1 Defining counterfactuals
1.1 Observing versus intervening
Aug 26. Slides After class, install R and Rstudio on your computer (see slide 17 from today’s lecture).
Statistical inference is about observing: if I observe a sample from a population, what can I infer about that population? Causal inference is about intervening: if I intervene to change some exposure, what average outcome would result?
Today we will discuss observing, intervening, and why the difference is so important.
1.2 Lab: Statistics review
In this lab, we will start by reviewing some basic statistical (random variables, expectation, conditional expectation, etc) and programming concepts.
1.3 Defining causal effects
Aug 28. Slides. After class, read Chapter 1 of Hernán and Robins 2020 and begin Problem Set 1.
Today we will define average causal effects in the potential outcomes framework.
By the end of class, you will be able to
- define potential outcomes
- explain the Fundamental Problem of Causal Inference1