03 · What You Need to Know
Migration status can matter ethically without defining the whole participant
“Refugee,” “migrant,” and “precarious legal status” are not interchangeable categories
These terms describe different legal, administrative, and social circumstances. A refugee recognized under an applicable legal framework, an asylum seeker awaiting a decision, a labor migrant with authorization to work, an international student, an internally displaced person, a person without regular immigration status, and a stateless person may face very different risks.
Researchers should use terminology accurately and avoid treating mobility itself as evidence of one uniform vulnerability.
This follows the broader principle that membership in a population associated with vulnerability does not make every member vulnerable in the same way.
Legal status can make ordinary data unusually consequential
Names, nationality, current location, immigration category, employer, migration history, family relationships, or documentation status may carry consequences beyond the research itself.
The exact risks depend on the jurisdiction and participant circumstances. Researchers should therefore avoid generic claims about what authorities can or cannot obtain and instead verify the legal protections, disclosure obligations, and data-governance rules that actually apply.
UNHCR's current data-responsibility guidance emphasizes that work involving forcibly displaced and stateless people should place rights, dignity, well-being, purpose, proportionality, and responsible handling of sensitive data at the center of design.
Watch Out
Do not collect immigration status, document numbers, exact locations, migration routes, or other sensitive identifiers merely because they might become analytically useful later. If the variable is not necessary for the research question, not collecting it may provide stronger protection than promising to secure it after collection.
Research participation itself can reveal something sensitive
Researchers often focus on protecting completed questionnaires and interview transcripts. But disclosure can happen earlier.
Being approached at a refugee service center, entering a room associated with an asylum study, receiving a visibly labeled message, or being seen speaking with a particular research team may reveal something about a person's circumstances.
For participants whose status is stigmatized or legally sensitive, recruitment privacy can therefore matter as much as database security.
This is an application of the broader problem of social stigma, legal status, and unusually consequential disclosure.
Language access is part of informed consent, not an administrative courtesy
A participant cannot make an adequately informed decision if essential research information is presented in a language they do not understand sufficiently.
Translation alone may not solve every problem. Technical research terminology, legal vocabulary, literacy differences, unfamiliar consent conventions, and culturally specific concepts can affect understanding even when words have been translated accurately.
| Potential communication problem |
Ethical concern |
Possible response |
| Participant has limited proficiency in the research team's language |
Incomplete understanding of participation |
Provide appropriate translated information and interpretation consistent with applicable requirements |
| Consent document is technically translated but highly legalistic |
Words are translated without making the decision understandable |
Use clear explanations appropriate to the participant population while preserving required information |
| Family member acts as interpreter |
Privacy, accuracy, or voluntariness may be affected |
Consider whether qualified independent interpretation is more appropriate, particularly for sensitive research |
| Participant has limited literacy |
Written consent becomes inaccessible |
Use an approved consent process that communicates information through appropriate formats rather than assuming inability to consent |
Communication barriers should not be confused with decision-making incapacity. A person who cannot read the research team's language may remain entirely capable of deciding once information is communicated accessibly.
Interpreters can become part of the confidentiality problem
An interpreter hears the research conversation. In a small linguistic or diaspora community, the interpreter may also know participants socially.
That can affect willingness to disclose sensitive experiences involving legal status, violence, employment, health, family relationships, or migration history.
Researchers should therefore consider interpreter confidentiality, training, role boundaries, and potential conflicts or community relationships. The most linguistically fluent person is not automatically the most ethically appropriate interpreter for sensitive research.
Dependence on humanitarian or social services can affect voluntariness
Refugees and displaced people may receive healthcare, shelter, food assistance, legal services, cash support, documentation assistance, or other resources through organizations that also facilitate research.
Participants may reasonably wonder whether agreeing will improve access or whether refusal could affect assistance.
Researchers should make clear that research participation is separate from ordinary aid and services and design recruitment so that this separation is credible in practice.
This is a form of dependency-based research vulnerability, not evidence that displaced participants are incapable of choosing.
Researchers should not blur research with legal or humanitarian decision-making
A participant may not know whether the interviewer is a researcher, aid worker, government representative, immigration official, clinician, or someone whose report could affect an asylum or service decision.
Consent should therefore clarify the researcher's role, the purpose of the study, who is conducting or funding it, how information will be used, and whether participation has any relationship to services or legal processes.
The more consequential the misunderstanding could be, the more important this role clarity becomes.
Trauma can matter without defining everyone who has been displaced
Some refugees and migrants have experienced war, persecution, violence, trafficking, family separation, detention, dangerous journeys, or other traumatic events. Others may not have experienced these events, may not experience them as traumatic, or may not be asked about them in the study.
Researchers should therefore avoid treating “refugee” as synonymous with “traumatized.”
When research does involve potentially distressing experiences, interview design, interviewer training, participant control over disclosure, stopping procedures, and appropriate support or referral planning may require attention. These safeguards should correspond to the actual topic and participant circumstances rather than a stereotype.
Repeated requests for personal stories can themselves become burdensome
Populations receiving humanitarian assistance can be approached repeatedly by researchers, evaluators, journalists, agencies, and organizations asking similar questions.
Researchers should ask whether the information already exists, whether another interview is scientifically necessary, and what participants receive from repeatedly recounting sensitive experiences.
Scientific curiosity alone is a weak justification for collecting burdensome personal narratives that do not materially contribute to a well-designed research question.
Data minimization is particularly important when status is sensitive
UNHCR emphasizes responsible data management across collection, storage, processing, analysis, sharing, dissemination, and use, with protection and “do no harm” considerations central to work involving forcibly displaced and stateless people.
For researchers, this suggests a practical sequence:
Ask why
Why is this sensitive variable necessary?
Ask how precise
Do you need an exact legal category, location, date, or identifier, or would a less granular variable answer the research question?
Ask who
Who genuinely needs access to identifiable information?
Ask what happens later
Will data be shared, archived, linked, reused, transferred across borders, or published in a form that changes identification risk?
The safest sensitive datum is sometimes the one the research never needed to collect.
Cross-border research complicates legal and data governance
Migration research may involve researchers in one country, participants in another, cloud services in a third, and collaborators elsewhere. Data may cross legal jurisdictions even when participants do not.
Researchers should therefore identify which privacy, data-protection, research-ethics, institutional, and contractual requirements govern collection, transfer, storage, access, and sharing.
There is no universal confidentiality statement suitable for every cross-border migration study. Current requirements need to be verified for the actual jurisdictions and systems involved.
Mobility can complicate follow-up and continuing consent
Participants may relocate, cross borders, enter different accommodation, lose telephone access, change numbers, or become unreachable. These circumstances can affect longitudinal research, safety monitoring, withdrawal requests, return of results, and continued communication.
Researchers should decide prospectively what happens when contact is lost and avoid collecting excessive tracking information simply to make follow-up easier.
Excluding migrants can also produce ethically important evidence gaps
Language differences, mobility, legal complexity, and additional ethics work can make migrant populations operationally difficult to include. Researchers may therefore exclude them through language criteria, documentation requirements, stable-address requirements, or recruitment procedures.
Some exclusions may be scientifically necessary. Others may merely make the study easier to run.
When the research concerns health systems, employment, education, social services, public health, or other areas affecting migrant populations, systematic exclusion can produce evidence that does not represent them. The ethical problem then shifts from risky inclusion to unfair absence.
