Urgent.News

What's breaking now, across thousands of outlets.

More in Finance & Markets

Are RCTs underpowered?

Maybe so: We study the statistical credibility of randomized controlled trials (RCTs) in development economics and the optimal allocation of resources to replication.

  • Only 30% of published rejections are actual false discoveries
  • Average directional power of RCTs at standard significance levels is 40%
  • Doubling sample sizes reduces average false discovery rate by 5 percentage points

More from Tuesday 10 November →