03 · What You Need to Know
The Major Types of Risk Researchers Should Consider
Physical Risk
Physical risk involves possible bodily harm, pain, injury, illness, physiological effects, or other adverse effects on physical health. It is often most obvious in biomedical and clinical research, but it can arise elsewhere.
Examples include adverse reactions to drugs, complications from invasive procedures, infection, radiation exposure, falls during movement studies, fatigue during exercise protocols, or injury caused by research equipment.
Physical risk should not be assessed merely by listing possible adverse events. Researchers need to consider their probability, potential severity, reversibility, duration, available treatment, and whether particular participants face greater risk because of their health or circumstances.
Psychological Risk
Psychological risk concerns possible emotional or mental harm associated with participation. Interviews, surveys, experiments, observations, or interventions can evoke distress even when nothing physically invasive occurs.
Research involving trauma, grief, abuse, discrimination, sexuality, stigmatized experiences, family conflict, violence, or other sensitive experiences may evoke anxiety, shame, fear, sadness, anger, intrusive memories, or other distress. Experimental manipulation can also create frustration, embarrassment, stress, or changes in self-perception.
Researchers should distinguish ordinary transient emotional responses from more consequential possible harms. Not every uncomfortable question constitutes serious psychological risk, but describing all emotional effects as “minor discomfort” without examining the participant population and research context can underestimate what participants may experience.
Social Risk
Social harm can occur when participation or research information changes how other people treat a participant. Possible consequences include damaged personal relationships, stigma, discrimination, exclusion, harassment, retaliation, or altered standing within a family, workplace, school, profession, or community.
These harms frequently arise through disclosure. For example, revealing that someone participated in research about domestic violence, substance use, workplace misconduct, political activity, or a stigmatized condition may itself communicate information the participant wanted to keep private.
Social risk can also exist at a group level. Research findings about identifiable communities may contribute to stereotyping or stigmatization even when individual participants cannot be identified. Ethical analysis may therefore need to consider effects beyond the individual participant.
Legal Risk
Legal risk exists when research participation or disclosure of research information could expose a person to investigation, prosecution, civil liability, regulatory consequences, immigration consequences, or other adverse legal action.
This may arise when researchers collect information about illegal behavior, undocumented activity, prohibited drug use, regulatory violations, or conduct that could become evidence in legal proceedings.
The precise legal exposure depends heavily on jurisdiction. Researchers should therefore avoid generic promises such as “your information can never be disclosed” unless the applicable legal and institutional protections genuinely support that claim. Confidentiality procedures, ethics approval, and consent language do not automatically override legal obligations or lawful demands for information.
Watch Out
Do not promise absolute confidentiality simply because your research team intends to keep information confidential. The limits of confidentiality may depend on applicable law, institutional policy, mandatory reporting requirements, data location, and the protections legally available in the relevant jurisdiction.
Economic Risk
Economic harm involves possible financial loss or adverse effects on employment, income, insurance, benefits, property, or other economic interests.
A participant could lose wages because research requires repeated visits. More consequentially, disclosure of sensitive information could contribute to loss of employment, denial of an opportunity, financial penalties, or other economic consequences.
Researchers should therefore distinguish between the cost of participation, such as transportation expenses or time away from work, and the risk of economic harm, such as the possibility that disclosure could result in job loss. The former may be a burden or cost that is already occurring; the latter concerns a possible harmful outcome.
Reputational Risk
Reputational harm occurs when research participation or disclosed information could damage how a person, organization, profession, or community is perceived by others.
Consider a qualitative study involving senior managers discussing questionable organizational practices. Even if publication removes their names, detailed combinations of job title, department, event, location, and quotation might allow colleagues to identify a speaker. The consequence may not be prosecution or direct financial loss. It could instead be damaged professional credibility or standing.
Reputational risk deserves particular attention in small populations, elite interviews, organizational case studies, ethnography, and research involving distinctive roles or events. Removing a name does not necessarily make a participant unidentifiable.
Privacy and Confidentiality Can Be Pathways to Several Kinds of Harm
Privacy and confidentiality are sometimes treated as another item in a list of risk categories. In practice, they often operate as pathways through which other harms occur.
A confidentiality breach may lead to psychological distress, social stigma, economic loss, legal consequences, or reputational damage. The breach is therefore not necessarily the endpoint of the risk analysis. Researchers should ask what could happen after information is exposed.
Research information is disclosed
A participant's sensitive responses become identifiable.
Another person gains access
An employer, relative, authority, colleague, or member of the public learns the information.
A consequence follows
The participant experiences stigma, conflict, retaliation, legal exposure, financial loss, or reputational damage.
Thinking in pathways prevents risk assessments from stopping at vague statements such as “there is a risk of loss of confidentiality.” The ethically important question is what that loss could plausibly do to the participant.
The Same Event Can Produce Several Harms
Risk categories are analytical tools, not mutually exclusive boxes. One event may create several consequences.
Suppose confidential information about an employee's report of workplace misconduct is accidentally disclosed. The participant may experience anxiety, strained relationships with colleagues, retaliation affecting employment, and damage to professional standing. The initiating event is one disclosure, but the possible harms span several domains.
This is one reason a study should not be classified simply as “psychological risk” or “social risk” and considered fully assessed. The categories help researchers search systematically for consequences that might otherwise be missed.
The Same Procedure Does Not Create the Same Risk for Everyone
Risk is contextual. Asking participants about political activity may have few consequences in one setting but substantial social or legal implications in another. Disclosure of a health condition may have different consequences depending on stigma, employment conditions, family relationships, access to insurance, or local law.
The participant population therefore matters when determining what counts as research risk and whether that risk is acceptable. Researchers should not estimate risk solely from the procedure while ignoring the circumstances of the people exposed to it.
Probability and Severity Still Matter Within Every Category
Finding a possible harm does not establish that a study is unacceptably risky. The Belmont Report distinguishes the probability of harm from its magnitude. Both require attention.
A mild but relatively common temporary reaction differs from an extremely uncommon but irreversible consequence. Researchers should be particularly careful when evaluating harms that are unlikely but potentially severe.
The goal is not to produce the longest possible catalogue of everything imaginable. Risk identification should be plausible, evidence-informed, and specific enough to support meaningful safeguards.
