03 · What You Need to Know
Why Research Ethics Looks Different Across Fields
Different Methods Create Different Ethical Problems
Ethics follows the research activity.
A drug trial may expose participants to physiological risks from an intervention. An interview study may involve disclosure of traumatic or illegal experiences. Ethnography may create long-term relationships and blurred boundaries between research and ordinary social interaction. Educational research can involve teachers studying their own students. Digital research may analyze data from thousands or millions of people with whom researchers never interact directly.
A single procedural template cannot respond equally well to all of these situations.
This is why the broad principles of research ethics need interpretation before they become concrete procedures.
Biomedical Research Has Strong Traditions of Prospective Participant Protection
Biomedical and clinical research ethics has been heavily shaped by concerns about interventions, physical risk, vulnerable patients, experimental treatments, therapeutic relationships, and the possibility that participants may confuse research with individualized medical care.
The World Medical Association's 2024 Declaration of Helsinki is specifically directed toward medical research involving human participants, including research using identifiable human material or data. It addresses matters such as scientific requirements, risk and burden, vulnerable groups, ethics committee review, privacy and confidentiality, informed consent, placebo use, post-trial provisions, registration, and publication.
These requirements make sense in their domain. They should not automatically be copied word for word into every form of historical, literary, educational, or computational research without asking whether the concepts fit.
Psychology Has Distinctive Concerns About Deception and Behavioral Manipulation
Experimental psychology sometimes relies on incomplete disclosure or deception because knowing the complete hypothesis or experimental manipulation could change participant behavior.
The American Psychological Association's Ethics Code therefore addresses deception explicitly. It permits psychologists to use deceptive techniques only under specified conditions, including prospective scientific, educational, or applied value and the absence of feasible effective nondeceptive alternatives. It also places limits on deception concerning research expected to cause physical pain or severe emotional distress and requires appropriate explanation of deception.
The same code contains specific provisions for research involving students and subordinates, recording voices and images, incentives, informed consent, and debriefing.
Those standards reflect recurring features of psychological research rather than an entirely separate moral universe.
Educational Research Often Has to Manage Dual Roles and Power Relationships
Educational research frequently occurs in settings where researchers already have roles as teachers, supervisors, administrators, or evaluators.
This creates ethical questions that may be less prominent in research involving unrelated community volunteers. Can students freely decline participation when the researcher assigns their grades? Will nonparticipation affect classroom opportunities? Can research activities be separated from required instruction? Who can access identifiable educational data?
The APA Ethics Code provides one example of discipline-relevant safeguards: psychologists conducting research with students or subordinates must take steps to protect them from adverse consequences of declining or withdrawing, and when participation is a course requirement or extra-credit opportunity, an equitable alternative is required.
The exact requirements for educational research depend on jurisdiction, institution, discipline, and research design, but the broader ethical issue is recognizable: existing authority relationships can change what voluntariness means in practice.
Qualitative Research Can Challenge One-Time Models of Consent
In many qualitative designs, researchers cannot specify every future conversational direction before data collection begins. Interviews evolve. Participants introduce unexpected topics. Ethnographic relationships develop over time. Researchers may learn information they never anticipated collecting.
This can make consent more relational and ongoing than a single transaction at enrollment.
That does not mean qualitative researchers are exempt from informed consent. It means the practical application may require continuing communication about participation, recording, quotation, anonymity, changing research focus, or use of unexpected information.
Qualitative research also raises distinctive questions about whether anonymization removes information essential to interpretation and whether richly contextualized quotations can indirectly identify participants.
Ethnography Creates Ethical Relationships That Extend Beyond Formal Participants
Ethnographic research may involve prolonged immersion in communities, organizations, workplaces, or social groups. Researchers can encounter people who have not gone through a conventional enrollment process yet become visible within fieldnotes and analysis.
Ethical questions may therefore concern relationships, reciprocity, representation, community consequences, confidentiality within networks, researcher positionality, and what can realistically be promised in a social environment where participants know one another.
A consent model designed for a single laboratory encounter may not fully resolve those issues.
Digital and Computational Research Can Involve People at Enormous Scale
Computational research creates a different ethical geometry.
Researchers may analyze millions of publicly accessible posts, infer sensitive characteristics from seemingly ordinary data, link datasets originally collected for unrelated purposes, or study algorithmic systems whose outputs affect populations.
Researchers may never meet the people represented in the data. That distance can make participant risk feel abstract even when re-identification, privacy, discrimination, manipulation, or group-level consequences remain plausible.
Questions such as whether data are public, whether users expected research use, whether consent is practicable, and whether quotations can be traced back to individuals may require methods-specific ethical reasoning.
The fact that research can be unethical without physical injury becomes particularly important in these fields.
Archival and Historical Research May Involve Different Relationships to Consent
Historical and archival researchers often work with documents created long before the present research project and sometimes long after the people described have died.
Conventional prospective informed consent may therefore be impossible or conceptually inappropriate. Yet ethical questions can remain concerning privacy, living relatives, sensitive information, culturally restricted materials, stigmatization, ownership, access conditions, and responsible representation.
The absence of a conventional research participant does not mean the absence of ethical judgment. It changes what the relevant judgment concerns.
Animal Research Uses Different Ethical and Regulatory Frameworks
Research involving animals cannot simply apply human informed consent principles because animals cannot provide consent in the human-research sense.
Ethical frameworks instead emphasize matters such as scientific justification, animal welfare, appropriate species and numbers, refinement of procedures, reduction of unnecessary animal use, and replacement with alternatives where feasible. Specific legal and institutional requirements vary by jurisdiction and setting.
This is a clear example of why research ethics cannot consist of one universal procedural checklist.
Disciplinary Variation Does Not Mean Ethical Relativism
If standards vary, does that mean whatever a discipline normally does is ethical within that discipline?
No.
Professional conventions themselves can deserve ethical criticism. A field may inherit practices that inadequately protect privacy, systematically exclude certain populations, normalize problematic power relationships, or fail to anticipate new technological risks.
“This is standard practice in my discipline” is relevant contextual information. It is not a complete ethical justification.
The Same Principle Can Produce Different Procedures
Respect for persons provides a useful example.
In a clinical trial, respect may require a detailed prospective consent process describing interventions and alternatives. In ethnography, respect may involve repeated negotiation of participation and careful decisions about representation as relationships develop. In low-risk secondary-data research, the relevant question may concern privacy, legitimate access, identifiability, and whether individual consent is required or can appropriately be waived under the applicable framework.
The principle remains recognizable while the operational procedure changes.
“Minimal Risk” Is Context Dependent
Risk itself can look different across disciplines.
A biomedical researcher may focus on adverse physiological events. A political scientist studying dissidents may be more concerned about surveillance or retaliation. An education researcher may worry about academic consequences. A workplace researcher may consider employment risk. A digital researcher may focus on re-identification or exposure of sensitive attributes.
Ethical review should therefore identify the harms plausible in the research context rather than importing only the most familiar risk categories from another discipline.
Different Fields Also Have Different Professional Codes
Professional organizations often develop ethics codes responsive to the practices of their members.
The APA Ethics Code, for example, contains research-specific standards addressing institutional approval, consent, recording, student and subordinate participants, inducements, deception, debriefing, animal research, and reporting research results.
Other disciplines may emphasize different matters because their methods and professional relationships differ.
Professional codes can provide valuable specificity, but researchers should determine their status. A disciplinary code does not automatically replace applicable law, regulation, institutional policy, or ethics committee requirements.
Interdisciplinary Research Can Be Subject to Several Ethical Traditions at Once
Contemporary projects increasingly cross disciplinary boundaries.
A study might combine machine learning, clinical data, qualitative interviews, and educational interventions. Which discipline's ethics applies?
Potentially several.
The better approach is to identify the activities and risks involved rather than selecting the disciplinary framework that imposes the fewest constraints. Clinical data may trigger health-research and privacy requirements. Interviews may require attention to consent and confidentiality. Algorithmic analysis may introduce bias, inference, or re-identification concerns. Educational recruitment may create teacher-student power relationships.
The ethical framework should follow the research, not the departmental letterhead.
Institutions Can Apply Common Protections Across Disciplines
Even when disciplines have different professional norms, institutions may impose a common human-research framework.
In the United States, OHRP notes that many institutions choose to apply Common Rule provisions to all human-participant research regardless of funding source. Institutions may also adopt policies providing protections beyond regulatory requirements.
A sociologist, psychologist, education researcher, and biomedical researcher at the same institution may therefore share some institutional requirements while still applying discipline-specific reasoning to their different methods.
Country and Discipline Are Separate Sources of Variation
Disciplinary variation should not be confused with geographic variation.
Two anthropologists working in different countries may share disciplinary traditions while operating under different laws and institutional requirements. Two researchers in the same country may work under very different professional standards because one conducts clinical trials and the other studies public online behavior.
The question of how ethical standards vary by country and setting therefore needs to be considered separately from disciplinary variation.
