Suppose we discover a propitious natural experiment where Dr. Phil starts pumping iron at the Mackey Center in December, but not at Anytime Fitness. Dr. Phil befriends all the bros at the Mackey Center and writes testosterone prescriptions for every there. We are interested in the causal effect of testosterone on bench press performance. Suppose we are considering an academic year (September to May.) Describe how you would get a nonparametric difference-in-difference estimate. Be sure to talk about the: Treatment, pre-period, post-period, treatment group, control group, the first difference, the second difference, the DID estimate, the treatment effect and the bias. Draw a picture.