Ethical Considerations in Conducting Research and Publishing Papers (2026)

Research ethics and publication ethics govern how researchers conduct studies and report findings. Violations — ranging from fabricated data to undisclosed conflicts of interest — damage scientific credibility and, in serious cases, cause direct harm. Understanding the ethical standards that apply at each stage of the research process is essential for any researcher, whether publishing for the first time or managing a programme of work.

Core Principles of Research Ethics

Research ethics across disciplines rests on four foundational principles established in the Belmont Report (1979) and subsequent frameworks:

• Respect for persons (autonomy): Participants have the right to make informed decisions about their involvement. Vulnerable populations (children, prisoners, cognitively impaired individuals) require additional protections.
• Beneficence: Research should maximise benefits and minimise harm. This applies to participants, communities, and the broader society affected by the research outcomes.
• Non-maleficence: Research should not cause unnecessary harm — physical, psychological, financial, or reputational — to participants or communities.
• Justice: The burdens and benefits of research should be distributed fairly. Research should not exploit vulnerable populations for the benefit of others.

Informed Consent

Before any research involving human participants begins, researchers must obtain informed consent. Informed consent requires that participants:

• Are told the purpose of the research, what participation involves, and how long it will take
• Understand what data will be collected and how it will be used and stored
• Are informed of any foreseeable risks or discomforts
• Know they can withdraw at any time without penalty
• Give consent freely, without coercion

Written consent is standard. For certain low-risk research (anonymous surveys, publicly available data), waiver of written consent may be appropriate — but this must be approved by your institutional review board (IRB) or ethics committee.

Research Integrity: Data Handling and Reporting

Research integrity means reporting findings honestly and completely — including negative or inconclusive results. The main integrity violations in research practice:

Fabrication

Inventing data that was never collected. The most serious integrity violation — treated as research misconduct with consequences including retraction of publications and termination of employment.

Falsification

Manipulating existing data — changing values, removing outliers without justification, or selectively reporting only the trials that produced favourable results. The difference between legitimate statistical exclusion (justified by pre-registered criteria) and falsification is documentation: if you decided to exclude outliers before collecting data and specified the criteria, that’s legitimate. If you excluded them after seeing the results to improve significance, it’s falsification.

P-hacking and HARKing

Running multiple statistical tests until a significant result appears (p-hacking) or presenting a post-hoc hypothesis as if it were pre-specified (Hypothesising After Results are Known — HARKing) inflate false positive rates and mislead readers. Pre-registration — publicly documenting your hypothesis and analysis plan before data collection — is the standard response to both.

Selective reporting

Publishing only significant results (publication bias) distorts the evidence base. Journals and funders increasingly require registration of all trials, and researchers are expected to report null results.

Publication Ethics

Authorship

All listed authors should have made substantive contributions to the work — typically defined by the ICMJE criteria: contributing to conception/design or data analysis, drafting or revising critically, and approving the final version. Honorary authorship (listing someone as an author for prestige, not contribution) and ghost authorship (failing to list someone who made substantial contributions) are both ethical violations.

Plagiarism and self-plagiarism

Presenting others’ ideas, data, or text without attribution is plagiarism. Self-plagiarism — republishing your own previously published work without disclosing the prior publication — is a separate but also serious violation, as it misrepresents the novelty of the work. Journals use CrossCheck/iThenticate to screen submissions for both.

Conflicts of interest

Financial relationships with industry, personal relationships with study participants, or ideological commitments to particular outcomes can all bias research design, data interpretation, or reporting. All potential conflicts must be disclosed in publications — even if you believe they didn’t affect your work.

Duplicate publication

Submitting the same manuscript to multiple journals simultaneously, or publishing substantially the same study in multiple outlets, violates publication ethics. Always confirm with the target journal that your submission is exclusive and not under review elsewhere.

Institutional Ethics Oversight

Any research involving human participants, animal subjects, or sensitive data requires ethics committee approval before data collection begins. Retrospectively seeking ethics approval — or conducting research without it — is not acceptable. Most institutions have an IRB or research ethics committee with defined review timelines; factor this into your research planning.