Achieving High Replicability: Psychology Protocol Boosts Success Rate to 86%
A prospective replication study involving four laboratories conducting social-behavioral research found a high replicability rate of 81% for self-confirmatory tests and 86% for replications. The average effect size (ES) of the self-confirmatory tests was smaller than the average ES of the published psychological literature. The replications had an average ES similar to the self-confirmatory tests and larger than previous attempts to replicate social-behavioral literature. The study suggests that high replicability is achievable when using rigour-enhancing practices such as preregistration, reporting all outcomes, large sample sizes, and commitment to high-fidelity replication procedures. The findings challenge the notion that low replicability and declining effects are inevitable in social-behavioral scientific investigation.