ENV330H5 Lecture Notes - Lecture 10: Counterfactual Conditional, Margarine, Confounding
Document Summary
Understanding causal relationships is a main goal of scientific inquiry. Lurking variables: when x appears to influence y, but only because both x & y are influenced by z (no real relationship between x & y) Confounding variables: x influences y, but z also influences y (so it"s difficult to tell the relative importance of z and x on y) Collinear variables: z and x influence y, but z and x cannot be separated from each other (e. g. you cannot hold z constant while x varies) (spurious relationships completely random noise that looks like a signal) example margarine and divorce rate. 3 levels of causality: association, intervention, counterfactuals. P(y|x)- given y (independent), with a given x. What if (there hadn"t been pollution in the lake) Not just correlation is there a cause. P(y|(do)x)- can x do something to y. Wikipedia: a procedure carried out to support, refute, or validate a hypothesis.
