01 · The Question
What Happens When Important Research Cannot Be Conducted Without Some Risk?
Research into domestic violence, political persecution, infectious disease, workplace exploitation, or educational discrimination may involve people whose circumstances already expose them to harm. Asking questions, collecting information, or documenting experiences can introduce additional risks, even when the research is carefully designed.
One response is to avoid these populations or topics altogether. That may eliminate the risks associated with conducting the proposed study, but it also leaves important questions unanswered.
When researchers refuse to investigate difficult problems because some possibility of harm remains, the decision itself may have ethical consequences. The challenge is determining which risks should prevent research and which can be responsibly managed.
03 · What You Need to Know
Why Responsible Research Requires Risk Assessment Rather Than Absolute Risk Avoidance
Why Is Zero-Risk Research Usually an Unrealistic Standard?
Research involving people rarely permits researchers to guarantee that nothing adverse will happen. Even apparently straightforward surveys can create inconvenience, fatigue, privacy concerns, or discomfort. More sensitive studies may involve risks of stigma, retaliation, psychological distress, or unintended disclosure.
The presence of a possible adverse outcome does not, by itself, establish that the research is unethical. Ethical evaluation must consider what could happen, how likely it is, how serious the consequences might be, and what measures could reduce those risks.
The Belmont Report frames beneficence through obligations to avoid harm and maximize possible benefits while minimizing possible harms. It does not establish an unconditional requirement that all research risks disappear.
The World Medical Association's Declaration of Helsinki similarly requires careful assessment of foreseeable risks and burdens in relation to potential benefits, with appropriate risk-minimization measures.
These principles are particularly relevant when excessive caution creates ethical problems of its own.
What Makes Research Socially Valuable?
Social value concerns the potential of research to generate knowledge that may improve understanding, inform decisions, address important problems, or contribute to human welfare.
A study need not promise immediate policy change or direct benefits to participants to possess social value. Foundational research, methodological investigations, and studies of poorly understood experiences may contribute knowledge that becomes useful later.
However, researchers should distinguish plausible social value from speculative claims of importance. A study is not automatically valuable because its topic is serious or because investigators describe its findings as potentially transformative.
Social value depends partly on whether the research question matters, whether the proposed design can produce credible evidence, and whether the findings could reasonably contribute to knowledge or practice.
The CIOMS International Ethical Guidelines emphasize the social and scientific value of health-related research. They also recognize that scientific validity matters because exposing participants to risk for research incapable of producing useful knowledge is ethically difficult to justify.
Why Can Avoiding Research Create Ethical Costs?
When potentially valuable studies are abandoned solely because some manageable risk remains, important populations or problems may remain insufficiently understood.
For example, avoiding interviews with people who experience discrimination may reduce opportunities to document institutional practices that are otherwise difficult to observe. Excluding people with complex health conditions from relevant research can also limit the applicability of findings to those populations.
These are potential consequences rather than proof that any particular study must proceed. Researchers should not assume that every knowledge gap justifies exposing participants to risk.
Nevertheless, a decision not to conduct research should consider what information will remain unavailable and whether that absence could perpetuate inequities or poorly informed decisions.
This creates a distinction between the ethics of conducting a study and the ethics of leaving a consequential question unanswered.
How Are Research Risks Evaluated?
Risk assessment generally considers the probability and magnitude of possible harms. Depending on the study, relevant harms may be physical, psychological, social, economic, legal, or informational.
Researchers should examine both the inherent risks associated with the research question and the additional risks introduced by particular procedures.
| Risk dimension |
Question to examine |
Possible response |
| Likelihood |
How plausible is the adverse event under the proposed conditions? |
Examine relevant evidence and identify exposure pathways. |
| Severity |
How serious would the consequences be if the event occurred? |
Prioritize prevention of severe or irreversible harm. |
| Duration |
Would the harm be temporary, prolonged, or potentially permanent? |
Consider follow-up procedures and longer-term consequences. |
| Distribution |
Who bears the risks, and who might benefit from the knowledge? |
Examine fairness and possible exploitation. |
| Mitigation |
Which safeguards could reduce the likelihood or seriousness of harm? |
Modify recruitment, consent, data collection, and security procedures. |
| Residual risk |
What risks remain after reasonable safeguards are implemented? |
Determine whether proceeding remains ethically defensible. |
Risk assessments may be qualitative or use structured scoring systems. A numerical score should not be mistaken for an objective measure of ethical acceptability. Serious consequences, uncertainty, and contextual factors may not be adequately represented by a simple multiplication of likelihood and severity.
What Is the Difference Between Minimal Risk and Acceptable Risk?
Minimal risk is a defined regulatory concept in some jurisdictions. Under the United States Common Rule, for example, minimal risk concerns the probability and magnitude of harm or discomfort anticipated in research relative to those ordinarily encountered in daily life or during routine physical or psychological examinations or tests.
Other jurisdictions and research frameworks may use different definitions or procedures. Researchers should consult the standard applicable to their study rather than assume that a single international definition governs all disciplines.
Acceptable risk is a broader ethical judgment. A study may involve more than minimal risk and still be ethically permissible under an applicable framework, provided the relevant requirements are satisfied.
Minimal risk
A classification defined by the applicable regulatory or ethical framework, often used to determine review and protection requirements.
Ethically acceptable risk
A contextual judgment about whether the remaining risks are justified and adequately managed under the applicable requirements.
Neither concept means that harm is impossible. Equally, a minimal-risk classification does not remove the obligation to protect confidentiality, obtain appropriate consent, or respond to emerging problems.
Does Greater Social Value Justify Greater Participant Risk?
Potential social value is relevant to risk-benefit assessment, but the relationship is not an unrestricted exchange. Researchers cannot simply claim that an important scientific objective outweighs any degree of participant harm.
The Declaration of Helsinki requires the importance of the research objective to outweigh the risks and burdens to participants and places participant rights and interests at the center of ethical obligations.
Consent, risk limits, independent review, and protections for participants in situations of vulnerability remain important even when a study addresses an urgent public concern.
Researchers should also distinguish benefits to participants from benefits to society. A study may offer no direct individual benefit while contributing valuable knowledge. That distinction must be communicated accurately during consent and considered during ethical review.
The difficult task of balancing scientific value against participant burden requires attention to who bears the costs of producing knowledge.
Why Risk Minimization Should Come Before Risk Acceptance
Recognizing that some risk may be acceptable does not permit researchers to retain avoidable hazards for convenience.
Before deciding whether residual risk is justified, investigators should ask whether the same research objective could be achieved through safer recruitment, less intrusive measurements, alternative data sources, shorter procedures, or stronger privacy protections.
For example, a study of workplace discrimination might use confidential interviews conducted outside working hours rather than identifiable surveys distributed through supervisors. Both approaches could address similar questions, but their implications for participant safety may differ substantially.
Risk minimization should be meaningful rather than ceremonial. A safeguard is useful only insofar as it addresses a credible source of harm and can be implemented reliably.
When Should Researchers Decide Not to Proceed?
Some research should not proceed in its proposed form, even when the question is important.
This may be the case when foreseeable risks are disproportionate to plausible benefits and social value, when valid consent cannot be obtained under the applicable requirements, when serious harms cannot be adequately mitigated, or when the study lacks sufficient scientific validity.
Researchers may also need to stop or redesign a study when circumstances change after approval. A protocol considered acceptable at the outset may become inappropriate because of new evidence, political conditions, security threats, or unanticipated adverse events.
Watch Out
Social importance must never become a blanket justification for exposing disadvantaged or dependent populations to risks that investigators would not adequately justify for others. The distribution of research burdens matters as much as the overall promise of the findings.
When risks cannot be ethically controlled, methodological rigor must not be used to justify an unacceptable procedure. Researchers may need to change the design, narrow the question, or discontinue the project.
04 · A Practical Example
Should Researchers Study Educational Discrimination When Disclosure Could Harm Students?
Hypothetical Example
Investigating discriminatory practices in higher education
A research team wants to investigate whether students from economically disadvantaged backgrounds experience discriminatory treatment during academic advising.
Interviews could reveal experiences that are not captured in institutional records. However, participants fear that their comments might be traced back to them, potentially affecting relationships with faculty or access to institutional opportunities.
The researchers initially consider abandoning the project because confidentiality cannot be guaranteed absolutely.
Establish the value of the question.
The team reviews existing evidence and determines whether the study could address a meaningful knowledge gap rather than merely repeat established findings.
Identify the actual exposure pathways.
The researchers examine recruitment records, interview locations, identifying details, data access, and the possibility that quotations could reveal participants indirectly.
Redesign procedures to reduce risk.
Recruitment is separated from academic advising, interviews are conducted privately, identifying details are minimized, and access to sensitive records is restricted.
Explain remaining limitations.
Consent materials clarify that confidentiality protections reduce risk but cannot guarantee that re-identification is impossible.
Obtain independent ethical assessment.
The revised protocol is submitted for the review required by the institution, including an assessment of residual risks and whether the proposed safeguards are adequate.
Suppose the reviewers determine that the revised procedures provide adequate protections and that the remaining risks are justified under the applicable framework. The study may then proceed, subject to its approval conditions.
If the institution is so small that participants could readily be identified despite the safeguards, the decision may be different. The team might need to recruit across several institutions, modify the questions, use a different method, or postpone the research.
The example illustrates why the ethical choice is not necessarily between conducting a dangerous study and conducting no study at all. Research design can change the level and distribution of risk.
06 · What This Means for You
How to Decide Whether Research With Unavoidable Risk Should Proceed
Begin with the research question rather than a preferred method. Establish what knowledge the study could plausibly contribute and why obtaining that knowledge matters.
Then identify the risks introduced by the proposed procedures. Distinguish risks that can be eliminated through redesign from those that remain inherent in addressing the question.
Where feasible, consult prospective participants or relevant communities about burdens and safeguards. Their perspectives may improve the assessment, although they do not replace independent review or establish that every proposed risk is acceptable.
A simple decision framework
If the study has limited scientific or social value
Reconsider the research question or design before exposing participants to avoidable burdens.
If risks are foreseeable and substantially reducible
Implement appropriate safeguards and reassess the remaining risks.
If meaningful risks remain after safeguards
Evaluate their likelihood, severity, distribution, and justification under the applicable ethical framework.
If risks remain unacceptable despite redesign
Do not proceed in the proposed form. Consider an alternative question, population, method, or timing.
When the available options involve competing values, reasoned ethical disagreement should be addressed through transparent justification rather than by assuming that the most cautious position is automatically correct.
Keep a written record of the assessment, safeguards, unresolved uncertainties, and review decisions. A defensible decision should explain both why the research matters and why its remaining risks are acceptable.
