Pre-Registration
The practice of publicly registering a study's hypotheses, methods, and analysis plan before data collection, to prevent HARKing (Hypothesizing After Results are Known) and selective reporting.
What it means
Pre-registration is the practice of publicly documenting a study’s research questions, hypotheses, data collection procedures, and planned statistical analyses before data collection begins. The registered plan is time-stamped and publicly accessible (typically via OSF, ClinicalTrials.gov, or AsPredicted), creating an auditable record of what was decided in advance versus what was decided after seeing the results.
Pre-registration is standard practice in clinical trials (ClinicalTrials.gov registration is required for publication in most medical journals) and is increasingly used in psychology, economics, and other empirical disciplines.
A Registered Report is a stronger form: the journal peer-reviews the introduction and methods before data collection and agrees to publish the results regardless of their direction, eliminating the publication pressure that can drive outcome-switching.
Why it matters for researchers
The replication crisis context: Pre-registration emerged as a response to widespread problems in scientific publishing — particularly p-hacking (trying multiple analyses until something is significant) and HARKing (presenting exploratory findings as if they were predicted in advance). Both inflate false-positive rates in the published literature.
How AI tools interact with pre-registration:
- AI tools do not reduce or increase the need for pre-registration — whether an analysis was planned in advance is a study-design question, not a computational one
- AI-assisted data analysis (Julius AI, ChatGPT code generation) runs the same risk of undisclosed exploratory analysis as manual analysis — the registered plan still constrains which results you can report as confirmatory
- The OSF platform is the most common venue for pre-registration outside of clinical trials and is free for academic use
Common misunderstanding: Pre-registration doesn’t prevent exploratory analysis — it just requires that exploratory findings be clearly labeled as such in the paper, rather than presented as the outcome of a pre-planned test.