03 · What You Need to Know
How to Reason Through Competing Ethical Principles
Ethical Principles Are Guides to Judgment, Not an Automatic Decision Algorithm
The Belmont Report identifies respect for persons, beneficence, and justice as broad principles intended to provide an analytical framework for ethical problems in research. It explicitly acknowledges that rules can conflict and that the principles themselves cannot always resolve particular ethical problems beyond dispute.
This limitation is not a defect in the framework. Ethical principles operate at a level broad enough to apply across many situations. That generality means researchers still need judgment when principles interact.
If research ethics could be reduced to “Principle A always beats Principle B,” ethics committees would require considerably fewer meetings.
First Determine Whether There Is a Genuine Conflict
Not every difficult decision contains a true conflict between principles.
A researcher might claim that participant privacy conflicts with scientific value when the real problem is simply that collecting more data would be convenient. A researcher might describe informed consent as conflicting with feasibility when the study could be redesigned without compromising either.
Before balancing anything, state exactly what each option protects and what it threatens.
For example:
Option A protects confidentiality but may prevent researchers from responding to a serious disclosed risk. Option B permits disclosure under specified circumstances but limits the confidentiality participants can be promised.
That is more useful than saying vaguely that “ethics is complicated.”
Get the Facts Right Before Weighing the Principles
Ethical disagreement can sometimes be a factual disagreement in disguise.
How likely is the harm? How serious would it be? Is the participant actually capable of making the decision? Is the information genuinely identifiable? Is the proposed procedure scientifically necessary? Does a legal reporting obligation apply? Is there really no less risky alternative?
Different answers can change the ethical analysis substantially.
Belmont's discussion of risk-benefit assessment emphasizes systematic, nonarbitrary analysis, careful assessment of information, and consideration of alternatives.
Ethical balancing based on inaccurate assumptions is simply precise reasoning about the wrong problem.
Identify Every Ethically Relevant Interest, Not Just the Two Most Obvious Ones
Dilemmas are often framed too narrowly.
A researcher may see a conflict between autonomy and beneficence while overlooking justice. Or between confidentiality and protection from harm while overlooking the effect that a broad disclosure policy could have on future participants' willingness to speak honestly.
Relevant considerations may include autonomy, welfare, avoidance of harm, fairness, privacy, confidentiality, scientific and social value, community interests, vulnerability, researcher obligations, institutional responsibilities, and applicable law.
The aim is not to produce the longest possible list. It is to avoid solving a two-variable version of a problem that actually has four.
Do Not Assume the Principles Have a Fixed Hierarchy
The Belmont Commission did not establish a hard ranking in which one of its three principles always prevails. In retrospective interviews, people involved in the Commission's work explained that the principles can conflict and that which consideration should give way depends on the facts and what is at stake.
This does not mean every principle always has equal weight.
In a particular case, one consideration may be substantially stronger because the threatened harm is severe, the infringement on autonomy is minor, the injustice is substantial, or the available alternative preserves most of the competing value.
The absence of a universal hierarchy creates a need for justification, not ethical paralysis.
Ask Whether the Conflict Can Be Reduced Through Redesign
The most useful ethical decision is often not choosing between Option A and Option B. It is designing Option C.
Suppose researchers want to study a vulnerable population because excluding it would create an important evidence gap, but inclusion raises concerns about consent and risk. The choice need not be unrestricted inclusion versus complete exclusion. Additional safeguards, modified recruitment, enhanced consent procedures, independent advocates, reduced-risk procedures, or narrower eligibility may preserve some of the scientific and justice-based reasons for inclusion while strengthening protection.
Ethical creativity matters because many apparent dilemmas are partly design problems.
Consider the Least Ethically Costly Way to Achieve the Legitimate Objective
When several options can accomplish the research purpose, compare their ethical costs.
If two methods can answer the question adequately but one requires substantially less sensitive information, the privacy conflict may largely disappear. If recruitment can be conducted by someone without authority over potential participants, the tension between scientific access and voluntariness may be reduced.
This does not mean always choosing the least burdensome procedure regardless of scientific quality. The relevant comparison is among options capable of achieving the legitimate research objective adequately.
Distinguish Infringing a Principle From Completely Disregarding It
Sometimes an ethically defensible decision cannot fully satisfy every principle.
Limited incomplete disclosure may restrict a participant's access to information but be necessary for a valid low-risk behavioral study under conditions that protect participants and provide appropriate debriefing. Belmont itself discusses circumstances under which incomplete disclosure may be justified when it is truly necessary, undisclosed risks are not more than minimal, and an adequate debriefing plan exists where appropriate.
The ethical argument is not that respect no longer matters. It is that a limited infringement may be justified under carefully constrained circumstances because of competing considerations.
Severity, Probability, Reversibility, and Scope Can Affect Ethical Weight
Not all competing interests have the same magnitude.
A small temporary inconvenience should not automatically outweigh prevention of severe irreversible harm. Conversely, a highly speculative benefit should not automatically justify a substantial and likely infringement of participant rights.
Useful questions include:
- How serious is the potential harm or infringement?
- How likely is it?
- How many people could be affected?
- Can the effect be reversed or remedied?
- How certain is the anticipated benefit?
- Is the burden necessary?
- What alternatives exist?
These questions do not produce a numerical ethics score. They make the comparison more disciplined.
Respect for Autonomy Can Conflict With Protection From Harm
One classic conflict concerns an informed adult who voluntarily accepts a substantial research risk.
Respect for persons supports taking the person's autonomous decision seriously. Beneficence and nonmaleficence require researchers to consider whether the risk itself is justified.
Consent therefore matters greatly without automatically authorizing every risk researchers could offer.
Belmont itself presents respect for persons as capable of generating competing claims. Its example of prisoner participation shows that respect can support allowing people to volunteer while simultaneously supporting protection against subtle coercion.
Protection From Harm Can Conflict With Justice
Researchers may want to exclude a population because participation is more complicated or risky. Protection may support that decision.
But systematic exclusion can produce evidence that does not apply to the population, deny opportunities to participate, and perpetuate knowledge gaps. Justice can therefore support appropriately protected inclusion.
The ethical task is to determine whether safeguards can make inclusion defensible rather than treating either inclusion or exclusion as automatically more ethical.
Confidentiality Can Conflict With Responsibilities Triggered by Serious Disclosures
Researchers may promise confidentiality to support privacy, trust, and honest participation. Yet some research settings can produce disclosures that create legal or ethical responsibilities to act.
The exact obligations vary by jurisdiction, profession, population, and research context, so researchers should not invent universal disclosure rules.
Where limits genuinely exist, the best ethical response often begins before data collection: identify foreseeable limits, describe them accurately during consent where appropriate, and establish a clear response procedure.
A dilemma handled prospectively is usually less chaotic than one discovered halfway through an interview.
Individual Interests Can Conflict With Social or Scientific Value
Research often seeks knowledge that may benefit people other than the participants who bear the immediate burden.
Beneficence can therefore operate at several levels. Belmont recognizes both individual and societal benefits from research while emphasizing that participant risks require careful consideration.
The possibility of substantial social benefit can strengthen the justification for some burdens. It does not create unlimited permission to sacrifice individual participants for aggregate benefit.
This is why the scientific or social value of research must be considered together with rights, risk, voluntariness, and fairness.
Legal and Institutional Requirements Can Constrain the Available Ethical Options
Researchers do not reason in a regulatory vacuum.
A researcher may conclude that two options are ethically defensible but discover that only one is legally or institutionally permitted. Conversely, a legally permitted action may remain ethically unattractive.
Ethical balancing therefore needs to occur within the set of actions researchers are actually authorized to take, unless a legitimate process exists for changing the rule, seeking an exception, or obtaining further review.
A researcher cannot resolve an ethical conflict simply by choosing an option that applicable law prohibits.
Consultation Is Part of Good Ethical Reasoning, Not Evidence of Failure
Difficult ethical problems often benefit from perspectives beyond the research team.
Relevant consultation may involve the ethics committee, institutional ethics staff, supervisors, community representatives, methodological experts, data-protection specialists, clinicians, legal advisers, or others depending on the issue.
Consultation is particularly useful when researchers themselves have a conflict of interest or when the decision affects a community whose perspective the research team does not adequately understand.
The researcher still needs to understand the reasoning. Consultation should improve judgment, not replace it with a game of “find someone willing to say yes.”
Document Why the Decision Was Made
Hard ethical decisions should be explainable after the fact.
Documentation can identify the competing principles, relevant evidence, alternatives considered, advice obtained, reasons alternatives were rejected, safeguards adopted, and why the final decision was judged proportionate.
This improves accountability and makes later reassessment possible if circumstances change.
It also prevents hindsight from rewriting a difficult decision as though the answer had always been obvious.
Some Ethical Conflicts Cannot Be Completely Resolved
A defensible decision can still carry an ethical cost.
Protecting confidentiality may limit what researchers can report. Including a population may expose participants to some unavoidable burden. Excluding them may perpetuate an evidence gap. A necessary deception may temporarily restrict autonomy despite careful debriefing.
Ethical reasoning does not always transform a dilemma into a situation where every value is perfectly preserved.
Sometimes the most intellectually honest conclusion is: this option is justified, but something ethically important is still being sacrificed.
And Sometimes No Available Option Is Clearly Good
There are cases in which every feasible option carries a serious ethical disadvantage.
At that point, the question is no longer simply how to balance principles. It becomes what to do when there is no clearly ethical option.
That distinction matters. A difficult trade-off with a defensible preferred option is not identical to a genuine moral remainder in which every available choice leaves substantial ethical concern.
