Product & Analytics / Experimentation
Minimum Detectable Effect MDE (with Statistical Power, 1 - beta)
The smallest true improvement a test is big enough to reliably catch.
The fork, why two teams get different numbers
MDE is quoted two incompatible ways.
In the Metric Library
- The full fork, both definitions worked all the way through
- The trap that makes the number lie, on a real export
- Every formula variant, spelled out
- The reconciliation anchor, what to tie it to and when to refuse
Reference: Neyman-Pearson power analysis convention; experimentation power-calculation conventions