What Is Pre-Registration?
Pre-registration is the practice of publicly documenting your research hypotheses, design, and analysis plan before collecting or analyzing data, typically by submitting the plan to an open registry like the Open Science Framework (OSF), AsPredicted, or ClinicalTrials.gov. The registered document becomes a timestamped, publicly accessible record that distinguishes planned analyses from exploratory ones conducted after seeing the results. Pre-registration doesn't prevent you from exploring your data, it just requires you to be honest about which analyses were planned in advance and which were data-driven. This transparency is a direct response to the replication crisis, which revealed that many published findings were shaped by questionable research practices like HARKing (Hypothesizing After Results are Known) and selective reporting. By committing to a plan before the data exist, pre-registration removes the temptation, and the opportunity, to unconsciously align your analyses with the results you hoped to find.
Why Pre-Registration Matters in Research
Without pre-registration, the line between confirmatory research (testing a pre-specified hypothesis) and exploratory research (discovering patterns in data) gets blurred, often unintentionally. A researcher who finds an unexpected result and then writes it up as if it were predicted all along isn't necessarily dishonest; they may genuinely believe the prediction was there from the start. Pre-registration protects against this hindsight bias by creating an external record. It also increases the credibility of significant findings, because readers know the analysis wasn't selected from many possibilities after peeking at the data.
How Pre-Registration Works
The process is straightforward but requires careful thought before you begin.
Decide What to Register
At minimum, a pre-registration should include your research questions or hypotheses, the study design (sample, conditions, measures), the primary outcome variables, the planned sample size and its justification (typically a power analysis), and the analysis plan (which statistical tests, which comparisons, how you'll handle exclusions and missing data). More detailed registrations also include secondary analyses, exploratory questions, and data quality checks.
Choose a Registry
The Open Science Framework (OSF) is the most widely used general-purpose registry. AsPredicted offers a streamlined template with fewer fields, suited for researchers who want simplicity. ClinicalTrials.gov is required for many clinical trials. Some journals maintain their own registries or accept registrations from any recognized platform.
Write the Pre-Registration
Be specific. "We will test whether the treatment group differs from the control group" is too vague. "We will compare mean scores on the Customer Effort Scale (CES) between the treatment and control groups using an independent-samples t-test, with alpha set at.05, two-tailed" is actionable. The level of detail should be enough that someone else could execute the analysis from your description alone.
Timestamp and Submit
Submit the registration before data collection begins (or before you access existing data you haven't analyzed yet). The platform generates a timestamped record with a unique URL. This URL goes in your manuscript so reviewers and readers can verify that your reported analyses match your registered plan.
Conduct the Study
Run the study as planned. When you analyze the data, clearly separate the registered (confirmatory) analyses from any additional (exploratory) analyses. Both can appear in the paper, the distinction is in how they're labeled and interpreted.
Report Deviations Transparently
If you deviated from the pre-registered plan (changed the primary outcome, used a different statistical test, excluded participants for reasons not specified in advance), report the deviations explicitly and explain why. Deviations aren't failures, they're expected when real-world research encounters complications. The requirement is transparency, not perfection.
When to Use Pre-Registration
- Hypothesis-testing studies. Anytime you're collecting data to test a specific prediction, pre-registration strengthens the inferential value of your results by confirming that the hypothesis preceded the data.
- Survey research with multiple variables. Surveys generate many possible analyses. Pre-registration helps you commit to a primary analysis plan and prevents selective reporting of whichever relationships happen to be significant.
- A/B testing and experiments. Pre-registration is particularly valuable in commercial research where there's pressure to find "winning" variants. Registering your primary metric, comparison plan, and stopping rule in advance protects against unconscious cherry-picking.
- Replication studies. When you're replicating someone else's finding, pre-registration demonstrates that your methods and analysis were fixed before data collection, making the replication more convincing regardless of whether it succeeds or fails.
- Multi-site or collaborative research. When multiple teams contribute data to a single analysis, pre-registration ensures everyone is working from the same protocol.
Common Mistakes to Avoid
- Being too vague. A pre-registration that says "We will analyze the data using appropriate statistical tests" doesn't constrain anything. The value of pre-registration comes from specificity, the more detailed the plan, the more credible the confirmatory results.
- Treating it as an all-or-nothing commitment. Pre-registration doesn't mean you can't explore your data or change course when circumstances demand it. It means you label exploratory analyses honestly and report deviations from the plan. Rigid adherence to a flawed plan isn't the goal, transparency is.
- Pre-registering after seeing the data. Some researchers have been caught pre-registering studies after data collection or analysis was already underway. This defeats the entire purpose. Registries timestamp submissions for exactly this reason.
How Quali-Fi Supports Pre-Registration
Quali-Fi's structured survey builder encourages the kind of upfront planning that pre-registration requires, defining measures, skip logic, and analysis variables before fielding begins. When your study design is locked in the platform, the gap between your pre-registered plan and your actual execution stays small, making transparent reporting straightforward.
Frequently Asked Questions
Does pre-registration stifle creativity?
No. Pre-registration separates confirmatory from exploratory analysis; it doesn't eliminate exploration. You can and should explore unexpected patterns in your data, you just label them as exploratory in your report. Many of the most interesting findings in research are serendipitous; pre-registration doesn't prevent them, it contextualizes them.
Is pre-registration required for publication?
Not universally, but many journals now encourage or require it, and some offer "registered reports" where the pre-registration is peer-reviewed before data collection. Funders like the NIH and the European Research Council increasingly expect registered protocols for funded research.
Can I pre-register qualitative research?
Yes, though the format looks different. Qualitative pre-registrations typically describe the research questions, the data collection approach, the sampling strategy, and the planned analytic method (e.g., thematic analysis, grounded theory), while acknowledging that qualitative analysis is inherently iterative.
Related Topics
Design studies that stand up to scrutiny. Start a free trial with Quali-Fi and build pre-registration-ready research from the start.