03 · What You Need to Know
How to Reach a Defensible Ethical Decision Without Assuming Universal Agreement
Why Can Reasonable Researchers Disagree About Ethics?
Ethical disagreement does not always arise from ignorance, poor judgment, or conflicting personal interests. Researchers may understand the same situation while placing different emphasis on autonomy, protection from harm, fairness, scientific value, or obligations to affected communities.
For example, one researcher may believe that excluding people with certain vulnerabilities protects them from disproportionate risk. Another may argue that automatic exclusion denies those individuals the opportunity to participate in research relevant to their lives.
Both positions identify legitimate ethical concerns. Their disagreement concerns how those concerns should be weighed in a particular situation.
Ethical principles often guide judgment without determining a unique answer. Respect for persons, beneficence, and justice, as articulated in the Belmont Report, provide important foundations for human-participant research, but applying them may require contextual interpretation.
The difficulty becomes greater when evidence about likely benefits or harms is incomplete. Researchers must sometimes decide before all consequences can be known.
Distinguish Ethical Disagreement From Disagreement About Facts
Before attempting to resolve an ethical conflict, determine whether the participants in the discussion actually disagree about values or about the circumstances of the research.
Consider a proposal to collect students' interactions with an AI tutoring system. One researcher argues that the data are anonymous. Another believes that the conversations contain information capable of identifying individual students.
This may initially appear to be a disagreement about privacy ethics. Yet an important part of the dispute concerns a factual and technical question: whether individuals can reasonably be identified from the records.
That question should be investigated using the actual data structure, system configuration, and relevant technical expertise.
Empirical disagreement
Different judgments about what is happening, what evidence establishes, or how likely particular consequences are.
Normative disagreement
Different judgments about what ought to be done, which interests deserve priority, or what level of risk is ethically acceptable.
Many real disagreements involve both. Resolving an empirical uncertainty may narrow the ethical dispute without eliminating it.
For instance, establishing that a dataset contains identifiable information clarifies the privacy risk, but researchers may still disagree about whether the proposed safeguards are sufficient.
Identify Binding Requirements Before Weighing Ethical Preferences
Ethical deliberation should not be used to bypass requirements that apply to the research.
A researcher may believe that a particular procedure offers substantial scientific value. That belief does not authorize ignoring a binding consent requirement, violating an applicable privacy law, or proceeding without a required ethics approval.
Begin by identifying relevant laws, institutional policies, regulatory requirements, approved protocols, contractual obligations, and professional standards.
Distinguish requirements that are mandatory from recommendations that permit contextual judgment. Where the scope of a requirement is uncertain, seek clarification from the responsible authority.
Some disagreements can be resolved at this stage because one proposed course of action is not permitted. Others remain because the governing framework allows more than one acceptable approach.
When applicable standards appear to conflict, the question of how to interpret competing research ethics requirements should be addressed before selecting a preferred option.
What Ethical Principles Are Actually in Conflict?
Ethical deliberation becomes more useful when researchers identify the particular interests and obligations involved rather than describe the problem simply as a disagreement about what is right.
Consider a study involving adults who have experienced a natural disaster. Interviews may provide valuable evidence about recovery needs, but discussing traumatic events could cause distress.
Respect for autonomy supports allowing informed adults to decide whether to participate. Beneficence requires attention to foreseeable harms and appropriate safeguards. Justice raises questions about whether excluding people who experienced severe harm would systematically remove their perspectives from the research.
These considerations need not be treated as mutually exclusive. Researchers might modify interview procedures, allow participants to skip questions, provide clear withdrawal arrangements, and establish appropriate responses to distress.
The resulting design may accommodate several ethical interests without fully eliminating the underlying tension.
Where principles genuinely conflict, the decision should explain why particular considerations receive greater weight in the circumstances.
How Should Researchers Evaluate Competing Ethical Arguments?
A defensible argument should connect ethical principles to the actual research circumstances.
For example, stating that participant autonomy must always prevail is insufficient when the proposed activity could create serious harm to people who have not consented. Likewise, asserting that all potential distress must be prevented may overlook participants' capacity to make informed choices and the social value of their participation.
Researchers should examine the factual assumptions, likely consequences, affected interests, and practical implications of each position.
Questions worth considering include whether the proposed activity is necessary to answer the research question, whether its risks are proportionate to its expected value, and whether an alternative could preserve the scientific purpose while reducing harm.
The strength of an ethical argument is not determined by how confidently it is expressed. It depends on whether its reasoning is coherent, evidence-informed, responsive to relevant objections, and consistent with applicable obligations.
Why Is Proportionality Important in Ethical Decisions?
Proportionality concerns whether the burdens or restrictions imposed by a decision are justified in relation to the interests the decision seeks to protect.
For example, excluding all students with disabilities from an educational technology study might reduce some anticipated accessibility-related difficulties. It could also create unjustified exclusion and undermine the relevance of the research to those students.
A more proportionate approach might involve accessible instruments, appropriate accommodations, and individualized assessment of risks.
Proportionality does not mean that any risk becomes acceptable when the expected scientific benefit is sufficiently large. Some actions remain impermissible under binding requirements or because they violate fundamental protections.
It also does not require assigning numerical values to every ethical consideration. In many situations, a qualitative comparison of likely consequences, available safeguards, and reasonable alternatives is more appropriate.
The broader problem of whether excessive caution can create ethical problems illustrates why risk reduction should be considered alongside fairness and scientific value.
Does Participant Autonomy Resolve Ethical Disagreements?
Participant autonomy is central to many ethical frameworks, particularly those governing research involving humans. However, autonomy is neither unlimited nor always sufficient to justify a research activity.
Meaningful autonomy requires adequate information, freedom from inappropriate pressure, and a genuine opportunity to make decisions.
In educational research, students may feel obliged to participate when recruitment is conducted by an instructor who also evaluates their academic performance. A signed consent form does not necessarily eliminate this concern.
Researchers should also consider whether participants' decisions could affect third parties who have not agreed to the research.
Conversely, protecting participants should not automatically become a reason to disregard their preferences or exclude them from meaningful opportunities.
The relationship between participant autonomy and the duty to prevent harm requires attention to the actual conditions under which choices are made.
Whose Perspectives Should Be Included in the Deliberation?
Ethical decisions may benefit from perspectives beyond those of the immediate research team.
Participants, community representatives, methodological experts, research ethics committees, and specialists in data protection or information security may identify concerns that researchers have overlooked.
For example, researchers studying automated educational assessment may focus on measurement accuracy, while students may be more concerned about the consequences of inaccurate classifications for their academic opportunities.
Including affected perspectives can improve understanding of the research context and reveal assumptions embedded in the proposed design.
However, consultation should not be reduced to a vote. A majority preference does not automatically override the rights of a minority or establish that an ethically consequential risk is acceptable.
Researchers should consider who is represented, whose interests may be missing, and whether participants in the consultation can express concerns without pressure.
Community consultation also does not necessarily replace individual consent or formal ethics review where those are required.
Can More Than One Ethical Decision Be Defensible?
Yes. Ethical reasoning may support more than one permissible approach when the available evidence does not establish a decisive reason to prefer one.
Suppose a research team is considering whether to provide participants with individual feedback about a nonclinical educational assessment.
One defensible approach might provide carefully contextualized feedback because participants could find it useful. Another might avoid individualized feedback because the instrument was designed for group-level research and lacks sufficient validity for individual interpretation.
The appropriate decision depends on the instrument's properties, participants' expectations, foreseeable consequences, and applicable requirements.
Recognizing multiple defensible options is not the same as claiming that every option is ethically acceptable. Some alternatives may rest on unsupported assumptions, disregard important interests, or violate binding standards.
Ethical pluralism permits reasonable disagreement while still allowing critical evaluation of arguments.
What Should You Do When the Research Team Cannot Reach Agreement?
Persistent disagreement should be handled through an established decision-making process rather than informal pressure or unilateral action.
The team should clarify the issue, document the competing positions, identify unresolved factual questions, and determine who has authority to make or approve the decision.
Where the issue affects participant welfare, consent, confidentiality, or an approved protocol, consultation with the relevant ethics committee may be necessary.
Independent advice may also be appropriate when the principal investigator has a conflict of interest or when institutional authority could discourage junior researchers from expressing concerns.
Seeking independent ethical advice does not imply that the research team has failed. It can provide a more impartial assessment when internal deliberation cannot adequately resolve the issue.
If required authorization remains pending, the affected research activity should not proceed merely because a deadline is approaching.
How Should the Final Ethical Decision Be Documented?
A decision record should preserve the reasoning that connects the ethical problem to the selected action.
For consequential decisions, documentation may include:
- The ethical question and the circumstances giving rise to it.
- The relevant facts, uncertainties, and applicable requirements.
- The competing ethical considerations and stakeholder interests.
- The reasonable alternatives examined.
- The evidence and arguments supporting each significant option.
- The selected decision and its justification.
- The consultations, approvals, and safeguards involved.
- The unresolved limitations and circumstances requiring reconsideration.
Not every minor judgment requires a lengthy written analysis. The documentation should be proportionate to the consequences and the applicable institutional procedures.
For substantial disagreements, recording why an alternative was rejected can help future reviewers understand the decision without assuming that dissenting researchers acted unreasonably.
The process of documenting difficult ethical decisions supports accountability, but documentation alone does not make an ethically indefensible decision acceptable.
Can a Defensible Ethical Decision Later Turn Out to Be Wrong?
Yes. Ethical decisions are often made under uncertainty, and subsequent evidence may reveal consequences that were not reasonably anticipated.
A researcher might implement safeguards that appear proportionate based on the available information, only to discover that the procedure creates an unexpected confidentiality risk.
The appropriate response is not necessarily to conclude that the original decision was made irresponsibly. Its quality should be evaluated in relation to the evidence, obligations, and reasonable alternatives available at the time.
Nevertheless, new information may create a duty to revise the procedure, introduce additional safeguards, notify relevant authorities, or discontinue an activity.
Ethical justification is therefore not merely a one-time exercise completed before data collection. It may require continuing attention throughout the research process.
When circumstances materially change, researchers should consider whether an ethically approved study remains appropriate to continue.
