Free with sign-up

Power and sample size

An underpowered test can't see the effect you're hoping for — so it reports 'no difference' or, worse, a false win you'll act on. The intuition behind minimum detectable effect, run length, and why peeking lies.

This lesson is free — you just need an account to read it.

The first lesson of every course is open. Create a free account to continue the rest.