Hypothesis-testing studies are either observational or experimental. In observational studies the epidemiologist observes without participating in the clinical process. In experimental studies, the epidemiologist plays an active role (usually by proxy) in the study. Observational studies have to be conducted when the epidemiologist cannot ethically or practically intervene (i.e., he/she cannot give someone a cancer or a heart attack; he has to wait for it to happen). Nor are people always willing to change their behaviors at someone else’s request (i.e., start or stop smoking or put on a lot of weight).