- History: 192? (Neyman, Wright)
- causal relationalship: direction is unclear
- Causality using an example, Causal effect of treatment (A = 1 or 1; doesnt have to be binary) on some outcome (Y = 0 or 1 or time until death)
- Potential outcome:
$Y^A$ . For example,$Y^1$ = 1. This means the outcome (disease) under treatment (A = 1). This could be$Y^(A=0)$ = 0; Outcome under no treatment is no disease
- Observed outcome is Y = Y^1 ; counterfactual outcome is
$Y^0$