A social work thesis studying “migrants” or “refugees” as though the terms were interchangeable, or as though the population were a single undifferentiated group, misses both the terminology precision and the ethical care this research area specifically requires. This guide covers how to scope the population correctly, the trauma-informed design and consent practices this research needs, interpreter and language considerations, and the access route that actually works with a population that has good reason to be cautious of researchers and institutions.

Terminology: migrant and refugee are not interchangeable

“Migrant” is a broad umbrella term covering anyone who moves across a border, for any reason — economic opportunity, family reunification, education, or displacement. “Refugee” refers to a specific legal status tied to displacement from persecution, conflict or serious harm, recognized under international and national legal frameworks with defined rights and protections attached. Using the terms interchangeably in a thesis is both imprecise and, for a reader familiar with the field, a signal that the writer has not engaged carefully with the population’s actual circumstances. State explicitly which term applies to your specific population and why — naming the relevant legal category (refugee, asylum seeker, undocumented migrant, labor migrant) rather than defaulting to a generic label, and noting where your population may include more than one category, which changes what access, consent and legal considerations apply.

Trauma-informed research design

A meaningful share of migrant and refugee populations, particularly those displaced by conflict or persecution, have experienced trauma, and a research design needs to anticipate this rather than treat data collection as identical to a general-population study. Core trauma-informed practices: give participants explicit control over what they choose to disclose rather than requiring a fixed set of questions be answered; avoid designs that require detailed retelling of traumatic events unless that retelling is genuinely central to the research question (and where it is, build in additional support); have a clear referral pathway ready — a named local mental-health or community-support resource — that a researcher can offer if a participant becomes distressed during data collection; and train yourself, if you are not already trained, in recognizing signs of distress and responding appropriately rather than pushing forward with a planned question sequence regardless. State your trauma-informed design decisions explicitly in the methods chapter, since a committee reviewing research with this population will specifically look for evidence that you considered this dimension, not just standard consent and confidentiality language.

Interpreters: professional, not a family member

Where language is a barrier, use a trained professional interpreter rather than a bilingual family member, friend, or community volunteer without interpreter training. A family member introduces several problems at once: confidentiality risk (the participant may not disclose sensitive information in front of a relative), potential coercion or filtering (a family member may soften, alter or withhold what the participant actually says), and translation-quality concerns, since accurate interpretation of nuanced or emotionally difficult content is a skilled task most untrained bilingual speakers are not equipped for. Budget for professional interpretation explicitly in your research plan and timeline — it adds cost and scheduling complexity, but is a standard expectation for rigorous research with a non-native-language population, not an optional enhancement. Once interviews are conducted through an interpreter, transcription needs its own care as well — our guide to transcribing research interviews covers verbatim versus intelligent-verbatim style and anonymization at the transcription stage, both of which matter more, not less, when a third-party interpreter was involved in the original exchange.

Power dynamics and researcher positionality

A researcher approaching a migrant or refugee population, particularly where the researcher holds citizenship, institutional affiliation, or language privilege the participant does not, occupies a position of relative power that shapes what a participant may feel able to say or decline. A brief, honest positionality statement addressing your own relationship to the population — your citizenship status, language background, any professional or volunteer history with the community, and how these might shape access, rapport, and interpretation of findings — is expected in this research area specifically, more so than in research with a general population. Our guide to reflexivity in qualitative research covers how to write this kind of positionality statement in more depth, including the personal, epistemic and methodological types of reflexivity relevant here. Address directly how your design mitigates the power imbalance where possible: emphasizing genuinely voluntary participation, making clear that declining to participate carries no consequence for any service the person receives, and ensuring the researcher is not in a position (such as a caseworker for the same agency) where participation could be confused with a service requirement.

Legal-status sensitivity: collect only what you need

A research design should ask whether collecting a participant’s specific legal immigration status is actually necessary for the research question, rather than gathering it as routine demographic information. If legal status is not directly relevant to your analysis, do not collect it — it introduces meaningful disclosure risk (data that could, in a worst case, be subpoenaed or otherwise compelled) with no corresponding analytic benefit. Where legal status genuinely is relevant to your research question, state explicitly in your consent process and data security plan how that information will be protected, including whether your data storage arrangement could ever be subject to legal compulsion, and what your institution’s policy is on protecting research data from such requests where relevant.

Access: community organizations, not cold outreach

Migrant and refugee communities, especially where trust in institutions and authorities has been damaged by prior experience, are rarely reachable through cold outreach or a generic recruitment flyer. Community-based organizations, resettlement agencies, faith communities, and established community leaders who already hold trust within the population are the access route that actually works — and building a genuine relationship with that organization, including being transparent about your research purpose and offering something back (a summary of findings, a connection to a relevant resource), matters as much as the formal ethics approval. Cold outreach not only produces lower response rates in this population specifically, it can also raise legitimate suspicion about your intentions given the population’s realistic concerns about surveillance, immigration enforcement, or exploitation by outside parties.

Reciprocity and community benefit

Because migrant and refugee communities are frequently studied without receiving anything back, building genuine reciprocity into your design — a plain-language summary of findings shared with the community, a connection to a relevant service the community organization identifies as needed, or simply respecting the community’s own priorities about what gets asked and reported — is both an ethical practice and, practically, what makes continued access and trust possible for future researchers working with the same community. State your reciprocity plan explicitly in your ethics application rather than treating it as an afterthought once data collection is complete.

Designing for a heterogeneous population

“Migrants” or “refugees” as a research population is rarely homogeneous — country of origin, time since arrival, family composition, legal status, and prior trauma exposure all vary substantially within what a thesis might otherwise treat as one group. A stronger design either narrows the population explicitly (a specific national-origin group, a specific arrival cohort, a specific legal-status category) or, where the population is intentionally broader, reports and discusses this heterogeneity directly rather than presenting findings as though the population were uniform. A thesis that collapses genuinely different sub-groups into one undifferentiated “migrant” or “refugee” category risks findings that do not actually generalize even within the stated population.

Data storage and security for a heightened-risk population

Beyond standard participant confidentiality, a study involving legal-status-sensitive information should document a specific data security plan: where data is stored, who has access, whether it is encrypted, and how long it is retained before secure deletion. State this plan in enough detail in your ethics application that a reviewer can judge whether it is proportionate to the actual sensitivity of what you are collecting, rather than relying on a generic institutional data-protection statement that does not address the specific risks this population faces.

A worked example

Weak framing: “This study examines the mental health needs of migrants.”

Stronger framing (illustrative): “This study examines post-resettlement mental health service access among refugees from one named conflict-affected country who were resettled in a single mid-sized U.S. city within the past three years, recruited through a partnership with a local resettlement agency, using a trauma-informed interview protocol with professional interpretation in the participants’ first language, and a data security plan addressing the sensitivity of any immigration-status information collected.”

The second version names the specific legal category and population, states the access partnership, and addresses the trauma-informed and interpretation considerations explicitly.

Comparing to other sensitive-population research

The scoping and ethics discipline this population requires overlaps significantly with the considerations any thesis studying a vulnerable or historically under-researched population faces. Our guide to social work dissertation methodology and research ethics covers the general vulnerable-population ethics framework — safeguarding, informed consent, gatekeeping — that this migrant- and refugee-specific application builds on, and our guide to qualitative research methods covers the broader methodological choices most studies in this area rely on.

Mistakes that create ethical or access problems

• Using “migrant” and “refugee” interchangeably without naming the specific legal category that applies to the population studied.
• No trauma-informed design considerations stated explicitly in the methods chapter.
• Using a family member as interpreter instead of a trained professional.
• Collecting legal-status information not actually necessary for the research question.
• Cold outreach instead of a community-organization partnership as the access route.
• Treating a heterogeneous population as uniform, without acknowledging or narrowing the variation within it.
• No reciprocity plan for the community providing access and data.
• No positionality statement addressing the researcher’s own relationship to the power dynamics involved.

Frequently asked questions

Should I use the term “migrant” or “refugee” in my thesis?

Use the term that accurately describes your specific population’s legal status and circumstances rather than treating the two as interchangeable, and state explicitly which term applies to your population and why.

Do I need an interpreter for interviews, or can a bilingual family member translate?

Use a professional, trained interpreter rather than a family member, since a family member introduces confidentiality risk, potential coercion, and translation quality concerns a trained interpreter is positioned to avoid.

Is asking about someone’s legal immigration status necessary for my study?

Only if it is directly relevant to your research question. If it is not necessary, avoid collecting it, since it introduces disclosure risk with no analytic benefit.

What is trauma-informed research design?

An approach that anticipates a population may have experienced trauma and designs data collection to minimize re-traumatization: giving participants control over what they disclose, avoiding unnecessary retelling of traumatic events, and having a referral pathway ready if a participant becomes distressed.

Who is the best access route for migrant or refugee research participants?

Community-based organizations, resettlement agencies, or established community leaders who already have trust within the population, rather than cold outreach, which is far less likely to succeed and can raise legitimate suspicion.

Do I need to write a positionality statement for this kind of research?

Yes, briefly — addressing your own citizenship, language background, and any professional relationship to the population, since these shape access, rapport and how findings should be interpreted, more so than in research with a general population.

How detailed does my data security plan need to be?

Detailed enough that a reviewer can judge whether it is proportionate to the actual sensitivity of what you are collecting — where data is stored, who has access, encryption, and retention period — rather than a generic institutional statement.

Design migrant and refugee research that is precise and trauma-informed

Tesify’s thesis workspace, used by 9,000+ students and 15,000+ chapters, helps social work researchers structure a precise population, a trauma-informed design and interpreter and access arrangements. Every word stays 100% written by you.