Ethics is not paperwork that sits beside a research project; it is part of the research itself. A study that harms participants, uses animals carelessly, or handles personal data recklessly is a bad study whatever its findings, and journals will reject it on ethical grounds alone. At the same time, many authors — especially students submitting for the first time — find ethics confusing, because the rules come from several places at once: their institution, their country's law, and the journal they submit to. This guide explains the core requirements in plain language: ethics approval, informed consent, data protection, animal welfare, and research integrity.
One important note before starting: ethics requirements vary by institution, country, and journal. This guide explains the widely shared principles and the questions you should ask, but it does not replace your own institution's ethics committee or your target journal's author guidelines. Always check both before you begin data collection, because approval obtained after the study is finished is usually worthless — ethics approval must come first, not afterwards.
Most research involving people, their data, or animals needs approval from an ethics committee or institutional review board before data collection begins. The committee's job is to check that the study's benefits justify its risks, that participants are protected, and that the methods are sound. Applying takes time — committees meet on a schedule, and they often ask for revisions — so build approval into your project timeline from the start, not as an afterthought in the week before fieldwork.
A typical application describes the research question, the methods, the participants, the risks and how they will be managed, the consent process, and how data will be stored and protected. For student projects, the supervisor usually reviews the application first, and many institutions have a lighter process for low-risk studies such as anonymous surveys. Do not assume your study is exempt: even a simple questionnaire can need approval if it asks about health, behaviour, or other sensitive topics, or if it involves children or other vulnerable groups. When in doubt, ask the committee — that question costs nothing, while collecting data without required approval can invalidate the whole study.
When you write the paper, the methodology section should state the approving committee and, where the journal requires it, the approval reference number. Journals routinely check this, and a manuscript without an ethics statement for research that clearly needed one will be returned before peer review. If your study genuinely did not need approval — for example, because it used only fully public, anonymised data — say so explicitly and explain why, rather than leaving the question unanswered.
Informed consent means that a participant understands what the study involves and freely agrees to take part. It has three parts, and all three matter. First, information: the participant must be told, in language they understand, what the study is about, what they will be asked to do, how long it will take, what the risks and benefits are, and how their data will be used. Second, comprehension: the researcher must check that the participant actually understood, not merely that a form was signed — this matters especially where participants have low literacy, a language difference, or a condition affecting understanding. Third, voluntariness: the participant must be free to refuse or withdraw at any time without penalty, and must know that.
Consent is normally recorded in writing, but the form is not the consent — the conversation is. A signed form from a participant who did not understand the study is not informed consent. For some studies, such as observation in public settings or analysis of anonymised records, a committee may approve a different consent arrangement or a waiver, but that decision belongs to the committee, not to the researcher alone. Special care is needed with children, patients, employees, and students, where a person may feel unable to refuse: a teacher researching their own students, or a clinician researching their own patients, must arrange consent so that refusal carries no consequence, often through an independent person.
Consent also covers what happens afterwards. Participants should know whether their data will be anonymised, whether quotations may be published, whether data may be shared or reused, and how long data will be kept. If you plan to share data openly — which many journals now encourage — consent must cover that sharing from the start, because data collected without permission to share usually cannot be shared later. Our guide on checking and avoiding plagiarism covers a different integrity question, but the principle is the same: permissions must be obtained before use, not assumed afterwards.
Confidentiality is a promise, and breaking it can harm real people. Personal data should be collected only where needed, stored securely with access limited to the research team, and anonymised as early as possible. In small studies, anonymisation needs care: a combination of age, village, occupation, and condition can identify a person even when their name is removed, so check that reported details cannot identify participants, especially in case studies and qualitative quotations.
Data protection rules differ by country, but the shared expectation is the same: tell participants how their data will be handled, keep data only as long as needed, and follow your institution's storage rules. Where a study involves health records, genetic data, or other sensitive information, extra safeguards apply, and the ethics committee will expect to see them in the application. If a data breach or an accidental disclosure happens, report it through your institution's process promptly — concealing it is a far more serious matter than the incident itself.
Research involving animals carries a special responsibility, and it is central to veterinary and animal science work. The widely used framework is the 3Rs: Replacement, Reduction, and Refinement. Replacement means using non-animal methods wherever they can answer the question — cell cultures, computer models, or existing data. Reduction means using the smallest number of animals that can give a reliable answer, which is a matter of good study design and statistics, not of cutting corners. Refinement means reducing suffering and improving welfare for the animals that are used: better housing, gentler procedures, proper pain management, and humane endpoints defined before the study starts.
Animal studies need approval from the relevant animal ethics committee, and the paper should report the approval, the species and numbers used, the housing and care conditions that affect the results, and the steps taken to minimise suffering. Reporting guidelines for animal research, such as the ARRIVE guidelines, set out the details journals expect, and following them makes both the ethics and the science of the paper stronger. Field studies on wildlife and farm-based studies have their own welfare and permission questions — landowner permission, wildlife regulations, and the welfare of animals in their normal environment — which should be settled before fieldwork, not discovered during it.
Veterinary researchers should also think about the dual role they sometimes hold: a veterinarian treating animals and a researcher studying them. When the same person is both clinician and researcher, the interests of the animal patient and the study can pull in different directions, and the study design should make clear which takes priority. In clinical veterinary research, owner consent is the equivalent of participant consent: owners must understand the study, its risks to their animal, and their right to withdraw, in plain language, before the animal is enrolled.
Ethics does not end with approval and consent. Research integrity covers the honesty of the work itself: reporting methods and results accurately, not inventing or altering data, not copying the work of others, and giving credit where it is due. Fabrication, falsification, and plagiarism are the most serious breaches, and journals now check for them routinely — plagiarism detection software is standard, and data manipulation is increasingly detected. The consequences go beyond one rejected paper: a finding of misconduct can end a career.
Authorship is an integrity question too. An author should have contributed substantially to the work and should approve the final manuscript; adding a senior name who did not contribute, or omitting a contributor who did, are both misconduct. Funding sources and conflicts of interest must be disclosed, because a reader cannot judge a study fairly without knowing who paid for it and what interests the authors hold. When in doubt about any of these questions, the published guidance of bodies such as the Committee on Publication Ethics (COPE) and the International Committee of Medical Journal Editors (ICMJE) is the standard reference, and your institution's research integrity officer can advise on specific cases. If you discover an error in your own published work, correcting it promptly through the journal is the honest course, and journals respect authors who do it.
An ethics application is much less intimidating once you know what the committee is actually asking. Start early, before your data collection dates are fixed, because review commonly takes weeks and committees often meet only once or twice a month. First, find the right committee: your institution usually has one for human participants and a separate one for animal work, and studies involving both need both. Ask your supervisor which committee handled similar projects, and read an approved application from your department if examples are available.
The form usually asks for the same set of things, so you can draft most of it before you open the portal. State the research question and why the study matters, describe who will take part and how they will be recruited, and explain the risks honestly, including minor ones such as inconvenience. Attach a participant information sheet written in plain language and a consent process that gives people time to decide and a real way to withdraw. For animal work, the equivalent section covers the species, the number of animals and its justification, housing and care, the procedures, and the humane endpoints at which an animal would be removed. Reviewers look hardest at whether you have applied the 3Rs: replacing animals where possible, reducing numbers without weakening the design, and refining procedures to lessen harm.
Expect questions back, and treat them as normal rather than as a rejection. Committees commonly ask for clearer wording in the information sheet, a stronger data storage plan, or a smaller sample with a better justification. Answer each point directly, revise the documents, and resubmit through the amendment route the committee provides. Once approval is granted, keep the approval letter with your study records and do not start data collection before the approval date. If anything changes later, such as a new site, a different questionnaire, or a change in animal numbers, submit an amendment first and wait for approval. That paper trail protects your participants, your animals, and the credibility of your published paper.
| Area | What is required | Common mistake |
|---|---|---|
| Approval | Committee approval before data collection | Starting fieldwork while approval is still pending |
| Consent | Information, comprehension, free choice, recorded | Treating a signed form as consent without understanding |
| Confidentiality | Secure storage, early anonymisation, careful reporting | Publishing details that identify a participant |
| Animal welfare | 3Rs: Replacement, Reduction, Refinement | Using more animals than the design needs |
| Integrity | Accurate reporting, proper authorship, disclosure | Gift authorship or undisclosed conflicts of interest |
These authoritative external resources explain the standards and guidance mentioned in this article:
Does an anonymous questionnaire need ethics approval?
Often yes. Anonymity removes some risks but not all, especially if the questions are sensitive or the participants are vulnerable. Ask your ethics committee rather than assuming an exemption.
Can I get ethics approval after collecting data?
Normally no. Approval must come before data collection, and retrospective approval is usually refused. If you have already collected data without approval, seek advice from your institution's ethics committee immediately rather than submitting the paper and hoping no one asks.
What if a participant wants to withdraw?
They may withdraw at any time without giving a reason and without penalty. Tell them this before they consent, and explain what happens to their data on withdrawal, since fully anonymised data often cannot be retrieved and removed.
Do I need owner consent for veterinary clinical research?
Yes. Owners must understand the study, its risks to their animal, and their right to withdraw, in plain language, before the animal is enrolled. Where the researcher is also the treating veterinarian, arrange consent so the owner feels free to refuse.
What are the 3Rs?
Replacement (use non-animal methods where possible), Reduction (use the smallest number of animals that gives a reliable answer), and Refinement (reduce suffering and improve welfare for animals that are used).
Where do I state ethics approval in the paper?
In the methodology section: the approving committee and, where required, the approval reference number, plus confirmation of consent or animal welfare standards. Some journals also require an ethics declaration alongside the manuscript.