mauigold6457 mauigold6457 Today at 1:49 AM Mathematics Answered suppose you have a binary randomized controlled experiment designed to measure the causal effect of x on y. let the control group be identified by x