01 · The Question
What Does It Actually Mean to Conduct Research With Integrity?
Most researchers know that they should not fabricate data, falsify results, or plagiarize someone else's work. But research integrity asks a much broader question: what does responsible conduct look like when no obvious misconduct has occurred?
That matters because many consequential research decisions happen far from the dramatic cases that attract misconduct investigations. Researchers choose methods, document decisions, manage data, assign authorship, disclose conflicts, review other people's work, communicate uncertainty, and sometimes discover mistakes that need to be corrected. Integrity is expressed through these ordinary decisions throughout the research process.
There is no single worldwide rulebook containing every requirement for every researcher. Disciplines, institutions, funders, jurisdictions, and types of research can impose different standards. Still, major research-integrity frameworks converge on a set of principles concerned with honesty, reliability, accountability, fairness, openness, and responsible stewardship of research.
03 · What You Need to Know
Research Integrity Is a Standard for How Research Is Done
Research integrity is broader than research misconduct
One of the easiest ways to misunderstand research integrity is to define it negatively: if you have not committed misconduct, your research must have integrity. That threshold is too low.
Formal definitions of research misconduct identify particularly serious forms of unacceptable conduct. For example, the U.S. Office of Research Integrity defines research misconduct in the research it oversees as fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or reporting research results. Its definition explicitly excludes honest error and differences of opinion.
Research integrity operates on a wider terrain. The National Academies describes integrity in research as planning, proposing, performing, reporting, and reviewing research in accordance with core values. The Singapore Statement similarly presents research integrity through principles and professional responsibilities that extend well beyond the avoidance of fabrication, falsification, and plagiarism.
Research misconduct
A category of serious wrongdoing defined by the applicable policy or regulatory framework. The precise definition and evidentiary requirements depend on the governing system.
Research integrity
The broader commitment to values, standards, and practices that support trustworthy and responsible research throughout the research process.
This distinction also means that behavior can raise legitimate integrity concerns without necessarily satisfying a formal definition of misconduct. The boundary between misconduct and questionable research practices therefore matters, particularly when institutions must decide how a problematic practice should be addressed.
Different frameworks use different words, but their principles substantially overlap
There is no universally mandated vocabulary for research integrity. That is not necessarily a problem. Major frameworks often emphasize similar obligations while organizing them differently.
The National Academies identifies objectivity, honesty, openness, accountability, fairness, and stewardship as foundational values of research integrity. The 2023 revised European Code of Conduct for Research Integrity, published by ALLEA, is organized around reliability, honesty, respect, and accountability. The Singapore Statement identifies honesty, accountability, professional courtesy and fairness, and good stewardship as its four principles.
| Principle |
What it can mean in research practice |
| Honesty |
Represent methods, data, results, contributions, limitations, and uncertainties truthfully rather than making the research appear cleaner or stronger than it is. |
| Reliability and rigour |
Use methods and procedures appropriate to the research question, maintain suitable quality controls, and base conclusions on the evidence actually obtained. |
| Openness and transparency |
Provide sufficient information about methods, decisions, limitations, interests, and, where appropriate and permissible, research materials or data so that the work can be scrutinized and understood. |
| Fairness and respect |
Treat collaborators, participants, reviewers, trainees, communities, and other contributors appropriately, including giving credit where it is due. |
| Accountability |
Take responsibility for your research decisions, contributions, claims, and obligations, and be prepared to explain or correct them. |
| Stewardship |
Use research resources, records, relationships, and the scholarly record responsibly, recognizing that research is part of a larger social enterprise. |
The terminology should not be treated as a universal checklist in which every framework must use precisely the same words. What matters is the underlying expectation that research should be conducted in ways that support the trustworthiness of both the process and the resulting record.
Responsible conduct begins before data collection
Integrity is not something added when a manuscript is written. Decisions made during study design can already affect whether the resulting evidence deserves confidence.
A researcher should formulate questions and choose methods without deliberately engineering the study to produce a preferred conclusion. Necessary approvals should be obtained before activities requiring them begin. Relevant regulations, institutional policies, contractual obligations, disciplinary standards, and study-specific requirements should be identified rather than assumed.
Responsible conduct also means recognizing the limits of your competence. If a project requires statistical, methodological, clinical, technical, cultural, or other expertise that the team does not possess, responsible practice may require consultation or collaboration rather than improvisation.
Data and records should preserve what actually happened
Research records are not merely administrative debris left behind after the interesting part of the project. They are part of the evidence supporting what was done.
The Singapore Statement calls for clear and accurate research records that permit verification and replication of the work by others. What adequate documentation looks like will differ across disciplines. A laboratory notebook, interview audit trail, analysis script, field note, version history, code repository, instrument record, or documented analytic decision may serve different purposes in different forms of research.
Responsible data conduct can involve acquisition, storage, documentation, analysis, security, confidentiality, sharing, retention, and ownership. NIH's responsible-conduct-of-research guidance accordingly includes data acquisition and analysis, recordkeeping, secure and ethical data use, confidentiality, management, sharing, and ownership among the subjects researchers may need to understand.
The underlying principle is straightforward: the research record should not misrepresent how the evidence was obtained or what the evidence shows.
Integrity applies to analysis and interpretation, not just raw data
You can preserve every original data file and still produce misleading research if the analysis or interpretation is handled irresponsibly.
Analytical decisions should have a defensible methodological basis. Researchers should not selectively suppress inconvenient results merely because they weaken a preferred narrative. Conclusions should reflect the strength and limitations of the evidence. Uncertainty should not disappear between the results section and the abstract.
This does not mean researchers can never make judgment calls. Research routinely requires judgment. The integrity question is whether those judgments can be justified by the research question, evidence, methods, disciplinary standards, and relevant prior decisions rather than by a desire to obtain a particular result.
Authorship and credit are integrity issues
Research integrity also governs how researchers represent who did the work. The Singapore Statement calls on researchers to take responsibility for their contributions and to include as authors those, and only those, who satisfy the applicable authorship criteria. Other significant contributions should be acknowledged appropriately.
The exact authorship criteria can vary by discipline, journal, and professional framework. Responsible conduct therefore requires checking the standards that actually apply rather than assuming that laboratory seniority, academic rank, funding, friendship, or supervision automatically settles authorship.
Conflicts of interest should be managed, not merely privately recognized
A conflict of interest does not automatically mean that a researcher has behaved improperly or that the research is invalid. The integrity problem arises when relevant interests are concealed or inadequately managed and could compromise, or reasonably appear to compromise, research judgment.
Depending on the context, responsible conduct may require disclosure to an institution, funder, ethics body, journal, collaborators, or other relevant parties. Financial interests are not the only possible concern. Professional commitments, competing roles, close relationships, or other interests may also require attention under applicable policies.
Peer review carries integrity obligations too
Integrity does not stop when you evaluate someone else's work. Reviewers may receive unpublished ideas, methods, findings, or other confidential material. Responsible review requires appropriate confidentiality, fair and rigorous evaluation, disclosure of relevant conflicts, and avoidance of using privileged information for personal advantage.
This is one reason research integrity should not be equated with "being honest about your own data." Researchers occupy several roles in the research system, and those roles create different responsibilities.
Compliance is necessary, but integrity is not reducible to compliance
Researchers must follow applicable laws, regulations, ethics requirements, institutional policies, funding conditions, and professional standards. The Singapore Statement explicitly identifies adherence to relevant regulations and policies as a researcher responsibility.
But "I followed the rules" is not always a complete integrity analysis. Written rules cannot anticipate every methodological judgment, collaborative dispute, emerging technology, or ambiguous situation. A technically permissible choice can still deserve scrutiny if it undermines reliability, fairness, transparency, or accountability.
Watch Out
Do not assume that ethics approval, journal acceptance, or the absence of a misconduct finding certifies the integrity of every aspect of a study. These mechanisms address particular questions under particular standards. They do not replace the researcher's continuing responsibility for the trustworthiness of the work.
Integrity includes what happens after the research is published
Publication does not end a researcher's responsibilities. Errors may be discovered later. New information may show that an analysis, figure, dataset, interpretation, or conclusion needs clarification or correction.
Responsible conduct generally requires taking problems seriously and responding proportionately. An honest error is not automatically misconduct, but recognizing that distinction does not make the error irrelevant. When a serious problem is discovered, the researcher may need to take action to protect the accuracy of the research record.
Research integrity is both an individual and a system responsibility
It is tempting to describe integrity entirely as a matter of personal character: good researchers make good choices. Personal responsibility certainly matters, but major integrity frameworks also recognize the research environment.
The National Academies argues that institutions, sponsors, publishers, professional societies, and other organizations can support or undermine integrity through their policies, incentives, practices, training, and responses to problems. The Singapore Statement likewise calls on research institutions to sustain environments that encourage integrity through education, clear policies, and appropriate standards.
That becomes especially important in collaborative research, where responsibility cannot simply be assigned to an abstract entity called "the team." Researchers should understand who is responsible for integrity within a research team, while institutions and research leaders should create conditions in which responsible practice is realistically possible.
06 · What This Means for You
How to Apply Research Integrity in Everyday Decisions
You do not need to approach every research decision as though you were preparing testimony for a misconduct hearing. A more useful habit is to ask whether another competent researcher could understand what you did, why you did it, and how your evidence supports what you claim.
When a decision feels uncomfortable, distinguish three questions. First, what do applicable law, regulation, institutional policy, funding conditions, ethics approval, or journal rules require? Second, what does accepted disciplinary or methodological practice support? Third, even if no explicit rule resolves the situation, which option better preserves the reliability, honesty, fairness, transparency, and accountability of the research?
A simple decision framework
If a law, regulation, approval condition, funder requirement, institutional policy, or other binding rule applies
Follow it and verify the current requirement with the responsible authority rather than relying on memory or informal custom.
If the issue is methodological
Ask whether the choice is scientifically or scholarly defensible and whether the research record represents that choice accurately.
If the issue concerns credit, collaboration, review, or competing interests
Ask whether responsibilities, contributions, interests, and decisions are being represented fairly and transparently under the standards that apply.
If no explicit rule gives you an obvious answer
Document the issue, seek appropriate methodological or integrity advice, and choose the course most consistent with trustworthy and accountable research.
If someone senior pressures you to use a practice you cannot defend
Do not treat hierarchy or the claim that
"everyone does it" as evidence that the practice is responsible.
One practical test is deceptively simple: would you be comfortable describing the decision accurately in the methods, research record, ethics documentation, response to reviewers, or conversation with a knowledgeable colleague? Reluctance to disclose a decision is not proof that it is improper, but it can be a useful signal that the reasoning deserves another look.
