03 · What You Need to Know
A Duty to Help Can Come From Several Different Sources
Ethical responsibility and legal duty are not the same thing
Researchers often ask whether they have a “duty of care” as though this were one universal research-ethics rule. It is not.
A legal duty depends on jurisdiction, the relationship between the parties, professional status, institutional arrangements, applicable safeguarding law, and the circumstances. A clinician conducting research may have obligations that differ from those of a sociologist. Rules involving children may differ from those involving competent adults. Requirements also vary internationally.
Ethical responsibility can be broader than what the law compels. A researcher may reasonably conclude that some response is ethically warranted even where no specific legal reporting requirement applies.
Researchers should therefore avoid saying either “ethics requires me to report everything” or “there is no law requiring me to act, so I have no responsibility.” Both collapse distinct questions into one.
Research ethics recognizes that researchers can acquire responsibilities outside the study
UKRI's ESRC guidance addresses this directly. Researchers working with potentially vulnerable participants may find themselves facing increased responsibilities or expectations outside the scope of the project, including situations where an interview reveals that a participant is in significant danger and the researcher is obliged to act.
ESRC recommends assessing the likelihood of such situations beforehand and developing strategies and a framework of clear responsibilities. Its guidance for international research makes a similar point where researchers may encounter urgent needs for intervention outside the research objectives.
The possibility of intervention should therefore be designed into foreseeable higher-risk research rather than treated as an ethical surprise.
The seriousness and immediacy of harm matter
A researcher does not acquire the same responsibility every time a participant makes a decision the researcher considers unwise. Respect for autonomy includes allowing adults to make choices researchers might not make themselves.
The case for action becomes stronger as the threatened harm becomes more serious, credible, and immediate, particularly where the person at risk cannot reasonably protect themselves or where safeguarding responsibilities apply.
Ordinary difficulty or disagreement
Usually does not justify overriding a participant's autonomy simply because the researcher believes another choice would be better.
Credible risk of serious harm
May activate safeguarding, professional, legal, institutional, or research responsibilities that justify or require action.
No simple formula can calculate the threshold. Context, capacity, vulnerability, immediacy, severity, credibility, and the researcher's responsibilities all matter.
Vulnerability can strengthen responsibilities without removing autonomy
Research involving children, people with impaired decision-making capacity, people experiencing abuse or exploitation, or other potentially vulnerable groups may create additional safeguarding responsibilities.
But “vulnerable” should not become shorthand for “incapable of deciding.” ESRC guidance emphasizes that vulnerability can be complex and context dependent. Researchers should protect people from serious harm without unnecessarily removing agency from participants who remain capable of making their own decisions.
The appropriate response should therefore reflect the person's circumstances rather than relying solely on a category assigned to them.
Sometimes the research itself creates a stronger responsibility
Researchers may acquire responsibilities because the study places them in a particular relationship with participants.
A team may deliberately recruit people to discuss traumatic experiences, invite disclosure of sensitive information, enter private homes, ask participants to revisit distressing events, or position researchers as a regular point of contact. These choices can make certain needs or disclosures reasonably foreseeable.
ESRC's ethics-application guidance requires procedures for dealing with information arising during fieldwork that raises significant concerns about the safety or wellbeing of participants or others. This reflects a basic principle: if the research design predictably creates circumstances in which serious concerns may emerge, the team should decide beforehand what researchers will do with them.
Professional roles may continue alongside the researcher role
A researcher who is also a physician, psychologist, nurse, teacher, social worker, lawyer, or other professional may have obligations arising from that professional role.
Those obligations cannot be generalized across professions or jurisdictions. Nor should researchers assume that merely holding a qualification means every research interaction creates a professional relationship.
Potential dual-role responsibilities should be clarified before the study where possible, especially if researchers are likely to encounter situations within their professional domain.
Confidentiality can conflict with the duty to protect
One of the hardest cases occurs when helping someone appears to require disclosing information they provided confidentially.
ESRC guidance recognizes that significant danger to a participant or another identifiable person may require action. Where researchers believe confidentiality must be breached to avert serious harm, the disclosure should be defensible, appropriately limited, and made to parties empowered to act. Where possible, participants should be told what action will be taken unless informing them would increase risk.
This is precisely why researchers should not casually promise absolute confidentiality in studies where serious disclosures are foreseeable.
Applicable law and institutional policy must be checked because the conditions under which disclosure is permitted or required vary.
Helping does not necessarily mean personally intervening
The phrase “duty to help” can suggest that the researcher must personally solve the situation. Often that is neither required nor wise.
A proportionate response might involve contacting emergency services, activating a safeguarding procedure, consulting a designated institutional officer, providing an approved referral, obtaining specialist advice, or alerting another professional who has authority to respond.
Researchers should work within their competence. Someone conducting an interview does not automatically become qualified to provide emergency mental health treatment, legal representation, child-protection investigation, or physical intervention.
Direct intervention can create new harms
Intervention may protect someone, but it can also expose them to retaliation, stigma, unwanted institutional involvement, family separation, legal consequences, immigration problems, or loss of trust.
These possibilities do not mean researchers should remain passive in the face of serious danger. They mean the form of intervention matters.
International research makes this particularly clear. Authorities considered protective in one jurisdiction may be distrusted or dangerous in another. ESRC guidance recommends understanding political and cultural sensitivities and working with knowledgeable local organizations or researchers where appropriate.
The researcher should not become endangered while helping
A duty toward another person does not automatically require researchers to expose themselves to serious physical danger. If a researcher witnesses violence, direct intervention may worsen the situation or create another victim.
Depending on the circumstances, the appropriate action may be to reach safety and summon qualified assistance. Researcher protection remains part of ethical research, not an optional consideration that disappears whenever another person needs help.
Participants asking for help and researchers witnessing harm are related but different
A participant may explicitly ask the researcher for money, advice, transportation, professional services, advocacy, or another form of assistance. Many such requests involve help outside the researcher's agreed role without creating a duty to provide exactly what was requested.
By contrast, a researcher may witness a serious problem without anyone asking for help. The ethical question then concerns the significance of the observed risk and the responsibilities activated by it.
In both situations, the response should follow the nature of the risk rather than simply whether the participant used the word “help.”
Researchers should know the escalation route before the dilemma occurs
Ethics applications should identify procedures for serious disclosures and observed incidents where these are reasonably foreseeable. Research organizations should also maintain systems capable of supporting researchers when ethics issues emerge.
That means researchers should know whom to contact, what information can be disclosed, how urgent situations differ from non-urgent concerns, and what happens if the designated person is unavailable.
A beautifully worded ethics protocol is of limited assistance if the fieldworker discovers at 9:00 p.m. that the only named safeguarding contact works Tuesdays from 10:00 to 12:00.
Watch Out
Do not describe a particular intervention or reporting obligation as universal unless the applicable law, professional standard, or institutional policy actually makes it so. Ethical concern about serious harm may be widely shared, but the required response can differ substantially across settings.
