Answer Happy

Q2. Standardization of risk ratio Assume a binary treatment, A € 0,1, binary outcome, Y E 0,1 and a covariate L taking values in a finite set C). The following algebra shows how one can identify the causal risk ratio, pytan), with observational data using standard assumptions. However justification of PY=1 each step has been not given. Add short justifications for each of the steps indicating where you need to use particular causal assumptions. 2 P(Y! = 1) P(Y0 = 1) E(Y) because.... E(Y) E(E(Y'|L)) because.... E(E(YL)) E(E(Y'|A = 1, L)) because.... E(E(YA = 0, L)) E(E(Y|A = 1, L)) because.... E(E(Y|A = 0, L)) E(P(Y = 1 A = 1, L)) because.... E(P(Y = 1 A = 0, L)) Liec P(L = 1)P(Y = 1|A = 1, L = 1) because.... Liec P(L = 1)P(Y = 1 A = 0, L = 1) What extra assumption, is necessary to estimate the final expression in observational data?