03 · What You Need to Know
Recognizing When Withdrawal Is Ethically Justified and How to Leave Responsibly
Why Can Continuing Research Become Ethically Unacceptable?
Research projects depend on conditions that support their scientific validity, ethical acceptability, and institutional authorization.
Those conditions may deteriorate during the research process. New evidence may reveal unacceptable participant risks, serious methodological defects, unauthorized data use, or conduct inconsistent with research integrity requirements.
In some cases, the problems can be corrected. In others, continuing the project would require investigators to disregard obligations that cannot reasonably be compromised.
The Singapore Statement on Research Integrity emphasizes honesty, accountability, professional courtesy, and good stewardship of research. The European Code of Conduct for Research Integrity similarly addresses responsible research practices and the need to maintain reliable research records.
These principles support the view that completing a project is not an overriding obligation when continuation would undermine the integrity of the work.
What Does Walking Away Actually Mean?
Researchers should distinguish several actions that are often described informally as leaving a project.
| Action |
What it means |
Important responsibility |
| Leaving the research team |
An individual stops participating in the project. |
Arrange appropriate transfer of duties, records, and outstanding obligations. |
| Withdrawing from authorship |
A contributor declines to be identified as an author of a manuscript. |
Clarify contribution, approval, accountability, and unresolved integrity concerns. |
| Pausing research activities |
Specified procedures are temporarily stopped pending assessment. |
Protect participants and follow applicable notification and approval requirements. |
| Terminating a study |
The research is formally discontinued. |
Follow authorized termination procedures and address participant and data obligations. |
| Reporting suspected misconduct |
A concern is referred to an appropriate institutional or regulatory authority. |
Provide accurate information and preserve relevant evidence without conducting an unauthorized investigation. |
These actions may occur together, but none automatically substitutes for the others.
A researcher who leaves a team may still need to report serious concerns. A person who withdraws authorship may remain responsible for their own earlier conduct. A study may continue under different personnel when the relevant problems can be resolved.
What Ethical Problems May Justify Withdrawal?
Withdrawal may be justified when continued involvement would require a researcher to participate in conduct that violates applicable ethical, legal, or professional obligations.
Examples include pressure to fabricate or falsify data, conceal consequential methodological limitations, misuse confidential information, disregard serious participant safety concerns, or publish findings that cannot be adequately supported.
Other situations may involve persistent institutional refusal to address credible concerns or fundamental disagreements about the ethical acceptability of the study.
Not every disagreement warrants withdrawal. Researchers should distinguish serious integrity problems from ordinary methodological debate, differences in interpretation, or interpersonal conflict.
The relevant question is whether continued involvement can remain consistent with the researcher's responsibilities after reasonable opportunities for clarification or correction have been considered.
When Is a Methodological Problem Serious Enough to Stop Research?
Methodological limitations vary in severity.
A smaller-than-planned sample, imperfect measurement, or unexpected missing data may reduce precision or generalizability without making a study scientifically worthless.
However, a project may become difficult to justify when essential evidence is irretrievably unreliable, the intended inference cannot be supported, or the remaining research activities impose burdens without a credible prospect of useful knowledge.
Researchers should assess whether the problem can be addressed through transparent limitations, revised analyses, additional safeguards, or a reformulated research question.
Where ethical constraints make the available methodology weaker, the appropriate response may be redesign rather than immediate abandonment.
Nevertheless, continuing a study solely to recover previous investment is not a scientific justification.
Why Sunk Costs Should Not Determine the Ethical Decision
Research projects require substantial investments of time, funding, institutional resources, and professional effort.
These investments may create pressure to continue even when the project's justification has materially weakened.
Economically, sunk costs are resources already expended that cannot be recovered through the current decision. Their existence does not establish that future investment is worthwhile.
In research ethics, the same reasoning is relevant. Two years of data collection do not make unreliable evidence valid. A substantial grant does not authorize unacceptable participant risks. A nearly completed manuscript does not justify concealing a serious methodological problem.
However, previous commitments may create continuing obligations. Participants may have relied on promised follow-up, funders may require reporting, and institutions may have responsibilities concerning records or equipment.
Researchers should distinguish the emotional difficulty of abandoning invested work from the practical duties that survive withdrawal.
Should Researchers Try to Resolve the Problem Before Leaving?
Where safe and appropriate, researchers should consider whether the concern can be addressed through clarification, correction, consultation, or an authorized change to the project.
For example, a disagreement about statistical interpretation may be resolved through independent analysis. A consent problem may require additional ethics review or restrictions on data use. A safety concern may be addressed through suspension and revised safeguards.
However, researchers should not assume that every concern must be discussed directly with the person responsible for the suspected problem.
Where misconduct, retaliation, evidence destruction, or serious safety risks are plausible, direct confrontation may be inappropriate.
In such circumstances, reporting concerns through the appropriate institutional channel may be more suitable than attempting informal resolution.
Reasonable attempts to resolve a problem should not become a requirement to remain involved in unacceptable conduct indefinitely.
What if the Project Leader Pressures You to Continue?
Power differences can make withdrawal particularly difficult for students, early-career researchers, employees, and collaborators whose careers depend on senior investigators.
A project leader may control access to data, funding, references, authorship opportunities, or future employment.
Researchers should distinguish legitimate expectations concerning assigned work from pressure to disregard ethical or research integrity obligations.
Where concerns involve a supervisor or senior collaborator, confidential advice from an appropriate research integrity officer, ethics office, ombudsperson, or other authorized support service may help clarify available options.
Relevant institutional protections and reporting procedures should be reviewed before disclosing sensitive allegations broadly.
Researchers should not assume that every institution provides identical confidentiality or anti-retaliation protections.
When professional retaliation is a credible concern, planning how to raise concerns safely may be necessary before taking further action.
Can You Withdraw Your Name From a Manuscript Without Resolving the Underlying Problem?
Authorship withdrawal and research integrity accountability are related but distinct.
The International Committee of Medical Journal Editors recommends that authors satisfy specified criteria concerning contribution, manuscript approval, and accountability. Its recommendations are influential in biomedical publishing, although disciplinary and journal policies vary.
A researcher who cannot approve a manuscript because its conclusions are unsupported should not be pressured to provide approval merely to preserve an authorship position.
However, removing a name does not automatically resolve suspected data fabrication, participant harm, or other problems affecting the research record.
Depending on the circumstances, the researcher may still need to communicate concerns to the appropriate institution, editor, or other authorized body.
Researchers should also avoid assuming that withdrawal from authorship erases responsibility for work they previously performed.
What Responsibilities Remain When a Researcher Leaves?
Departure from a project does not necessarily end obligations concerning research records, confidential information, participant welfare, or institutional property.
Researchers may need to transfer approved responsibilities, return equipment, preserve records, provide information needed for ongoing participant follow-up, and comply with applicable data retention requirements.
The institution or principal investigator may retain responsibility for the project, but departing personnel should follow the relevant handover procedures.
Research records should not be deleted, altered, or removed without authorization merely because a researcher disagrees with the project leader.
Where integrity concerns exist, preservation of relevant evidence may be particularly important.
The procedures for handling research records when a team member leaves should be followed alongside applicable confidentiality and institutional requirements.
What Should Happen to Participants if a Researcher Withdraws?
Participants should not be abandoned because a researcher decides to leave.
The appropriate response depends on the researcher's role, the study's nature, and whether the project continues under other authorized personnel.
If the researcher provides an essential intervention or monitoring function, an orderly transition may be necessary to protect participants from disruption.
Where the entire study is suspended or terminated, investigators and institutional authorities should consider communication, follow-up, data handling, and any continuing obligations arising from participation.
The Declaration of Helsinki addresses responsibilities concerning participant welfare and arrangements following medical research, although its specific requirements should not be assumed to apply identically to every nonmedical study.
Researchers should avoid treating their personal decision to withdraw as authorization to terminate procedures or commitments beyond their authority.
Should Suspected Misconduct Be Reported Before or After Withdrawal?
There is no universal sequence suitable for every situation.
Where immediate participant safety or evidence preservation is at stake, prompt reporting may be necessary. In other circumstances, a researcher may first seek confidential advice about reporting obligations and protection from retaliation.
Researchers should preserve relevant information through authorized means and avoid conducting an independent investigation beyond their role.
Allegations should be communicated accurately, distinguishing direct observations from suspicions or interpretations.
Withdrawal should not be used to conceal known serious problems, but researchers should also avoid making unsupported public accusations.
The appropriate reporting pathway depends on the institution, jurisdiction, funding arrangements, and nature of the concern.
How Should the Decision to Leave Be Documented?
A consequential withdrawal decision should be recorded sufficiently to explain the circumstances, concerns, alternatives considered, advice obtained, and actions taken.
The record should distinguish verified facts from unresolved allegations and preserve the chronology of relevant events.
Researchers should also document the transfer of responsibilities, location of authorized records, outstanding participant obligations, and any required institutional notifications.
Confidential information should be protected. The departure record should not become an uncontrolled collection of participant data or allegations about colleagues.
Where the decision involves substantial ethical uncertainty, a structured ethical decision record can help demonstrate why withdrawal was considered necessary.
Documentation does not guarantee that the decision was correct, but it permits appropriate scrutiny and reduces reliance on retrospective recollection.
When Is Walking Away Premature?
Withdrawal may be premature when the concern is based on incomplete information, the disagreement concerns a legitimate methodological choice, or a reasonable corrective process remains available without exposing anyone to unacceptable risk.
For example, disagreement about whether to use thematic analysis or another qualitative approach does not ordinarily establish a research integrity violation.
Similarly, a study producing unexpected results may remain scientifically valuable even when the findings are disappointing.
Researchers should examine whether the problem is substantive, whether its consequences are serious, and whether available alternatives could preserve the project's ethical and scientific acceptability.
However, the possibility of resolution should not be used to pressure someone into continuing conduct they reasonably believe violates binding obligations.
Where the appropriate course remains uncertain, independent ethical advice may help clarify the distinction between a resolvable disagreement and a fundamental reason to withdraw.
