A Registered Report’s Stage 1 in-principle acceptance (IPA) locks the study’s hypotheses, methods, and confirmatory analysis plan — but it does not, by itself, lock the author list. Between Stage 1 protocol acceptance and Stage 2 results submission, the research team that actually runs the study is very often not identical to the team named on the accepted Stage 1 manuscript, and researchers new to the format frequently assume, incorrectly, that the same “nothing changes” rule applies to authorship. It doesn’t. This guide sets out what does and doesn’t hold constant about a Registered Report’s byline across the two stages, and how authorship changes are actually handled when they happen.
For the mechanics of the two-stage format itself — what in-principle acceptance guarantees, how Stage 2 review differs from a conventional review, which journals and services offer the track — see CASRAI’s companion guide, Registered Reports: How Preregistered Study Plans Get Accepted Before Data Collection, which this page assumes as background. For guidance on drafting the Stage 1 manuscript itself, see How to Write a Registered Report.
What Stage 1 acceptance actually locks — and what it doesn’t
In-principle acceptance is a commitment from the journal to publish the study’s results, in the format of a Stage 2 manuscript, provided the authors execute the approved protocol as registered. What’s locked is the content of the protocol: the stated hypotheses, the method, the sample size and how it was determined, the confirmatory analysis plan. Anything not specified at Stage 1 can still appear at Stage 2, but only as clearly labelled exploratory analysis carrying less evidentiary weight than the pre-registered confirmatory tests.
Authorship is a separate question, governed by separate rules. The Center for Open Science’s own Registered Reports guidance describes the in-principle-acceptance mechanism in terms of the protocol, not the byline — it does not specify a Registered-Reports-specific authorship-change procedure. That’s left to each individual journal’s ordinary authorship policy, which in turn generally follows the same norms — largely traceable to ICMJE — that govern authorship changes on any manuscript, applied to the specific circumstance of a Registered Report.
Why Registered Report author lists commonly change mid-study
Registered Reports create a structural reason for authorship to shift that most manuscripts don’t: a real gap, often many months or multiple years, between when the protocol is accepted and when the results exist to write up. During that gap, data collection and analysis actually happen — and a predictable set of personnel changes tends to occur within it:
- A collaborator joins for execution. A research assistant, postdoc, or lab member hired specifically to run the approved protocol, collect data, or carry out the registered analysis plan may not have been involved at the conception/design stage that produced the Stage 1 manuscript.
- A specialist is added for analysis. A statistician or methodologist is sometimes brought in once data collection is complete, to execute the pre-specified analysis plan or to conduct the exploratory analyses that Stage 2 permits beyond it.
- An original author departs. A graduate student graduates and loses institutional access, a postdoc changes institutions or leaves academia, or a collaborator becomes unresponsive during the sometimes lengthy data-collection period.
- The corresponding author or principal investigator changes. Particularly on multi-year protocols, the person accountable for communicating with the journal at Stage 1 is not always the same person available to do so at Stage 2.
None of this is unique to Registered Reports — personnel turnover happens on any long-running study — but the two-stage structure makes the timing unusually visible: the journal has a documented, dated record of exactly who was named at protocol acceptance, against which any later change is a legible departure rather than something that simply appears, unremarked, in the final submitted manuscript.
How the authorship-change process actually works
Because COS leaves this to individual journal policy, the exact procedure varies by outlet, but it generally follows the same pattern publishers apply to any authorship change requested after initial submission, per ICMJE‘s widely adopted recommendation: the editor should be notified in writing of the requested change, with an explanation for why it’s warranted, and a signed statement of agreement from every author currently on the manuscript — including, specifically, the author being added or removed. Publishers such as Springer Nature apply a similar standard: additions or removals are not routine, but are permitted with a stated reason and editorial sign-off, typically at the discretion of the editor-in-chief or handling editor.
Applied to a Registered Report specifically, the practical steps are:
- Notify the handling editor as soon as the change is known — not retroactively at Stage 2 submission. Most Registered Reports guidelines already ask authors to flag the editor about any deviation from the approved protocol; an authorship change is worth surfacing the same way, even though it isn’t a protocol deviation itself.
- Confirm the new or departing contributor actually meets (or no longer meets) authorship criteria. The applicable test is the same one that applies to any manuscript — see CASRAI’s guide to ICMJE’s four authorship criteria — not a Registered-Reports-specific standard. A person who joined only to collect data under existing, protocol-specified procedures, with no role in drafting or revising the manuscript, may be better credited in the acknowledgments than the byline; see ICMJE’s own guidance on the distinction, and CASRAI’s dictionary entry on authorship for the underlying criteria.
- Get written, signed agreement from all continuing authors and from the author being added or removed, consistent with standard ICMJE-style practice.
- Update the CRediT contribution statement to reflect what each person actually did across the full study, not just at whichever stage they joined — see the next section.
- Disclose the change plainly in the Stage 2 cover letter, alongside any protocol deviations being reported, so the editor and reviewers have the full picture when evaluating the Stage 2 manuscript.
A useful mental model: an authorship-change request on a Registered Report is processed as an ordinary authorship change, on an unusual timeline. The IPA guarantees the journal’s commitment to the protocol; it does not pre-approve, or forbid, any particular roster of names attached to that protocol later.
Documenting a changed roster with CRediT
The CRediT taxonomy (ANSI/NISO Z39.104-2022) is well suited to exactly this situation, because it records what each person did rather than simply that they contributed. A Stage 1-only author’s contribution statement can genuinely differ from a Stage 2-only author’s: someone who wrote the original hypotheses and design might carry Conceptualization and Methodology, while someone who joined afterward to run the protocol might carry Investigation and Data Curation, and a statistician added post hoc might carry Formal Analysis. See CASRAI’s guides to writing a CRediT author statement and how CRediT roles can be shared or held by multiple people for the mechanics of assembling a statement that reflects contributions accurately across a roster that changed mid-project.
This is also where an early authorship agreement pays off on a Registered Report specifically: because the Stage 1 protocol is written and locked well before the study is executed, a team that agrees up front on how authorship and CRediT roles will be assigned if the roster changes during data collection — rather than negotiating it retroactively once Stage 2 is being drafted — avoids exactly the kind of dispute COPE’s guidance exists to resolve.
If the corresponding author changes
A change to who serves as corresponding author is its own sub-case, since that person is the journal’s primary point of contact and, in most authorship frameworks, the one attesting to the accuracy of the final submission. Where the corresponding-author role needs to pass to someone else between Stage 1 and Stage 2 — commonly because the original corresponding author has left the institution or the project — that handover should be made explicit to the editor at the same time as any other authorship notification, not left implicit in the Stage 2 submission metadata.
What an authorship change does not affect
It’s worth being explicit about the boundary, since it’s easy to conflate the two: a change in the author list does not reopen the locked protocol. The hypotheses, methods, and confirmatory analysis plan that earned in-principle acceptance still bind the study regardless of who is named on it. If a newly joined author wants to run the study differently from what was registered, that’s a protocol-deviation question, handled the way any deviation is handled — disclosed to the editor and reported transparently at Stage 2, distinct from and unaffected by whatever authorship-change process is running in parallel. A new team member does not get to retroactively alter what was pre-registered simply by virtue of joining the author list.
Conversely, unresolved disagreement about who belongs on the author list, or in what order, is an authorship dispute like any other, and is not resolved by the Registered Reports format itself. COPE’s flowcharts for ghost, guest, and gift authorship and for handling authorship disputes generally apply here exactly as they would on any manuscript — see CASRAI’s overview of COPE’s guidelines. The two-stage structure changes when a dispute is likely to surface (often at the roster-change point during execution, rather than only at final submission), not how it gets resolved.
Practical checklist for research teams
- Confirm, at Stage 1 drafting, whether your target journal or review service (a journal’s in-house track, or a service such as PCI Registered Reports) publishes any authorship-change policy of its own, rather than assuming COS’s general guidance covers it.
- Agree early — ideally in a written authorship agreement — on how roles and CRediT credit will be handled if the team who executes the study differs from the team who wrote the protocol.
- Treat any addition, removal, or reordering of authors as its own notification to the handling editor, separate from (but disclosed alongside) any protocol-deviation reporting at Stage 2.
- Get signed agreement from every continuing author and from the author being added or removed before submitting the change.
- Update the CRediT statement to reflect actual contributions across both stages, not just the roster present when the manuscript is finally submitted.
- Keep the corresponding-author handover explicit if that role changes hands.
Frequently asked questions
Does in-principle acceptance lock the author list the same way it locks the methods?
No. In-principle acceptance is a commitment to the protocol — the hypotheses, methods, and confirmatory analysis plan — not to a specific roster of names. Authorship changes are handled through the journal’s ordinary authorship-change process, generally following ICMJE-style norms, independently of the protocol lock.
Can someone be added as an author after Stage 1 acceptance?
Yes, provided they meet the applicable authorship criteria (typically ICMJE’s four criteria or the journal’s equivalent policy) and the addition follows the journal’s authorship-change process — written justification, signed agreement from all authors including the one being added, and editorial approval. This is common on Registered Reports specifically because data collection and analysis, which can bring in new contributors, happen after Stage 1 rather than before it.
What if an original Stage 1 author didn’t do enough to remain an author by Stage 2?
The same authorship criteria that apply at initial submission apply throughout — contribution is not credited simply for having been named at Stage 1. Removing an author who no longer meets the criteria, or who did not ultimately contribute to the executed study, follows the same signed-consent, editor-notified process as any other authorship-change request, and should be handled directly and transparently rather than left unaddressed.
Does a change in authorship count as a protocol deviation that has to be reported at Stage 2?
Not on its own. Protocol deviations concern the study’s design, methods, and analysis plan; authorship changes concern who is credited for the work. They’re reported through separate channels — deviations in the Stage 2 manuscript’s methods/deviations reporting, authorship changes through the journal’s authorship-change process — though both should be disclosed to the handling editor as soon as they’re known.
Who decides disputes over authorship on a Registered Report?
The same bodies and frameworks that handle authorship disputes on any manuscript: the journal’s editor in the first instance, drawing on COPE’s guidance and flowcharts, escalating to the authors’ institution(s) if it can’t be resolved editorially. The Registered Reports format itself does not include a separate dispute-resolution mechanism.
