrow the possibilities. In a small sample, repeated participant codes can also allow readers to assemble information about the same person across an article.

Morse and Coulehan have argued that demographic information in qualitative publications should be limited to what is pertinent and that participant codes used analytically should not automatically be carried into publication when they provide no interpretive value and create additional linkability.

This does not mean participant characteristics should always be hidden. The question is whether each descriptor helps readers interpret the finding enough to justify the additional disclosure risk.

Removing identifying details may be possible without changing the meaning

Some identifying information can be omitted or generalized while preserving the analytical point. A specific hospital might become "a regional hospital." An exact age might become an age range. A person's highly specific job title might be represented by a broader occupational category if the distinction is not analytically important.

Minor editorial changes can also be indicated transparently when scholarly conventions require it. Researchers should preserve the substance of what the participant said rather than quietly rewriting a quotation into a more convenient version.

However, anonymisation is not simply an editing exercise. If changing a characteristic alters the interpretation of the quotation, the researcher faces a substantive methodological and ethical decision. The question of whether researchers should alter details in qualitative quotations therefore requires careful attention to both confidentiality and fidelity.

Sometimes removing the identifier removes the point

Suppose a participant describes discrimination specifically because they occupy a rare occupational role. Replacing that role with "employee" may protect identity, but it may also erase the mechanism the study is trying to explain.

Researchers should not pretend that confidentiality and analytical richness always align neatly. Sometimes they conflict.

Possible responses depend on the circumstances. Researchers may omit the quotation, paraphrase the finding, reduce other contextual details, combine supporting evidence, seek appropriate participant permission where ethically and methodologically suitable, or use controlled rather than unrestricted dissemination. Each approach has trade-offs.

When removing identifying information changes the meaning of qualitative data, that tension should be resolved deliberately rather than disguised through silent alteration.

Quotations can identify people mentioned by the participant too

A quotation may protect the speaker while revealing someone else. Participants commonly mention supervisors, spouses, patients, colleagues, students, relatives, organizations, or other individuals during interviews.

Researchers should therefore review quotations for third-party identifiers as well as participant identifiers. A statement such as "my supervisor, the only cardiologist at our clinic..." may expose another person even if the participant remains anonymous.

Not every study promises anonymity

Confidentiality should not be treated as a universal requirement to erase participant identity regardless of research design or participant preference.

The Canadian Tri-Council Policy Statement, for example, recognizes qualitative contexts such as oral history and biographical research in which participants may wish to be identified and requires researchers planning identity disclosure to discuss that choice with participants and record their consent.

Named participation can therefore be legitimate when appropriate to the methodology, consent process, ethics review, and participant wishes. What researchers should avoid is accidental identification when confidentiality was promised.

A quotation that names nobody but still identifies the speaker

The quotation is not identifying because it contains a secret code. It is identifying because the story itself narrows the possible speaker.

Common mistakes when protecting identity in participant quotations

Read each quotation as someone who knows the participant

Before publishing a quotation, temporarily stop reading it as the researcher who already knows everyone's identity. Imagine instead that you are the participant's colleague, supervisor, relative, classmate, or community member.

What clues would you notice?

The goal is not to remove every trace of context from qualitative research. It is to retain the context needed to support interpretation without casually providing a trail back to someone whose identity you promised to protect.

Before publishing a participant quotation

Questions about anonymous and confidential participant quotations

A quotation is confidential only if the person behind it remains protected

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.