Correlation vs Causation: Confounding & Simpson's Paradox
This lesson is part of the course
In this lesson
You show that a coupon which looks harmful in the totals actually helps in every segment, and compute a segment-adjusted effect.
Buy the course once for $9.99 to unlock every lesson and lab, the final exam and the certificate – lifetime access, no subscription.
Lessons in this course
- Correlation vs Causation: Confounding & Simpson's Paradox Locked
- The Gold Standard: Randomised Experiment & Average Treatment Effect Locked
- Difference-in-Differences: The 2×2 Table and the Regression Locked
- Checking Parallel Trends: Pre-Period Slopes, a Placebo Test & a Trend Chart Locked
- Matching: Nearest-Neighbour Controls with MIN, MATCH & INDEX Locked
- Propensity Scores & Inverse Probability Weighting (IPW) Locked
- Regression Discontinuity: The Jump at the Gold-Status Cutoff Locked
- Instrumental Variables: The Wald Estimator from a Randomised Nudge Locked