Treat a field’s replication rate (percentage of independent, well‑powered replication attempts that reproduce original effects) as a formal metric of empirical credibility, reported by journals and funders. Embed this metric in grant review and policy citations so evidence used for regulation or large public programs must come from literatures with demonstrably high replication‑rate scores.
— Using replication rates as a governance metric would change how governments and institutions rely on social‑science findings, redirect funding to more robust research practices, and reduce policy built on fragile results.
Ethan Siegel
2026.09.15
78% relevant
The article warns that a solitary experimental 'blip' is insufficient to declare a discovery and stresses the need for reproduction and statistical robustness — directly connecting to the idea that replication rates (and replication standards) are the proper metric for scientific accountability and public confidence.
2015.01.04
100% relevant
The Open Science Collaboration (2015) produced a field‑level replication rate (~36% for 100 studies) and concrete effect‑size shrinkage, showing this metric is measurable and policy‑relevant.
← Back to all ideas