Read FAQ 5.2.2 about randomness and answer the following questions.
a. Suppose we get a small p-value in an observational study comparing two proportions with no random assignment. What can we rule out as an explanation for the difference in proportions? Can we conclude there is a cause? Can we conclude what that cause is?
b. Why does random assignment allow us to determine what the cause is in a study?
c. Can you have random assignment when you are testing a single proportion? Explain.