Can Researchers Protect Confidentiality When They Know Participants Personally?

What Happens to Confidentiality When the Researcher Already Knows the Participant?

A lecturer interviews colleagues from their own department. A principal studies teachers in their school. A clinician conducts research involving patients they already know. A community researcher interviews neighbours.

The researcher does not need a code key to work out who Participant 7 is. They knew before the study began.

That does not make confidential research impossible. It does mean that removing names from a spreadsheet solves only part of the problem. Prior relationships affect what the researcher already knows, what participants may assume, what others can infer, and how research information must be kept separate from other roles.

Yes, but Confidentiality Requires More Than De-Identifying the Dataset

Knowing the Participant Changes Where Confidentiality Risk Comes From

Confidentiality Does Not Require the Researcher to Be Unable to Identify Participants

Confidentiality and anonymity are different protections. A researcher can know exactly who a participant is while restricting how that person's research information is used and disclosed.

This distinction matters especially in insider research. The researcher may interact directly with participants, recognise their voices, know their workplaces, or already understand the events they describe.

The study is therefore generally not anonymous to that researcher merely because transcripts later use pseudonyms. It may still be confidential if identifiable information is appropriately controlled.

This follows the broader distinction between anonymous and confidential research.

Prior Knowledge Can Identify Participants Even After Their Names Are Removed

Suppose a researcher interviews six colleagues and labels the transcripts P1 through P6. An external analyst may not know who P4 is. The researcher may recognise P4 immediately from the person's role, experiences, speech, or interview itself.

Under the US Common Rule, identifiable private information includes private information for which the participant's identity is or may readily be ascertained by the investigator or is associated with the information. OHRP also treats coded information as identifiable to investigators when they can link it to specific individuals directly or indirectly through coding systems.

Coding therefore remains useful for reducing unnecessary exposure, but it does not transform known participants into strangers.

The Researcher May Know More Than the Dataset Contains

Insider researchers bring contextual knowledge to the data. A transcript might say only "a senior colleague," while the researcher knows exactly which colleague the participant means. A participant may mention "the incident last semester," assuming the researcher already understands the reference.

This means identifiability does not reside solely in the written transcript. It can arise from the combination of the research material and the researcher's prior knowledge.

The ICO's anonymisation guidance treats identifiability as contextual and identifies singling out and linkability as key considerations. Whether information distinguishes a person depends partly on what the relevant party already knows and what additional information can be connected to the data.

Participants May Also Know That the Researcher Knows

The prior relationship can affect what participants disclose. An employee interviewed by their manager may wonder whether criticism will affect future treatment. A student interviewed by their lecturer may worry about assessment. A patient may have difficulty separating a research invitation from clinical care.

These concerns extend beyond confidentiality into voluntariness, power, and role conflict, but they also shape confidentiality expectations. Participants may reasonably want to know whether information disclosed for research could influence the researcher's behaviour toward them in another role.

A strong confidentiality plan should therefore specify how research information will be compartmentalised from decisions the researcher makes as supervisor, teacher, clinician, administrator, colleague, or community member where applicable.

Knowing Someone Before the Study and Learning Something Through the Study Are Different

An insider researcher may already know that a colleague holds a particular position or experienced a public event. During an interview, that colleague may disclose sensitive information about workplace conflict, health, finances, relationships, or other matters.

The fact that the researcher already knew the participant does not make the newly disclosed research information ordinary workplace knowledge.

Researchers should treat information obtained through the research relationship according to the confidentiality commitments and requirements governing the study. Existing familiarity should not become a shortcut for carrying research disclosures into unrelated conversations or decisions.

Recruitment Can Reveal Participation Before the Interview Even Begins

In a close workplace or community, participation itself may reveal something. Colleagues may see someone entering the researcher's office, attending an interview, or responding to a recruitment message.

If eligibility relates to a sensitive characteristic or experience, merely being identified as a participant can disclose information.

Researchers should therefore consider privacy during recruitment and scheduling, not only confidentiality after data collection. Private communication channels, neutral meeting arrangements, and recruitment procedures that do not unnecessarily reveal who agreed or declined can matter.

Third-Party Information Becomes Especially Difficult in Insider Research

Participants frequently talk about other people. In insider research, the researcher may know those people too.

A participant does not need to say, "Dr. Santos did this." They might say, "the only associate dean who joined us last year," and the researcher may immediately know whom they mean.

OHRP's exploratory work on privacy in qualitative research has highlighted the possibility that interviews, focus groups, observation, and other qualitative methods can intentionally or unintentionally elicit sensitive or identifiable information about third parties.

Researchers should therefore plan how third-party identifiers will be treated in notes, transcripts, analysis, and publications rather than focusing exclusively on the identity of the enrolled participant.

Removing Names From Transcripts Still Has Value

The fact that the interviewer knows participants does not make de-identification pointless.

Replacing names with study codes or pseudonyms can reduce unnecessary exposure when transcripts are accessed by coders, supervisors, collaborators, transcription services, or other authorised people who do not need participant identities.

This is an application of role-based access. The interviewer may inevitably know who participated while another analyst can work with appropriately prepared material without receiving the same identifying information.

Researchers should determine who actually needs access to identifiable research information rather than concluding that because one researcher knows, everyone might as well know.

Shared Social Worlds Increase Contextual Recognition

Insider research creates a second audience problem: readers may know the participants too.

A paper might describe someone as "a newly appointed department head with 27 years of service." That description could be anonymous to an international reader and obvious to every employee in the institution.

Audio recordings can pose similar problems. HHS advisory materials have recognised that identifiability of recordings can depend on circumstances such as a limited sample and recognisable voices.

Researchers should therefore assess dissemination from the perspective of people inside the shared setting, not only anonymous journal readers.

Pseudonyms Can Fail When the Story Identifies the Person

Changing "Professor Reyes" to "Professor A" does little if the surrounding text describes a unique research centre, unusual career history, public dispute, distinctive quotation, or event known throughout the institution.

Qualitative confidentiality may require decisions about contextual detail, quotations, role descriptions, site names, timing, or combinations of attributes.

That does not mean every detail should be stripped away until the findings become methodologically empty. Researchers must balance interpretive value against disclosure risk and explain any transformations that materially affect the evidence.

Researchers Should Not Promise Themselves Impossible Cognitive Separation

A researcher cannot realistically promise to forget what a participant disclosed once the interview ends.

The more defensible commitment concerns behaviour: research information will not be intentionally used or disclosed outside the purposes and circumstances described by the study, subject to applicable limits of confidentiality.

This distinction is important when researchers occupy ongoing roles with participants. The goal is not fictional amnesia. It is disciplined handling of information across roles.

Another Researcher Can Sometimes Reduce the Conflict

Where the prior relationship creates substantial power or confidentiality concerns, a study may be designed so that another team member handles recruitment, consent, interviews, identifiable data, or particular analyses.

This is not always necessary or feasible. Insider knowledge can also contribute legitimate methodological insight, particularly in ethnographic, organisational, participatory, and community research.

The decision should therefore reflect the specific risks rather than treating researcher familiarity as inherently unethical.

Confidentiality in a Study of the Researcher's Own Department

The researcher's prior knowledge never disappears. Confidentiality is protected by controlling what happens because of that knowledge and by reducing unnecessary identification for everyone else.

Common Mistakes When Researchers Know Their Participants

Treat Prior Relationships as Part of the Confidentiality Architecture

If you know potential participants personally or professionally, map what that relationship changes before recruitment begins.

Insider research does not require pretending that relationships disappear at the interview-room door. It requires acknowledging that those relationships exist and designing confidentiality around the information pathways they create.

Before Researching People You Already Know

Frequently Asked Questions About Insider Research and Confidentiality

Knowing Participants Does Not Eliminate Confidentiality, but It Changes What Must Be 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.