03 · What You Need to Know
How Preregistration and Registered Reports Work Differently
Preregistration Creates an Advance Record
In ordinary preregistration, researchers specify important elements of the research before the relevant data or results are known and deposit the plan in a system that creates a time-stamped record.
The Open Science Framework, for example, currently describes preregistration as posting a time-stamped, read-only study plan before beginning the relevant data collection or analysis.
The plan may contain hypotheses, sampling procedures, outcome definitions, exclusion criteria, stopping rules, and analyses. The exact content depends on the research design and on which decisions need advance specification.
Researchers then conduct the study. Later, they may submit the resulting manuscript to a journal through the conventional publication process.
Nothing about preregistration alone means that a journal has evaluated or agreed to publish the study.
A Registered Report Moves Peer Review Before the Results
The Registered Report model changes the publication sequence.
Rather than completing the study and then asking a journal to evaluate the entire manuscript, researchers submit a Stage 1 manuscript before the main results are known. The exact requirements vary among journals, but Stage 1 typically includes the research question and rationale, hypotheses where appropriate, methodology, and proposed analyses.
Editors and reviewers can therefore evaluate the research question, design, feasibility, methodological rigor, and analysis plan at a point when substantive changes can still be made.
This prospective review is a defining difference.
Preregistration
The research plan is recorded in advance, but ordinary preregistration does not itself subject that plan to journal peer review.
Registered Report
The Stage 1 research plan is evaluated through a participating journal's peer-review process before the main results are known.
In-Principle Acceptance Changes the Publication Incentive
If a Stage 1 Registered Report successfully completes peer review, the journal may issue in-principle acceptance, often abbreviated IPA.
This is a conditional commitment to publish the completed research provided the authors satisfactorily conduct the approved study, meet the journal's quality requirements, appropriately handle or explain deviations, and draw conclusions supported by the evidence.
Crucially, publication is not supposed to depend on whether the hypothesis is supported or whether the result is statistically significant, positive, novel, or exciting.
For example, Elsevier's current Registered Reports guidance states that an accepted Stage 1 entails in-principle acceptance for Stage 2 independently of whether the eventual findings are negative, inconclusive, positive, or unexpected, subject to the requirements of the format. Scientific Reports similarly states that final acceptance does not depend on the results obtained but on methodological rigor and the development of the research question.
Ordinary preregistration offers no equivalent publication commitment.
The Two Approaches Address Overlapping but Different Problems
Preregistration primarily makes the chronology of research decisions more transparent. Readers can compare the advance plan with the completed study and identify which hypotheses, outcomes, exclusions, or analyses were specified before the relevant results were known.
Registered Reports retain that prospective element but also intervene in the publication process itself.
Traditional journal review normally occurs after results exist. Editors and reviewers therefore know whether the study produced striking, statistically significant, null, or otherwise appealing findings when making publication decisions.
Registered Reports move the initial editorial decision earlier. By evaluating the question and methodology before the outcomes are known, the format is intended to reduce the role of result direction in determining publication and to provide methodological feedback while the design can still be improved.
The Workflow Is Different
| Feature |
Preregistration |
Registered Report |
| Advance research plan |
Yes |
Yes |
| Time-stamped prospective record |
Central to preregistration |
Typically incorporated into the Registered Report process |
| Journal peer review before main results are known |
No, not inherently |
Yes |
| Opportunity for reviewers to improve the design prospectively |
Not provided by preregistration itself |
Yes, through Stage 1 review |
| In-principle acceptance |
No |
Yes, after successful Stage 1 review under the journal's policy |
| Publication protected from unfavorable result direction |
No |
A central feature, subject to Stage 2 requirements |
| Second journal review after the study is completed |
Only through ordinary manuscript review if submitted |
Yes, as Stage 2 review |
| Ability to conduct exploratory analyses |
Yes |
Yes, when appropriately distinguished and consistent with journal requirements |
Stage 1 Happens Before the Main Results Are Known
Stage 1 is where the Registered Report differs most visibly from conventional publishing.
Reviewers are evaluating the proposed research rather than reacting to its eventual outcome. Current Elsevier reviewer guidance, for example, asks Stage 1 reviewers to consider the importance of the research question, rationale for the hypotheses, soundness and feasibility of the methodology and analysis pipeline, reproducibility, and prespecified outcome-neutral quality checks.
That timing creates an unusual advantage: methodological criticism can arrive while researchers can still do something about it.
A reviewer who identifies a serious design problem after a conventional study has already been completed may be correct but rather late to the methodological party. In a Registered Report, the same criticism can potentially improve the research before the principal data are collected.
Stage 2 Evaluates the Completed Research
After the study is completed, researchers prepare the Stage 2 manuscript containing the results and interpretation.
Reviewers then assess issues such as whether the approved methodology was followed, whether deviations were appropriately handled, whether prespecified quality criteria were satisfied, and whether the conclusions are supported by the evidence.
Specific workflows differ among publishers. For example, Elsevier currently publishes accepted Stage 1 and Stage 2 articles as linked publications under its model, whereas other journals may register or archive the Stage 1 protocol and publish the completed work as a single final article.
Researchers should therefore consult the target journal's current Registered Report instructions rather than assuming that all implementations are procedurally identical.
In-Principle Acceptance Is Conditional, Not an Unconditional Guarantee
The phrase “accepted before the results are known” can be misunderstood.
In-principle acceptance does not mean that whatever researchers eventually submit must be published. Authors still need to complete the study satisfactorily, adhere sufficiently to the approved protocol or appropriately manage deviations, meet prespecified quality requirements, and interpret the results defensibly.
Journals can have additional conditions.
The important protection is narrower: a hypothesis failing to receive support should not by itself turn an otherwise acceptable study into an unpublishable one.
Watch Out
Do not describe in-principle acceptance as an unconditional publication guarantee. It is a prospective commitment subject to the journal's Stage 2 requirements. Always check the exact Registered Report policy of the journal you are considering.
A Registered Report Can Still Deviate From the Stage 1 Plan
Prospective peer review does not make research perfectly predictable.
Problems can arise after Stage 1. Recruitment may fail. Equipment may become unavailable. An analysis may prove inappropriate. An unforeseen methodological issue may require modification.
Registered Report policies generally provide mechanisms for handling justified deviations, although the procedures vary. Elsevier, for example, currently advises authors to consult the handling editor before deviating from proposed materials and methods.
This makes the distinction between legitimate deviations and problematic undisclosed changes particularly important. Under a Registered Report, the journal may also be involved in evaluating or approving substantial modifications.
Registered Reports Can Include Exploratory Analyses
A Registered Report does not mean that researchers must ignore everything they did not predict at Stage 1.
Unexpected findings can be investigated. Additional analyses can be conducted. New hypotheses can emerge.
The key is to distinguish those analyses from the work that was prospectively reviewed and approved. The Registered Report model is intended to protect confirmatory inference without eliminating scientific exploration.
The same general principle applies to preregistration: exploratory and confirmatory analyses can coexist when their status is reported clearly.
You Can Preregister Without Submitting a Registered Report
This is the more common conceptual direction: preregistration can stand on its own.
You can register a study plan, conduct the research, and later submit the completed manuscript through ordinary journal channels. The preregistration remains useful as a record of the advance plan even though no journal reviewed that plan prospectively.
You therefore do not need to find a Registered Report journal simply because you want to preregister.
A Registered Report Normally Includes a Registration Component
The reverse relationship is closer.
Registered Reports are built around a prospectively specified protocol. Journal implementations commonly require the Stage 1 protocol to be registered or archived after in-principle acceptance. The Open Science Framework currently provides a dedicated Registered Report Protocol Preregistration template for this purpose.
Still, the defining feature of a Registered Report is not simply that the protocol was deposited somewhere. It is the combination of prospective journal review and in-principle acceptance before the main results are known.
Not Every Journal Offers Registered Reports
Preregistration can be conducted independently of journal participation. Registered Reports, by contrast, require a journal that offers the format and a study that fits that journal's scope and eligibility requirements.
Policies can differ in accepted study designs, timing, review criteria, Stage 1 requirements, treatment of pilot data, procedures for deviations, and publication format.
If you are considering a Registered Report, identify potential journals early. Waiting until the study is complete defeats the prospective logic of Stage 1 review.
