It is tempting to imagine that research integrity is secured by rules: define misconduct clearly enough, publish the policy, and trust will follow. But anyone who has worked in research knows that the hardest integrity questions are rarely about clear-cut fabrication or plagiarism. They are about the grey areas — how to handle an awkward result, who deserves to be an author, what to do about a supervisor’s expectations, how much detail to report. These are resolved less by rulebooks than by the training researchers receive, the culture of the environments they work in, and the institutions that support good practice and respond when it fails. Together these form an integrity ecosystem. This article examines that ecosystem — responsible conduct of research training, research culture, and the bodies that uphold integrity — drawing on the research integrity domain of the CASRAI Dictionary.
Beyond the binary of misconduct
Research integrity is often framed around its most serious violations — fabrication, falsification and plagiarism — and those are rightly treated as gravely wrong. But focusing only on the extremes misses where most integrity is actually won or lost. Far more common, and far more corrosive in aggregate, are the everyday questionable practices: selective reporting of results, inappropriate authorship, sloppy record-keeping, inadequate description of methods, the small compromises made under pressure. These rarely trigger a formal investigation, yet they undermine the reliability of the literature just as surely. A mature view of integrity, therefore, is not merely about catching the worst behaviour but about cultivating the everyday good practice that prevents the slide toward it. That is fundamentally a matter of training and culture, not enforcement alone.
Responsible conduct of research training
Responsible conduct of research (RCR) training is the educational component of the ecosystem: structured instruction in the principles and practices of doing research well. Good RCR training goes well beyond reciting the definitions of misconduct. It covers the design and management of data, the responsible use of methods, the norms of authorship and contributorship, the handling of conflicts of interest, the ethics of working with human participants and animals, mentoring relationships, peer review, and how to navigate the genuine dilemmas that arise in practice. Its purpose is formative rather than merely cautionary: to help researchers internalise good practice as a professional habit, and to give them the vocabulary and confidence to recognise and discuss integrity questions before they become problems. Training works best early and repeatedly, woven through a research career rather than delivered once as a compliance exercise.
Research culture: the environment that shapes behaviour
Training has limited effect in a hostile environment. The single most powerful determinant of whether researchers act with integrity is the culture around them — the incentives, pressures, examples and norms of their immediate setting. A culture that rewards quantity over quality, that prizes positive results, that tolerates exploitative mentoring, or that punishes the admission of error, will undermine the best training. A culture that values careful work, supports the reporting of negative or null results, models good authorship practice, and makes it safe to raise concerns will reinforce it. Much of the recent attention to research integrity has accordingly shifted toward research culture — recognising that you cannot train your way out of an environment whose incentives push in the wrong direction. Changing culture is slower and harder than running a course, but it is where the deepest gains lie.
The institutions of the integrity ecosystem
Surrounding training and culture are the bodies that set expectations, offer guidance and respond to allegations. These operate at different levels and play complementary roles:
- National oversight and policy bodies establish standards and, in some systems, handle allegations involving publicly funded research. In the United States, for example, the Office of Research Integrity (ORI) oversees integrity in research funded through its remit.
- National advisory and support organisations promote good practice and offer guidance to institutions and researchers. In the United Kingdom, the UK Research Integrity Office (UKRIO) provides independent, confidential advice and education on research integrity.
- Institutions themselves carry the front-line responsibility: providing training, fostering culture, maintaining policies, and investigating concerns fairly when they arise.
- Journals, publishers and editorial bodies uphold integrity at the point of publication, through editorial policies, correction and retraction processes, and shared community guidance.
No single layer is sufficient alone. The strength of the ecosystem comes from their overlap: training informs culture, culture is reinforced by institutional expectations, and oversight bodies provide both standards to aspire to and a backstop when prevention fails.
Where integrity meets everyday practice
The integrity ecosystem is most visible not in dramatic misconduct cases but in the ordinary disputes it helps prevent and resolve. Authorship is the classic example: disagreements over who should be named, in what order, and on what basis are among the most common and most damaging integrity problems in everyday research life, and they are best handled through clear norms, transparent contribution practices and, where needed, fair institutional processes — the subject of our guidance on resolving authorship disputes. Many such disputes never arise at all when contribution is recorded honestly and explicitly from the outset. Structured contributorship through the CRediT taxonomy — whose roles are described in our overview of the CRediT roles — supports integrity in exactly this way, by making who-did-what a matter of explicit record rather than later contention.
A shared language for integrity
Integrity training, culture and oversight all depend on people meaning the same things by the same terms — what counts as authorship, what constitutes a conflict of interest, what a particular contribution amounts to. When those terms drift between institutions, training schemes and policies, the ecosystem fractures. A consistent vocabulary keeps it coherent, which is what the CASRAI Dictionary provides: shared definitions so that the concepts at the heart of responsible conduct are understood the same way across the training, the culture and the institutions that together make research trustworthy.