1 Definition and purpose
Pre-registration is a research practice in which investigators document a study’s hypotheses, design, methods, and analysis plans before examining the outcomes. It is most common in empirical disciplines where researchers can specify in advance what they intend to test and how they intend to analyze the data. By creating a dated record of these plans, pre-registration helps separate confirmatory work from later exploratory analysis.
1.1 Core meaning
At its core, pre-registration is a commitment to a research plan made before the relevant data are observed. The record may include the main research questions, the sample to be studied, the measurements to be used, and the statistical tests to be applied. The purpose is not to forbid later changes, but to make any changes visible and interpretable.
1.2 Goals of pre-registration
Pre-registration serves several closely related aims. It encourages researchers to think carefully about their study before data collection begins and gives readers a clearer basis for judging how much flexibility existed in the analysis process.
1.2.1 Transparency
A pre-registered study makes the intended procedures visible to other researchers, reviewers, and the public. This transparency can reduce ambiguity about what was planned in advance and what emerged during the course of the project.
1.2.2 Reduction of bias
By specifying hypotheses and analytic decisions beforehand, pre-registration can limit selective reporting, post hoc hypothesis creation, and other forms of research bias. It is especially useful when many possible analyses could be performed on the same dataset.
1.2.3 Reproducibility
Clear advance documentation can make it easier for others to reproduce a study or evaluate whether the reported results follow from the stated methods. Reproducibility is strengthened when the protocol, outcomes, and analysis rules are recorded in a stable form.
1.3 Relationship to open science
Pre-registration is often grouped with open science practices such as open data, open materials, and data sharing. Together, these practices aim to make research more inspectable and easier to verify. Pre-registration does not require all data or code to be public, but it complements other transparency measures.
2 Historical development
Pre-registration developed gradually as researchers sought stronger safeguards against hidden analytic flexibility. Its earliest widespread use was in fields where outcomes could be strongly influenced by study design and where advance specification was already valued for practical and ethical reasons.
2.1 Early use in clinical research
Clinical research was among the first areas to adopt formal trial registration. Recording study protocols in advance helped document what treatments were being tested, what outcomes were planned, and how participants would be enrolled and followed. This practice supported oversight and made it easier to compare published findings with original plans.
2.2 Expansion to psychology and social science
Later, pre-registration spread into psychology and related social sciences, where concerns about flexible analysis, small samples, and publication bias became more prominent. Researchers began registering hypotheses and statistical plans to distinguish predicted effects from patterns discovered after data inspection.
2.3 Adoption in other disciplines
The practice has since expanded into economics, education, neuroscience, and other empirical fields. In each area, the details differ, but the central idea remains the same: documenting a study plan before observing the results.
3 Key components of a pre-registration
A useful pre-registration is specific enough that another researcher could understand what was planned and, in many cases, reproduce the intended analysis. The amount of detail may vary by discipline and study type, but several elements are commonly included.
3.1 Research questions and hypotheses
The registration usually states the main research question and any directional or non-directional hypotheses. Clear hypotheses help distinguish confirmatory claims from broader exploratory goals. If a study has multiple hypotheses, they are often listed separately.
3.2 Study design and sampling plan
Researchers typically describe the design of the study, such as an experiment, survey, cohort study, or field intervention. The sampling plan may include target population, recruitment strategy, expected sample size, and any randomization procedures.
3.3 Variables and measurements
Pre-registration often identifies the primary variables, outcome measures, predictors, and covariates. It may also specify how each variable will be operationalized, including scoring rules, instrument versions, and thresholds for defining categories.
3.4 Data collection procedures
The protocol commonly outlines how data will be gathered, when collection will occur, and what procedures will be used to maintain consistency. In experimental studies, this may include instructions, treatment delivery, and blinding procedures.
3.5 Statistical analysis plan
The analysis plan is often the most detailed portion of the registration. It explains how the data will be examined and how the main claims will be tested.
3.5.1 Primary analyses
Primary analyses are the main tests intended to answer the central research question. These may involve specific models, comparisons, or effect estimates. Stating them in advance reduces ambiguity about which results are the principal ones.
3.5.2 Secondary analyses
Secondary analyses address additional questions that are related to, but distinct from, the primary aim. They can include subgroup analyses, robustness checks, or supplementary comparisons. These analyses are often presented as supportive rather than definitive.
3.5.3 Exclusion criteria
Registrations commonly specify how data points, participants, or observations will be excluded, if at all. Predefining exclusion rules can reduce the risk that decisions are made after the researcher sees which cases help or hinder a preferred result.
3.5.4 Handling missing data
Because missing observations can affect results, many protocols state in advance how missing data will be treated. This may include complete-case analysis, imputation methods, or rules for determining whether a case remains eligible for analysis.
4 Common types of pre-registration
Different research settings use related but not identical forms of pre-registration. The underlying purpose remains advance specification, but the format may be tailored to the study’s goals and publication pathway.
4.1 Standard pre-registration
Standard pre-registration is a time-stamped record of hypotheses and methods submitted before data analysis. It is commonly used for one-off studies and may be public, embargoed, or restricted to certain reviewers until publication.
4.2 Registered reports
Registered reports combine pre-registration with journal review before results are known. In this format, the proposed study design and analysis plan are peer reviewed in advance, and journals may offer in-principle acceptance if the protocol meets standards. This can reduce publication bias against null findings.
4.3 Clinical trial registration
Clinical trial registration is a formal requirement in many medical contexts. It usually records trial objectives, eligibility criteria, interventions, outcomes, and sponsor information. The emphasis is on public accountability and the protection of participants as well as scientific transparency.
4.4 Exploratory pre-analysis plans
Some researchers prepare pre-analysis plans for exploratory or policy-related studies in which all outcomes cannot be fully anticipated. These plans still set out the intended framework for analysis, even if the research remains partly open-ended.
5 Pre-registration workflow
The process of pre-registering a study usually follows a sequence that begins before data analysis and ends with a dated, archived record. The exact steps depend on the platform and discipline.
5.1 Choosing a registry or platform
Researchers first select a suitable registry or platform. The choice may depend on field norms, whether the study is clinical or nonclinical, and whether the registration needs to be public, private, or embargoed.
5.2 Writing and submitting the protocol
Next, the investigator prepares the protocol or registration form, entering details about the hypotheses, sample, variables, and analysis plan. The submission should be completed before the relevant outcomes are examined, so the record can serve as an authentic pre-data commitment.
5.3 Time-stamping and version control
A reliable registry assigns a date and preserves the submitted version. Time-stamping helps establish precedence, while version control shows whether later edits were made and what those edits involved.
5.4 Public and embargoed registrations
Some registrations are immediately visible, while others remain hidden until a later date or until publication. Embargoed registrations can be useful when confidentiality is needed, but they still preserve the advance record within the system.
6 Benefits
Pre-registration is valued because it improves the interpretability of research findings and strengthens the documentation surrounding a study. Its advantages are strongest when the research question allows meaningful advance specification.
6.1 Improved research credibility
When readers can compare the final report with the original protocol, they can better assess the reliability of the conclusions. This often increases confidence that the main findings were not shaped by undisclosed choices made after looking at the data.
6.2 Less p-hacking and HARKing
Pre-registration can discourage p-hacking, the practice of trying many analyses until a statistically favorable result appears, and HARKing, or hypothesizing after the results are known. These practices can inflate apparent effects and make findings less trustworthy.
6.3 Better planning and documentation
Writing a study plan in advance can improve organization, clarify expectations for the research team, and reduce confusion later in the project. It also creates a structured record that may be useful during manuscript preparation or internal review.
6.4 Easier peer review and replication
Reviewers and other researchers can evaluate whether the analysis followed the intended design more easily when a pre-registration exists. Replication efforts also benefit from a clear description of the original procedures and outcomes.
7 Limitations and criticisms
Despite its advantages, pre-registration is not a complete solution to the problems of research bias or weak study design. It works best as one element in a broader framework of good research practice.
7.1 Reduced flexibility
Some critics note that a rigid protocol may make it harder to respond to unexpected data issues or emerging scientific insights. In practice, researchers can still adapt their methods, but they should explain deviations clearly and distinguish them from the original plan.
7.2 Incomplete or vague registrations
A registration that is too brief or nonspecific may fail to prevent selective interpretation. If the analysis plan leaves too many choices open, the value of pre-registration is reduced because the researcher retains broad discretion after seeing the data.
7.3 Practical barriers in some fields
In some disciplines, studies develop iteratively, data sources change rapidly, or field conditions are difficult to forecast. These circumstances can make detailed advance planning challenging. Even so, many researchers still use partial registrations or structured analysis plans where full pre-specification is not feasible.
7.4 Distinguishing preregistered and exploratory work
A common challenge is separating the registered analyses from later exploratory discoveries in the final report. Good reporting practice usually requires that both are labeled clearly so readers can see which claims were specified beforehand and which arose from subsequent examination of the data.
8 Applications across disciplines
Pre-registration is used in a range of empirical fields, although the content and strictness of registrations differ. The approach is most influential where statistical inference plays a major role.
8.1 Biomedical research
In biomedicine, pre-registration is especially important for clinical trials and intervention studies. It supports patient safety, outcome accountability, and comparison between the original protocol and the published report.
8.2 Psychology
Psychology has been a major driver of broader pre-registration adoption. Researchers often use it to separate predicted effects from post hoc interpretations and to improve the credibility of experimental and survey-based findings.
8.3 Economics
In economics, pre-analysis plans are often used in field experiments and policy evaluations. They help specify outcome variables, econometric models, and subgroup analyses before the data are examined.
8.4 Education
Educational research uses pre-registration to clarify intervention studies, classroom experiments, and evaluation designs. This can be especially helpful when multiple student outcomes and school-level factors might otherwise lead to flexible analysis choices.
8.5 Neuroscience
Neuroscience studies, particularly those involving imaging or behavioral tasks, may benefit from advance specification because data can be complex and analytically rich. Pre-registration helps researchers identify primary contrasts and reduce the temptation to interpret every pattern as meaningful.
9 Platforms and registries
A number of systems support pre-registration, each with its own format and intended use. Some are designed for clinical research, while others are built for general-purpose academic use.
9.1 ClinicalTrials.gov
ClinicalTrials.gov is a widely used registry for clinical studies. It records trial descriptions, eligibility criteria, outcomes, and other protocol details, making it a central platform for medical study registration.
9.2 OSF Registrations
OSF Registrations is a flexible platform for many kinds of research projects. It allows investigators to create time-stamped records of hypotheses, methods, and analysis plans, and it is widely used across disciplines.
9.3 AsPredicted
AsPredicted offers a streamlined format for concise pre-registration. It is designed for quick documentation of the key elements of a study and is often chosen when researchers want a simple, standardized form.
9.4 Discipline-specific registries
Some fields use specialized registries tailored to their own methods and norms. These may include repositories for clinical, behavioral, economic, or educational studies, each with its own required fields and review standards.
10 Related concepts
Pre-registration is part of a broader ecosystem of research practices that aim to improve rigor and openness. It is often discussed alongside several closely related ideas.
10.1 Registered reports
Registered reports are a publication format in which the study plan is reviewed before results are known. They extend pre-registration by linking the protocol to editorial evaluation.
10.2 Replication
Replication is the attempt to repeat a study or analysis to see whether similar results are obtained. Pre-registration can make replication easier by documenting the original research plan in advance.
10.3 Open data
Open data refers to making datasets available for inspection and reuse, subject to ethical and legal limits. It complements pre-registration by allowing others to examine the recorded outcomes and verify analyses.
10.4 Open materials
Open materials are the shared instruments, code, stimuli, or questionnaires used in a study. When combined with pre-registration, open materials help others understand exactly how the research was conducted.