Deterministic vs. probabilistic causation
- Q: Does smoking A (0/1) cause cancer B (0/1)?
- Deterministic causation (Wikipedia)
- “if A causes B, then A must always be followed by B”
- Probabilistic causation
- “A probabilistically causes B if A’s occurrence increases the probability of B” (Wikipedia).
- Q: Can you give me an example for both DC and PC?
- Q: What could that look like in a dataset?
- Deterministic: All individuals with Smoking = 1 have Cancer = 1.
- Probabilistic: Individuals with Smoking = 1 have higher likelihood of having Cancer = 1.
- Terminology
- Cause = Treatment (Q: Where does “treatment” come from?)
- Causal effect = Treatment effect