Registered Reports is a peer-reviewed journal publication format, not a self-filed registry entry. Where ordinary preregistration is something a researcher does unilaterally — time-stamping a plan before starting a study, with no outside evaluation and no promise attached — a Registered Report moves peer review itself earlier, before data are collected or outcomes are known, and attaches a binding editorial commitment to the outcome of that review. This guide walks through the mechanics: how Stage 1 review works, what in-principle acceptance actually guarantees, how Stage 2 review differs from ordinary manuscript review, which journals and services currently offer the format, and exactly where it diverges from plain preregistration.
The two-stage review process
Stage 1: reviewing the plan, not the results
At Stage 1, authors submit a manuscript containing the study’s background and rationale, its hypotheses, and a fully specified method and analysis plan — sample size (and how it was determined), measures, procedure, exclusion criteria, and the statistical tests or models to be used. Pilot data is often included to demonstrate feasibility. No results exist yet, or, for studies using secondary or archival data, the outcomes have not yet been examined.
Reviewers at this stage evaluate the same things a normal review would judge about the introduction and methodology — is the question worth asking, is the logic sound, is the design capable of actually answering it, is the analysis plan appropriate — but they are structurally unable to judge the study on its results, because none exist. That single fact is the entire mechanism by which Registered Reports address HARKing (hypothesizing after results are known) and publication bias: an editor cannot favor a statistically significant or novel-looking finding over a null one when no finding has happened yet.
In-principle acceptance (IPA)
If a Stage 1 manuscript passes review — typically after one or more rounds of revision, the same as conventional peer review — the journal grants in-principle acceptance. This is a binding editorial commitment to publish the completed study once it is executed as approved, regardless of whether the results are statistically significant, null, or contrary to the original hypothesis. It is not an unconditional guarantee: publication still depends on the authors actually following the approved protocol, disclosing and justifying any deviations, and executing the study competently. A study that abandons its registered design without explanation, or is executed so poorly that the data can’t support any conclusion, can still be rejected at Stage 2 — but a study that was carried out as planned cannot be rejected because the editor doesn’t find the result exciting.
Because the approved Stage 1 protocol is normally deposited on a public registry (frequently OSF Registries) at the point of IPA, the Stage 1 manuscript itself typically functions as the study’s preregistration — see the differences section below for exactly how that overlap works.
Stage 2: reviewing adherence, not excitement
After data collection, authors submit a Stage 2 manuscript that adds the results and discussion to the already-approved introduction and methods. Stage 2 review checks the completed study against the approved Stage 1 protocol: did the authors run the analyses they said they would, on the sample and measures they specified, using the exclusion rules they pre-registered? Any analysis that wasn’t specified at Stage 1 can still be reported, but it must be clearly separated from the confirmatory results and labelled exploratory — it did not go through the same pre-data scrutiny and doesn’t carry the same evidentiary weight.
What Stage 2 reviewers are explicitly not supposed to do is relitigate whether the finding is interesting, novel, or likely to be cited. That editorial judgment was already made at Stage 1, before anyone knew what the data would show.
Which journals and services offer Registered Reports
Registered Reports were independently proposed in 2012 at the journals Cortex and Perspectives on Psychological Science; Chambers’ 2013 paper in Cortex, “Registered Reports: A new publishing initiative at Cortex,” is the commonly cited founding description of the modern format. The Center for Open Science (COS), which maintains the Registered Reports initiative, reports that over 300 journals across psychology, neuroscience, medicine, ecology, and other fields now offer the format, either as a standing submission track or through one-off special issues.
Two concrete routes worth knowing apart:
- A journal’s own in-house track. Nature Human Behaviour is a widely cited example: it runs a standing Registered Reports article type with its own submission guidelines, and Nature Portfolio maintains a cross-journal collection page indexing Registered Reports coverage across its titles.
- PCI Registered Reports (rr.peercommunityin.org). Part of Peer Community In, this is a journal-independent review service: it recruits its own recommenders to run Stage 1 and Stage 2 review on preprints, entirely outside any single journal’s editorial office. A published list of “PCI RR-interested” journals — Nature Human Behaviour among them — have agreed to be notified when PCI Registered Reports issues an in-principle acceptance or Stage 2 recommendation, and may then offer direct publication or a streamlined further review rather than re-running peer review from scratch. In practice, one approved Stage 1 protocol can lead to publication through either path: a journal’s own track, or a PCI RR recommendation picked up by a participating journal.
Because the exact list of participating journals changes as more titles adopt the format, check a target journal’s own author guidelines and, for the current count and journal list, the Center for Open Science’s registered-reports initiative page directly rather than relying on any fixed list.
How Registered Reports differ from plain preregistration
The two are related but not interchangeable, and the terms are often used loosely as if they were:
| Plain preregistration | Registered Report | |
|---|---|---|
| Who acts | The researcher, unilaterally, on a registry (OSF Registries, AsPredicted, ClinicalTrials.gov) | The researcher and a journal, jointly, through a submission and review process |
| Independent review of the plan | None — the registry time-stamps whatever is submitted without evaluating it | Peer review at Stage 1, by the same kind of reviewers who’d evaluate a conventional manuscript |
| Publication guarantee | None — registering a plan does not obligate any journal to publish the eventual results | In-principle acceptance is a binding commitment to publish, contingent on the protocol being followed |
| When it happens | Before data collection, entirely at the researcher’s discretion and timeline | Before data collection, but gated by journal submission deadlines and review rounds |
| Effect on publication bias | Documents what was planned, so deviations are detectable — but a null result can still simply go unpublished | Removes the incentive at the point of editorial decision, since the decision is already made before results exist |
In practice a Registered Report’s Stage 1 protocol typically is the study’s preregistration — it’s usually deposited on a registry (often OSF) at the point of in-principle acceptance — so the two aren’t competing choices. A Registered Report is best understood as one specific, journal-mediated way of formalizing a preregistration and attaching a publication commitment to it. For the registry-based mechanics of preregistration itself — what a registration needs to specify, the major platforms, and how it interacts with clinical trial registration and funder policy — see CASRAI’s Preregistration of a Study Protocol guide, which this page complements rather than repeats.
Why the format exists
Registered Reports were designed to address two related problems in conventional publishing: HARKing and publication bias. When editorial decisions happen only after results are known, there is a structural incentive — for authors and, some evidence suggests, for editors — to favor findings that are statistically significant, novel, or a clean confirmation of the original hypothesis, and for null or messy results to go unpublished or get quietly reframed. By locking the publication decision to the quality of the question and method rather than the outcome, Registered Reports remove that incentive at its source rather than trying to correct for it after the fact (for example through funnel-plot detection of missing null results in a meta-analysis). See CASRAI’s metascience entry for the broader research-on-research context this format sits within, and the p-hacking entry for the related analytic malpractice a fixed, reviewed analysis plan also constrains.
Frequently asked questions
Does in-principle acceptance guarantee publication no matter what happens?
No. It guarantees publication provided the approved protocol was actually followed, any deviations are disclosed and justified, and the study was executed competently. A study that departs substantially from its approved plan without explanation, or is executed so poorly the data can’t support a conclusion, can still be rejected at Stage 2.
Can a Registered Report include analyses that weren’t part of the original plan?
Yes, but they must be clearly separated from the pre-specified confirmatory analyses and labelled exploratory. Reviewers evaluated the confirmatory plan at Stage 1; anything added afterward didn’t go through that same pre-data scrutiny.
Is Registered Reports only for experimental psychology?
No. It originated in experimental psychology and neuroscience but has since spread to clinical research, ecology, and computational fields, including some analyses of secondary or archival data where the “before outcomes are known” standard applies instead of “before data collection.”
Is PCI Registered Reports the same thing as a journal’s own Registered Reports track?
No. PCI Registered Reports is a community-run, journal-independent review service that evaluates Stage 1 and Stage 2 submissions on preprints. A separate list of participating (“PCI RR-interested”) journals can then offer publication based on that review rather than commissioning their own from scratch.
If I preregister a study on OSF, do I automatically have a Registered Report?
No. A standard preregistration on a registry is a unilateral, unreviewed record — no editor has evaluated it and no journal has committed to publishing the results. It only becomes a Registered Report if a journal (or a service like PCI Registered Reports) has actually peer-reviewed the plan and granted in-principle acceptance before the results existed.
Related CASRAI resources
- Registered Report — the dictionary definition, with additional detail on the Chambers & Tzavella (2022) history of the format.
- Preregistration of a Study Protocol: What It Is and How to Do It — the registry-based practice in full, including OSF Registries, AsPredicted, and clinical trial registration.
- Pre-registration
- HARKing (Hypothesising After Results are Known)
- p-hacking
- Metascience
