library(dagitty)Exercise 8
Instruction
There are four DAGs in the pictures attached to this exercise on Brightspace. For each DAG, identify (a) which path is a causal path from X to Y, (b) which variables need to be adjusted to satisfy the back-door criterion for the causal effect of X –> Y. Below is an example:
Q1: Top Left (Example)
Causal paths are X –> Y. The paths (i) X <– Z –> Y and (ii) X <– Z <– A –> Y are not causal. Conditioning on Z is sufficient to block all backdoor paths.
dag1 <- dagitty("dag{ X -> Y; X <- Z -> Y; X <- Z <- A -> Y }")
# The part below is optional; it just matches the appearance in the questions
coordinates(dag1) <- list(x = c(X = 0, Z = 1, A = 2, Y = 2),
y = c(X = 0, Z = -1, A = -1, Y = 0))
plot(dag1) # verify it's the same as the given graph
# Variables to be conditioned on
adjustmentSets(dag1, exposure = "X", outcome = "Y")