Этика · Бастапқы
Plan ethical research and protect participants
Plan review, consent, permissions and privacy before collecting data.
11 мин оқу · 2026 ж. 26 қаз.
Бұл бет әлі аударылмаған, сондықтан ағылшын нұсқасын оқып отырсыз.
Негізгі тұжырымдар
- Separate review from consent and publication permission.
- Protect vulnerable participants and identifiers.
- Report actual determinations.
Мазмұны
Plan review, consent, permissions and privacy before collecting data.
Ethical research begins when you decide what to study, whom to involve and what information to collect. It continues through analysis, writing, publication and data sharing. A manuscript cannot repair a recruitment process that pressured students or a public dataset that exposed participants who expected confidentiality.
Requirements depend on the institution, jurisdiction, population and activity. This article explains questions to resolve with the responsible supervisor, institution or review body. It does not establish legal permission for a particular project. A student project, online survey, public dataset or intention to publish on a student website does not by itself determine whether review is required.
Connect ethical principles to everyday decisions
Three influential principles in the Belmont Report are respect for persons, beneficence and justice. Respect concerns autonomy and appropriate protection; beneficence concerns avoiding unnecessary harm and considering benefits; justice concerns fair treatment in the distribution of research burdens and benefits. The report arose in the United States and is a foundational ethical source, not a substitute for the rules governing every country or discipline. S107
Translate principles into concrete decisions. If a project asks classmates to discuss stress, respect requires a meaningful opportunity to decline. Considering harm means thinking about embarrassment, unwanted disclosure and distress, not only physical injury. Justice asks why these classmates are being recruited and whether convenience places the burden on people with little freedom to refuse.
Scientific quality also matters ethically. Collecting burdensome information with no coherent question or usable analysis may waste participants’ effort. Plan the design and sample carefully so that the information has a reasonable prospect of answering the stated question. Do not promise direct benefit when the project is primarily educational or exploratory.
Separate the permissions that are often confused
Several decisions may be needed, and one does not automatically supply another.
| Decision | What it concerns | What it does not automatically establish |
|---|---|---|
| Institutional or ethics review | Whether the proposed activity meets applicable requirements | Every later change is authorised |
| Informed participation | Whether an individual understands and voluntarily agrees to participate | Permission for every future use of identifiable material |
| Access permission | Whether researchers may obtain or use a dataset, site or record | Permission to redistribute it publicly |
| Publication consent | Whether identifiable material may appear in a publication | Permission for unrelated analysis or reuse |
| Data licence or agreement | Conditions for use, redistribution and attribution | Removal of privacy or ethical responsibilities |
For medical research, the World Medical Association’s 2024 Declaration of Helsinki addresses independent review, participant protections and informed consent. ICMJE also specifies protections for identifiable patient information in medical publishing. Their scope should be stated correctly rather than presented as a universal legal code for all student research. S109 S09
Before collecting data, identify who is responsible for the relevant determination. Keep the actual decision and approved documents. If the institution determines that a particular form of review is unnecessary, record that determination accurately. Do not turn your own assumption into an official exemption statement.
Make consent an understandable process
Informed consent is a voluntary decision based on understandable information, not merely a signature or a ticked box. Explain the purpose, activities, duration, foreseeable risks, realistic benefits, compensation if any, withdrawal arrangements and contact route. State whether recordings are made, who will access them, how data will be stored, and whether quotations or datasets may be shared.
Use language suited to the participants. Ask a person unfamiliar with the project to explain what they think participation involves. If they misunderstand, revise the explanation. Avoid implying that participation is required for good grades, access to a service or approval from a teacher.
Assent generally describes an affirmative agreement from a person who cannot provide the legally required consent independently. For research involving minors, the roles of guardian permission, assent and consent must be established with the responsible body. Ages, exceptions and requirements vary. A website account or a parent’s acceptance of platform terms is not automatically consent to a research project.
Withdrawal also needs an honest explanation. It may be possible to remove identifiable records before a stated processing stage, but impossible to retrieve a contribution once it has been irreversibly anonymised or combined into published summaries. Explain the approved arrangement before participation rather than promising unlimited deletion that cannot be delivered.
Recognise power and social pressure
A teacher, supervisor or team leader may unintentionally pressure people even while saying participation is voluntary. Recruitment during class, public requests to raise a hand, or an interviewer who controls future opportunities can make refusal visible or costly.
Consider an approved arrangement in which a person outside the grading relationship handles invitations and keeps participation decisions private from the teacher. Provide an equivalent alternative if participation is connected to a course activity and institutional rules require one. The appropriate solution depends on the setting; describe the actual arrangement in the Methods rather than merely asserting that there was no coercion.
In peer research, friendship can also create pressure. A classmate may disclose more than intended because the interviewer is a friend. Discuss boundaries, confidentiality and the limits of the researcher’s role with the supervisor. Sensitive topics require a plan for responding to distress or unexpected disclosures through the approved institutional process.
Distinguish privacy, confidentiality and anonymity
Privacy concerns access to people and their information. Confidentiality concerns how entrusted information is protected. Anonymity concerns whether a person can be identified from the data and available context. Pseudonymisation replaces direct identifiers with codes while a link or other means of identification remains. Terminology can have specific legal definitions, so use the definitions required for your setting.
Removing names does not necessarily make data anonymous. A combination of age, school, role and an unusual event can identify someone. Audio recordings contain voices. Exact quotations from online posts can be searchable. A file labelled “anonymous survey” may still contain email addresses, precise timestamps or platform identifiers.
Collect only the information needed for the question. Store an identity key separately when it is needed, restrict access and use approved storage. Plan retention and deletion with the institution. Do not place participant identifiers in filenames, figure labels, file properties or public code examples. A public repository should be checked for accidental identifiers in all files, not only the final spreadsheet.
Work through a social science example
A fictional student proposes interviewing classmates about pressure during research placements. The first draft says: “All interviews are anonymous and will be uploaded so others can check the work.” This promise is internally inconsistent if recordings retain identifiable voices and contextual details.
The revised plan identifies a supervisor, obtains the required institutional determination, recruits without linking participation to assessment, and explains recording and quotation arrangements. The analysis uses participant codes. The identity key and recordings have restricted access. The team considers whether particular combinations of details could identify a person before including quotations.
A fictional excerpt says, “As the only first year student in Professor X’s laboratory, I felt unable to disagree.” Removing the name alone may not protect the speaker. Depending on consent and the approved analytic approach, the team may omit identifying details, use an explicitly described paraphrase or leave the excerpt out. The meaning should not be altered to strengthen the argument. If detailed context is essential to interpretation, explain the ethical tradeoff and seek appropriate guidance.
A useful Methods paragraph would describe recruitment, consent, the researcher’s relationship to participants, recording, access and protection of quotations. It should not claim that data are anonymous if the research team can reconnect codes to identities.
Address online and secondary data thoughtfully
Public visibility is not the same as unrestricted ethical use. People may share material in one context without anticipating its aggregation, analysis or republication in another. The Association of Internet Researchers recommends considering context, vulnerability and potential consequences across the research process. S108
For a proposed online study, check access terms, data licences, the institutional determination and the possibility of identification. Do not assume that quoting a searchable post is equivalent to reporting an anonymous interview excerpt. Explain how sampling, quotation and sharing choices reduce foreseeable harm while preserving the integrity of the analysis.
For an existing dataset, record its provenance, version, licence and permitted uses. Check whether the proposed analysis fits the original permissions and any data agreement. A dataset supplied by a collaborator is not automatically available for public redistribution. If only controlled access is permitted, describe that access route accurately in the paper.
A review using published literature often has different oversight needs from new participant research. It still requires responsible handling of identifiable case details, copyrighted figures and restricted supplementary data. Do not copy another paper’s ethics statement into your review as if it described your own work.
Apply the same reasoning in science and engineering
Natural science example. A student uses animal tissue supplied by an approved laboratory. The manuscript should identify the provenance and applicable approval or institutional determination accurately, based on the actual arrangement. Receiving a sample does not authorise a new procedure outside the permitted project. ARRIVE supports transparent reporting of animal research; it does not grant permission to conduct it. S03
Engineering example. A prototype records video to evaluate pedestrian detection. The paper may focus on technical accuracy, but the recordings can still contain identifiable people and locations. The team needs an appropriate plan for collection, access, retention and publication. Replacing a video filename with a number does not remove the faces inside it. A demonstration image should be assessed separately from the numerical performance table.
Laboratory hazards, field access, equipment use and environmental impacts also require the relevant supervision and assessments. A publication tutorial cannot replace training or authorisation. Describe procedures relevant to interpreting the results without presenting absent approvals as completed facts.
Plan responsible sharing and use of external tools
Data sharing can strengthen research when permissions and protections allow it. It can also expose people when done without care. Consider whether a deidentified analysis dataset, aggregate results, a controlled access route or synthetic demonstration data can meet the transparency goal. Synthetic data must be labelled and should not be represented as the observations underlying the study.
Do not upload confidential recordings, unpublished sensitive data or identifiable participant material to an external writing or analysis service without the required authorisation and safeguards. Check the actual service and institutional requirements. Removing a name from a prompt may be insufficient when the remaining text identifies a person through context.
Keep disclosure about AI assistance separate from claims of ethics approval or human authorship. See Article 13 for authorship and declarations. Responsibility for the submitted work remains with the human authors and the applicable research process.
Write declarations from verified facts
An ethics statement should identify the responsible body and reference number where applicable, explain consent arrangements and describe relevant limits. If an authorised body determined that review was not required, state the basis and actual determination without inventing an approval identifier. Avoid the phrase “ethical approval was unnecessary” when no responsible assessment occurred.
Data availability should describe what is actually available, where it is held and under what conditions. “Available on request” should name a realistic access process and should not promise access that consent or a licence prohibits. Explain material restrictions directly.
Practice. A fictional author writes: “Names were removed, so all transcripts are anonymous and can be published freely.” Identify two problems.
Model answer. Removing names does not establish anonymity because quotations and contextual details may identify participants. It also does not establish permission to publish transcripts. The author must review identification risk, consent and data agreements with the responsible institution and describe a permitted sharing approach.
Complete the ethics fields in Template 2 before the study and the verified declarations in Template 12 before submission. Ethical writing accurately reports the protections that were implemented and the limits that remain.
Sources
S03: ARRIVE. The ARRIVE guidelines 2.0. Official guidance, accessed 11 November 2026.