Is Covert Observation a Form of Deception?

If people do not know they are being studied, have researchers deceived them?

A researcher sits in a public plaza and records patterns of pedestrian behavior without approaching anyone. An ethnographer joins a group without revealing that observations will become research data. A researcher enters an online community under an ordinary user account and quietly studies conversations.

All three situations can be described as covert observation because the people being observed do not know that research is occurring. Yet they are not necessarily ethically equivalent, and they do not necessarily involve deception in exactly the same sense.

The central distinction is between failing to disclose the researcher's role and actively creating a false belief about that role. Even when no explicit lie occurs, covert observation can still raise serious questions about consent, reasonable expectations of privacy, confidentiality, identifiability, and whether researchers have entered or influenced the setting they are studying.

Covert observation can involve deception, but it is not automatically active deception

Covert observation sits at the intersection of disclosure, consent, and privacy

Observation without disclosure is not necessarily the same as telling a lie

Active deception ordinarily involves intentionally giving participants false or misleading information. Covert observation can occur without any communication between the researcher and the people being observed.

A researcher quietly counting how many pedestrians use stairs rather than an escalator in a public transportation hub, for example, may never tell anyone anything false. People simply do not know that their publicly visible behavior is being recorded for research.

That differs from a researcher entering a private group and falsely claiming to be an ordinary member in order to observe conversations. The second design adds an intentional misrepresentation of identity or purpose.

This parallels the broader difference between deception and incomplete disclosure. Absence of information and provision of false information can overlap ethically without being conceptually identical.

Non-participant and participant observation raise different issues

TCPS 2 distinguishes non-participant observational research, sometimes called naturalistic observation, from participant observational research.

In non-participant observation, the researcher observes activity without intervening. In participant observation, the researcher participates in the activity while also observing it. Participant observation is common in ethnographic research and may be overt or covert.

The distinction matters because joining the activity can change the researcher's relationship with the people being studied. A researcher who interacts, forms relationships, asks questions, influences events, or assumes a social role creates ethical issues beyond those involved in merely witnessing publicly visible behavior.

Public observation may sometimes occur without prospective consent

Under the U.S. Common Rule, certain research involving observation of public behavior can qualify for an exemption when the regulatory conditions are satisfied. The revised Common Rule's exemption at 45 CFR 46.104(d)(2) covers specified research involving educational tests, surveys, interviews, or observation of public behavior, including visual or auditory recording, when at least one of the applicable conditions is met.

These conditions concern matters such as whether identities can readily be ascertained, whether disclosure of responses could create specified harms, or whether limited IRB review provides appropriate privacy and confidentiality protections in qualifying circumstances.

HHS's preamble to the revised Common Rule explains part of the rationale: when behavior is genuinely public, it can ordinarily be observed by others even when the person does not know that a researcher is watching.

That does not mean every covert observational study is exempt, nor should researchers make exemption determinations merely from the everyday meaning of “public.” Institutional procedures should be followed for determining whether a study qualifies.

A public location does not eliminate every expectation of privacy

One of the most important mistakes in covert research is equating physical accessibility with ethical publicity.

TCPS 2 cautions that something publicly accessible may nevertheless be considered private within the culture or context of prospective participants. People involved in religious practices, online communities, or other activities may reasonably expect some degree of privacy even when an outsider could technically gain access.

Researchers should therefore ask what people reasonably understand about the setting, who they expect to be present, what kind of observation they anticipate, and how the resulting information may be used.

Online spaces make “public” especially complicated

An unrestricted public webpage is different from a closed support group, invitation-only forum, private messaging channel, or community in which members reasonably understand conversations to remain within the group.

SACHRP guidance concerning internet research has treated online material legally available to any internet user without specific permission as potentially constituting public behavior for Common Rule purposes, while also cautioning that public-behavior status does not automatically make every observational activity exempt from IRB review.

TCPS 2 similarly emphasizes expectations of privacy in virtual settings. Researchers studying online behavior should therefore avoid assuming that technical accessibility alone determines the ethical status of the material.

Identifiability can change the ethical stakes

Counting anonymous behavior in a public space is different from recording faces, usernames, license plates, distinctive quotations, or other information that allows individuals to be identified.

Identifiability can increase privacy and confidentiality risks, particularly when the behavior concerns stigmatized, illegal, professionally damaging, or otherwise sensitive activity.

The Common Rule's public-behavior exemption explicitly considers how information is recorded and the consequences of disclosure. TCPS 2 likewise instructs researchers to consider whether dissemination through photographs, recordings, or other outputs could identify individuals observed in public settings.

Recording can create additional obligations

Watching behavior and making an anonymous tally is not ethically identical to recording identifiable audio or video.

APA Standard 8.03 requires psychologists to obtain informed consent before recording voices or images for research unless specified exceptions apply. One exception covers research consisting solely of naturalistic observations in public places when the recording is not anticipated to be used in a manner that could cause personal identification or harm. Another concerns research designs involving deception, with consent for use of the recording obtained during debriefing.

Researchers should therefore examine the recording itself, not merely the fact that observation occurs in a public place.

Covert participation can cross more clearly into deception

Suppose an ethnographer joins a private organization under a fabricated biography, develops relationships with members, participates in meetings, and records observations without revealing the research role.

That situation differs substantially from watching pedestrian behavior in a plaza. The researcher has not merely withheld a scholarly purpose. A false identity or role may actively shape other people's decisions about access, trust, communication, and interaction.

The ethical analysis therefore needs to consider whether the deception itself can be justified, in addition to questions about observational research and consent.

Covert observation should not be justified merely because overt observation changes behavior

Researchers may argue that people would behave differently if they knew they were being observed. That concern can be methodologically important. Observation awareness can alter behavior and threaten the validity of naturalistic research.

But the same principle applies here as elsewhere in deceptive research: methodological usefulness is not automatic ethical permission. Researchers should explain why overt observation would materially compromise the research and why the proposed covert method satisfies the applicable requirements.

The fact that knowing about the research might change participant behavior is part of the justification, not the end of it.

Privacy may matter even when consent is not required

Consent and privacy are related but distinct ethical issues.

A framework may permit observation without prospective consent in a particular public context while still requiring researchers to minimize collection of identifiable information, protect confidentiality, avoid unnecessary intrusion, or limit how observations are disseminated.

Researchers should therefore avoid reasoning that “consent is not required, so anything observable may be recorded and published.” The method of collection and the form of dissemination can create risks independently of the consent question.

Debriefing covert observation is not always straightforward

If researchers never know who the observed individuals are, individual debriefing may be impossible. In other settings, approaching people afterward to explain that they were covertly observed may itself reveal sensitive information, increase intrusion, or undermine the rationale for the research.

Other covert studies may permit debriefing and, depending on the applicable framework, require it or strongly favor it. If the researcher interacted deceptively with identifiable participants, the argument for subsequent explanation may be considerably stronger than in anonymous observation of ordinary public behavior.

The question therefore connects directly to when research may proceed without individual debriefing. The answer depends on the nature of the observation, governing ethics framework, feasibility of recontact, participant welfare, and whether the study involved active deception.

Three covert observations with very different ethical profiles

Calling all three simply “covert observation” hides ethically important differences. Researchers and ethics reviewers need to examine what the researcher actually does, what people reasonably expect, and what information is collected and retained.

Common misconceptions about covert observational research

Classify what you are doing before deciding whether covert observation is acceptable

Begin with the setting and researcher behavior rather than the label “covert.” Identify whether you are simply observing, recording, interacting, intervening, joining the group, or actively misrepresenting your identity or purpose.

Then ask what the people being studied reasonably expect. The fact that a researcher can gain access does not necessarily mean participants understand their behavior as available for systematic research and publication.

Finally, plan dissemination as carefully as data collection. A covert observation that creates little risk when recorded anonymously can become much more consequential if a publication includes recognizable photographs, usernames, quotations searchable online, or descriptions that make individuals or small communities identifiable.

Before conducting covert observation, check the setting, role, and privacy expectations

Frequently asked questions about covert observation

Covert observation is not defined by deception alone

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.