Preregistration refers to the practice of depositing a detailed description of a study plan — including the research question, hypotheses, study design, data collection methods, and analysis plan — in a publicly accessible repository or platform before data collection begins. Doing so can encourage researchers to commit to a study plan in advance and avoid making changes to the study methodology or reporting based on the results that they obtain.

What are the benefits of preregistration?

Preregistering a study benefits the scientific community and integrity of research in a number of ways:

• It compels the researcher to formulate a study rationale for a specific research question, which helps to distinguish between hypothesis-testing (confirmatory) and hypothesis-generating (exploratory) research. Confirmatory analyses are planned in advance in order to retain the validity of their statistical inferences, and exploratory analyses are reported as post hoc investigations that might inspire confirmatory tests in future studies.
• It may mitigate publication bias (which occurs when studies with statistically significant results are more likely to be published than studies with nonsignificant results) because not publishing studies due to nonsignificant findings would result in a public record of unpublished studies, which may reflect poorly on the researcher.
• It reduces selective reporting, in which the researchers selectively report results that support a specific hypothesis or align with desired outcomes.
• It helps prevent “HARKing” (hypothesizing after the results are known), in which the researchers form or modify hypotheses based on the results that they obtain.

Is preregistration foolproof?

Although preregistration can promote research integrity by encouraging transparency, fostering accountability, and reducing questionable research practices, it’s certainly not foolproof. Here are some of its limitations:

• Preregistration is typically a voluntary process, which means that many researchers choose not to preregister their work.
• Even when researchers do preregister their work, they can still deviate from their initial study plans (with or without making public and trackable changes to the preregistration records that reflect the deviations from the initial study plans). That said, such deviations without sufficient justification can provide valuable information about the study quality.
• Researchers may not provide complete information in their preregistration entries, leaving “wiggle room” for making adjustments to the study plan along the way. For example, although a researcher may list “muscle strength” as a study outcome, they may not specify the muscle group for which strength will be examined, the method that will be used to assess muscle strength, or the time points at which this outcome will be measured.
• Although preregistration may encourage researchers to publish studies even when the findings are nonsignificant, publication bias can still arise if the researchers choose not to publish the results of their studies or if studies with nonsignificant results are less likely to be accepted for publication by scientific journals.