Preregistration

Preregistration is the practice of recording hypotheses, methods, and analysis plans before doing or completing a study. Don’t hate the replicator, hate the game mentions it as one reform social scientists were already using to respond to the Replication Crisis.

The source treats preregistration as useful but incomplete. It can reduce P-Hacking by fixing some decisions before researchers see outcomes, but it does not replace [[ReplicationPackage|replication packages]], Robustness Checks, or institutions that make follow-up inspection credible.

Key Claims

• Preregistration separates planned tests from after-the-fact exploration.
• It can lower the temptation to tune analysis around a Statistical Significance Threshold.
• The practice is strongest when later readers can compare the preregistered plan with the actual analysis.
• It does not solve weak data documentation or coding errors by itself.
• Research Integrity Incentives matter because a preregistration only constrains behavior if deviations can be noticed.

Connections

• Replication Crisis and P-Hacking - problem context.
• Replication Package and Robustness Checks - complementary transparency and testing tools.
• Institute For Replication and Replication Games - source’s broader reform environment.
• Scientific Self-Correction - methodological value of making claims corrigible.