01 · The Question
How Do the Major Research Ethics Principles Work Together?
Respect for persons, beneficence, nonmaleficence, and justice are frequently presented as four principles of research ethics. That can make them look pleasantly tidy: respect people's choices, do good, avoid harm, and be fair.
Actual research is less cooperative.
Respecting a participant's choice may permit a risk that concern for their welfare makes uncomfortable. Protecting a population from risk may exclude that population from research that could benefit it. A scientifically valuable study may impose burdens on people who will never receive its benefits.
The principles therefore make more sense as interacting ethical considerations than as four independent requirements that can always be satisfied separately.
03 · What You Need to Know
What Each Principle Contributes to Ethical Research
First, a Terminological Caution: Belmont Has Three Principles, Not Four
It is common to encounter respect for persons, beneficence, nonmaleficence, and justice presented together. That framework is useful, but it should not be attributed verbatim to the Belmont Report.
Belmont identifies three basic principles for research involving human subjects: respect for persons, beneficence, and justice. Within beneficence, it gives two complementary rules: do not harm, and maximize possible benefits while minimizing possible harms. In other ethical frameworks, particularly the four-principles tradition associated with biomedical ethics, beneficence and nonmaleficence are distinguished.
There is therefore no need to choose between saying that research ethics has “three principles” or “four principles” without specifying the framework. The more precise approach is to identify which framework you are using and understand the ethical work each concept performs.
Respect for Persons Protects Agency and Recognizes When Agency Is Limited
Belmont describes respect for persons through two related commitments: treating individuals as autonomous agents and protecting people whose autonomy is diminished.
Autonomy means more than allowing someone to tick a consent box. It concerns a person's ability to deliberate about participation and make a sufficiently informed and voluntary decision. Researchers can undermine autonomy by withholding material information, making information unnecessarily difficult to understand, using inappropriate pressure, or creating circumstances in which refusal carries real or perceived consequences.
Respect also explains why identical treatment is not always appropriate. Someone with limited decision-making capacity may require additional safeguards. A student being recruited by the professor who grades them may face pressures that an unrelated community participant does not. A person may also be autonomous for one decision while requiring assistance with another.
The ethical question is therefore not merely, “Did this person agree?” It is, “Was this person's capacity for meaningful choice appropriately respected and protected in these circumstances?”
Beneficence Directs Attention Toward Welfare and Benefit
Belmont treats beneficence as an obligation, not simply as charitable behavior. Researchers should consider how their work can secure legitimate benefits while reducing risks.
Benefits may occur at different levels. A participant might receive a direct benefit, although many studies provide none. A community or patient population might benefit from improved knowledge. Society may gain information that supports future research, policy, or practice.
Potential benefit should nevertheless be described realistically. Researchers should not transform a remote possibility into a promised outcome merely because a stronger benefit makes the protocol easier to defend.
Beneficence therefore asks both what good the research might reasonably accomplish and whether the study has been designed to realize that value responsibly.
Nonmaleficence Makes Avoidable Harm a Separate Question
Separating nonmaleficence from beneficence can be analytically useful even when a framework such as Belmont discusses avoiding harm under beneficence.
A study can have considerable potential value and still contain an unnecessary risk. Researchers should not reason that a sufficiently valuable objective automatically cancels avoidable harm. Instead, they should ask whether the same objective could be achieved through a safer procedure, less sensitive data, fewer interventions, stronger confidentiality measures, or a different population.
Harm also extends beyond physical injury. Belmont explicitly recognizes psychological, physical, legal, social, and economic harms. Contemporary research can additionally raise substantial concerns about privacy, re-identification, stigma, discrimination, and consequences for groups or communities.
Nonmaleficence does not make all risk impermissible. It forces researchers to justify why remaining risk is necessary rather than merely convenient.
Justice Examines Who Carries the Research and Who Benefits From It
Justice changes the unit of ethical analysis. Respect and harm may initially draw attention toward what happens to an individual participant. Justice asks researchers to look across participants and populations.
Who is being recruited? Why them? Who bears the inconvenience, risk, or opportunity cost? Who is excluded? Who is expected to benefit from the resulting knowledge?
Belmont connects justice to fair procedures and outcomes in participant selection. Its historical discussion warns against repeatedly placing research burdens on disadvantaged or easily available populations while benefits flow elsewhere.
Justice does not require every population to bear identical burdens or receive identical benefits. Differences can be justified by the research question, biological or social characteristics relevant to the study, need, or other defensible considerations. What requires scrutiny is whether the distribution has an ethically relevant justification.
The Four Principles Often Reinforce One Another
Many research decisions support several principles simultaneously.
Removing unnecessary identifiers from a dataset may reduce harm while respecting privacy. Designing recruitment so that a professor does not know which students declined may strengthen voluntariness and reduce possible academic pressure. Ensuring that an intervention is studied in a population likely to use it can advance both scientific relevance and justice.
Ethical design is therefore not normally a matter of making one decision for respect, another for beneficence, and another for justice. A well-designed safeguard can respond to several ethical concerns at once.
Sometimes the Principles Pull in Different Directions
The difficult cases are those in which satisfying one ethical consideration appears to weaken another.
| Ethical tension |
One consideration |
Competing consideration |
| An adult voluntarily accepts substantial research risk |
Respect supports taking the person's autonomous choice seriously. |
Nonmaleficence and beneficence still require scrutiny of whether the risk is ethically acceptable. |
| A population is considered especially vulnerable |
Protection from harm may support additional safeguards or exclusion. |
Justice may weigh against routine exclusion if it leaves the population without evidence relevant to its needs. |
| Sensitive information could produce valuable knowledge |
Beneficence may support collecting information needed to answer an important question. |
Respect and nonmaleficence require attention to privacy, consent, and consequences of disclosure. |
| A highly efficient sample is drawn from a dependent population |
Scientific efficiency may make recruitment attractive. |
Respect and justice may raise concerns about voluntariness and unfair concentration of research burdens. |
A conflict does not mean one principle has failed. It means the principles are doing what ethical principles are supposed to do: revealing considerations that a simple procedural rule might miss.
Respect for Autonomy Does Not Give Participants Unlimited Power to Authorize Risk
A particularly important tension arises between respect and protection from harm.
Suppose an informed adult freely agrees to participate in an extremely hazardous study. Does voluntary consent settle the matter?
No. Consent is ethically important, but it does not automatically convert an otherwise unjustifiable risk into an acceptable one. Ethical review still asks whether risks have been minimized and whether remaining risks are reasonable in relation to anticipated benefits and the importance of the knowledge expected to result.
Researchers therefore cannot outsource every ethical judgment to participants. Respecting autonomy matters enormously, but asking someone to choose does not remove the researcher's responsibility for what choices are ethically appropriate to offer.
Protection Can Become Unjust When It Turns Into Automatic Exclusion
The opposite problem is excessive protection.
Researchers may believe that excluding populations perceived as vulnerable is the safest ethical response. Sometimes exclusion is justified. But systematic exclusion can also create evidence gaps and deny groups the potential benefits of research relevant to them.
The 2024 Declaration of Helsinki explicitly recognizes that vulnerability can be contextual and dynamic and requires fair and responsible inclusion with appropriately considered support and protections. It also asks researchers to consider how research benefits, risks, and burdens are distributed.
The ethical question is therefore not simply whether a group is vulnerable. Researchers should ask what creates the vulnerability, what additional protection is needed, whether inclusion is justified, and what harms could result from exclusion.
No Universal Formula Tells You How Much Weight Each Principle Gets
Ethical reasoning would be considerably easier if researchers could assign numerical weights to autonomy, benefit, harm, and justice and let a spreadsheet settle the matter. Ethics committees would probably enjoy shorter meetings too.
Major research ethics frameworks do not provide such an algorithm. The relevance and weight of competing considerations depend on the circumstances, including the magnitude and probability of harm, importance of the research, participant population, available alternatives, and applicable ethical and regulatory standards.
This is why understanding the core principles of research ethics is only the beginning. Difficult cases require researchers to explain why a particular balance is justified.
When principles genuinely point toward different decisions, the problem becomes one of reasoning through competing ethical principles, not searching for a hidden rule that makes the conflict disappear.
