Sampling, estimation, confidence, hypothesis testing and the A/B test they add up to.
9h · 12 lessons · 1 challenge
Before this: Probability
Every model comparison is a statistical claim, and most of them are made badly. This module builds the machinery in order: sampling and the central limit theorem, point estimation and MAP, then confidence intervals, hypothesis testing, t-tests and power.
Power gets its own lesson rather than living inside the A/B lab, because the lab depends on it.
Run an A/B test and watch it every day. Stop when it looks significant. Then see how often that decision was wrong.
Worth 350 XP.
