M31.2 CONNECT THE MECHANISM
State the assumptions behind effects that cannot all be observed
Ana got a coupon and spent $60. Would she have spent it anyway? You can never see both answers, but with the right design you can still measure the effect.
LESSON OVERVIEW17 min lesson
Lesson overview
Ana got a coupon and spent $60. Would she have spent it anyway? You can never see both answers, but with the right design you can still measure the effect.
What you’ll explore
- Define an effect with potential outcomes, compute an average treatment effect, name the assumptions (ignorability, positivity, SUTVA) that let observed data identify it, and use difference-in-differences or an instrument when randomization is impossible.
GO TO THE SOURCE
Original explanations, connected to the research.
Causal Inference: What If (Hernán & Robins)Estimating causal effects of treatments in randomized and nonrandomized studies (Rubin, 1974)Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania (Card & Krueger, NBER working paper, 1993; American Economic Review, 1994)The Book of Why (Pearl & Mackenzie, 2018)Causal inference in statistics: An overview (Pearl, 2009)Suggest a correction
A precise note can make an explanation better.
Choose the scene and describe what needs attention. Download a feedback file to share through a channel you already use. This page does not send feedback or connect you with a reviewer.