COPE flowcharts are decision-tree diagrams published by the Committee on Publication Ethics (COPE) that walk an editor through a fixed sequence of questions and recommended actions for a specific integrity scenario. Separate flowcharts cover suspected plagiarism, duplicate or redundant submission, authorship disputes, and image manipulation, translating COPE’s Core Practices into a case-specific decision path.
The Committee on Publication Ethics (COPE) is best known to most researchers as a standards body — but its most-used practical output isn’t a policy statement, it’s a library of flowcharts. These are decision-tree diagrams that translate COPE’s Core Practices into a specific sequence of questions and recommended actions for a specific scenario: a reviewer flags suspected plagiarism, a reader spots the same figure in two papers, two co-authors can’t agree on the author list. This guide walks through what the flowcharts actually contain and how editors are meant to use them for four scenario types: suspected plagiarism, duplicate or redundant submission, authorship disputes, and image manipulation. It doesn’t redefine COPE itself — see the COPE dictionary entry for that — and it’s a different angle from CASRAI’s existing guide on how a retraction actually happens, which covers the editorial process once a retraction is already the likely outcome. The flowcharts are the earlier decision layer: how an editor works out what’s actually going on, and what the proportionate response is, before that process starts.
What a COPE flowchart actually is
COPE publishes and periodically updates an extensive library of flowcharts, organized by scenario — separate diagrams exist for a submitted manuscript versus a published article, because the available remedies differ (you can reject a submission; you can only correct, add an expression of concern to, or retract something already published). Each flowchart starts from a trigger (“a reviewer suspects plagiarism,” “a reader alleges an image has been manipulated”) and branches on an assessed degree of severity, ending in a small set of recommended actions — contact the author for an explanation, contact the author’s institution, issue a correction, add an expression of concern, retract, or take no further action. COPE itself, and secondary sources that describe them (including publisher and journal ethics pages that reproduce COPE’s own guidance), frame the flowcharts as a starting point for an editor who isn’t sure how to handle a specific ethical query — not a substitute for judgment, and not a legal or adjudicative ruling. They’re published openly on COPE’s website at no cost and in several languages other than English, and no COPE membership is required to consult them, though COPE membership does give an editorial office additional case-consultation support beyond the flowcharts themselves.
It’s worth being explicit about what the flowcharts are not. They are not COPE deciding anything on the journal’s behalf — the editor retains final decision authority in every case, the same principle that applies to retraction itself (COPE’s role is advisory and standards-setting, not adjudicative). They are not a replacement for gathering actual evidence — several flowcharts explicitly route through a “request documentary evidence and investigate” step before any action is taken. And they don’t resolve the underlying misconduct question (was this actually plagiarism, was this actually fabricated or falsified data) — that’s frequently a question for the author’s institution, not the journal, a point that comes up repeatedly across the specific flowcharts below.
Suspected plagiarism
COPE maintains separate flowcharts for plagiarism suspected in a submitted manuscript and plagiarism found in an already-published article, since the available response differs once a paper is in the published record. For a submitted manuscript, the guidance runs roughly as follows: when a reviewer or editor raises a concern, the editor thanks the person who raised it and requests full documentary evidence rather than acting on an unsubstantiated claim. The next step is an assessment of the degree of copying — COPE draws a line between minor overlap (for example, short recycled phrases in a discussion section, sometimes associated with a non-native English speaker paraphrasing awkwardly, with no misattribution of data or findings) and clear plagiarism (substantial unattributed reuse of text and/or data). Where the copying is more than minor, the authors are contacted directly and the journal’s plagiarism policy and process are explained to them; a manuscript found to contain clear plagiarism is typically rejected, and depending on severity the editor may also consider contacting the author’s institution. The published-article version of the flowchart follows the same overall logic but adds retraction, correction, or expression of concern as the possible published-record outcomes rather than a straightforward rejection.
This is a distinct concern from self-plagiarism (reuse of an author’s own previously published material without disclosure), which COPE and most journals treat as a related but separately assessed issue — the flowchart for reused-own-material cases asks different questions than one for reused third-party material.
Redundant (duplicate) submission and publication
COPE also separates this into a submitted-manuscript flowchart and a published-article flowchart. The submitted-manuscript version applies when an editor suspects a manuscript substantially overlaps with something the same authors have already published or submitted elsewhere. The recommended sequence: assess the degree of overlap; where it’s substantial, contact the authors and ask for an explanation, and review the context of the submission (sometimes there’s a legitimate reason — a conference proceedings paper being expanded into a full journal article, for instance, which is why the assessment step matters rather than a flat rule). If the authors’ response is unsatisfactory, the recommended action is to reject the manuscript, tell the authors clearly why, and set expectations for future submissions; for more serious or apparently deliberate cases, contacting the author’s institution is also part of the guidance. The published-article version of this flowchart follows the same overlap-assessment logic but ends in a correction or retraction decision rather than rejection, since the paper is already part of the record. See the duplicate publication dictionary entry for the underlying definition this flowchart operationalizes.
A practical note that instructions-to-authors pages frequently borrow from this same COPE guidance: journals are advised to state their redundant-publication policy explicitly in their submission instructions, and to have authors affirmatively confirm at submission that the manuscript is original and not under consideration elsewhere — both reduce how often this flowchart needs to be invoked at all.
Authorship disputes
“Authorship dispute” flowcharts actually cover two related but distinct situations, and it’s worth keeping them separate. The first is suspected ghost, guest, or gift authorship — a quality-control concern about who’s listed, independent of whether anyone involved is complaining. COPE defines a ghost author as someone who qualified for authorship but was left off the list; a guest author as someone added (with or without their knowledge) to make the author list look more credentialed despite no real involvement; and a gift author as someone included as a favor, courtesy, or out of seniority rather than for a genuine contribution. Where an editor suspects one of these, COPE’s guidance is to raise it directly with the authors, and — for confirmed cases involving senior researchers who should have known better — to consider informing the relevant department head or research-governance office at the author’s institution.
The second situation is a genuine dispute between named co-authors — over author order, over whether someone should be added or removed, or over credit generally. Here COPE’s stated position is notably restrained: journals are advised to generally avoid trying to adjudicate authorship disputes themselves, and instead to refer the dispute to the authors’ home institution(s), which is better positioned to investigate who actually did what. Separate flowcharts exist for the mechanics of the resulting change requests — adding an author before publication, adding an author after publication, and requesting removal of an author after publication — and a recurring rule across all of them is that a change to the published author list generally requires consent from every listed author, not just the person requesting the change; an editor should not remove or add a name unilaterally on one party’s say-so.
This is where CASRAI’s own infrastructure is directly relevant, not just adjacent: a properly completed CRediT statement (see sample CRediT statement formats) creates a documented, role-by-role record of who contributed what at the time of submission — Conceptualization, Investigation, Writing, and the other 13 CRediT roles disclosed up front, tied to each author’s ORCID iD where publishers support it. That documentation is exactly the kind of contemporaneous evidence an editor needs before triggering one of these flowcharts, and disputes are less likely to reach this point at all when contribution was disclosed clearly from the start. CRediT doesn’t replace the ICMJE authorship-qualification test or resolve a dispute on its own — it’s a contribution-disclosure taxonomy, not an adjudication mechanism — but it materially changes what evidence exists when a dispute does arise.
Image manipulation
COPE’s image manipulation flowchart applies specifically to concerns about digital images — duplication of parts within a single image, reuse of the same image across different figures or panels represented as different experiments, or other alterations to photographic or scan-based figures. This sits alongside, but is handled somewhat differently from, more general fabricated data concerns, because image manipulation is often identifiable through direct visual or forensic comparison in a way that other forms of data fabrication are not. As with the other flowcharts, the recommended path branches on severity and intent: editors are generally guided to distinguish adjustments that don’t change the underlying scientific meaning of an image (for example, uniform brightness or contrast adjustment applied to an entire image) from alterations that add, remove, duplicate, or obscure specific features — the latter is treated as a data-integrity problem regardless of the author’s stated intent, and depending on how central the affected image is to the paper’s conclusions, the outcome can range from a correction to a full retraction. This general severity distinction is consistent with how COPE’s broader retraction guidance treats data fabrication versus honest error.
What the flowcharts don’t do
None of these flowcharts hand an editor a verdict. Every one of them routes through an evidence-gathering and author-response step before recommending any action, and several explicitly push the underlying factual question — did this person actually contribute, was this image actually altered with intent to deceive — outward to the author’s institution rather than asking the journal to investigate and rule on it directly. That division of labor matters for research offices as much as for editors: a research misconduct finding is, in most systems, an institutional determination, not an editorial one, and a journal contacting an institution under one of these flowcharts is often the first formal signal that an institutional process needs to start. The flowchart’s job ends at the editorial decision — correction, expression of concern, rejection, or retraction — not at determining whether misconduct occurred.
Frequently asked questions
Are COPE’s flowcharts mandatory, or just recommended?
Recommended. COPE is a membership and standards organization, not a regulator — its flowcharts and Core Practices are guidance that member and non-member journals alike are free to consult, but nothing in a COPE flowchart is independently enforceable against a journal. Many publishers adopt COPE’s flowcharts as their own internal editorial process by reference, which is a large part of why they’re so widely used in practice.
Where can I find COPE’s official flowcharts?
COPE publishes and maintains its current flowchart library at publicationethics.org, organized by scenario type. That’s the authoritative, up-to-date source — this guide summarizes the general logic and common scenarios rather than reproducing every branch of every flowchart, and COPE updates individual flowcharts periodically as its guidance evolves.
What’s the difference between a COPE flowchart and COPE’s Core Practices?
Core Practices is the higher-level framework — the set of principles COPE expects member journals and publishers to follow across the full publication lifecycle. The flowcharts are the operational layer underneath it: practical, scenario-specific decision aids for applying those principles to one concrete situation (duplicate submission, disputed authorship, and so on) rather than a general statement of policy.
Does COPE itself decide whether a paper gets retracted?
No. COPE’s own materials, and the flowcharts specifically, are advisory — the retraction (or correction, or expression of concern) decision is made by the journal’s editor, using COPE’s and other bodies’ guidance as an input. See CASRAI’s guide on how a retraction actually happens for the fuller editorial process once that decision point is reached.
What happens after a flowchart points toward contacting the author’s institution?
That step hands the underlying factual question — often, whether misconduct actually occurred — to the institution’s own research-integrity process, which is typically better positioned to investigate than the journal (it can access lab records, interview staff, and take actions like requiring retraction that the journal alone cannot). The journal’s own next step usually still needs to be decided independently: it can proceed to a correction or retraction based on what it can establish on its own record, or wait for the institution’s findings, depending on the flowchart and the severity of the concern.
