4-I really enjoyed reading your post. I agree that interaction is when one variable relies on the effects of another variable. I really like the example you gave about over weight men having higher cholesterol than overweight woman. This The variable being that if it is a man or woman that is over weight will affect how high the cholesterol is.
5-If I was comparing chantix, nicotine patch, and placebo these three variables would be appropriate for ANOVA versus a t-test that compares less variables.
6-An interaction is when one variable is dependent on another variable. This means that if one variable changes it can affect the outcome of another variable. There are many factors that go into my job working as an Occupational therapy assistant at a hand and shoulder rehabilitation center. First the doctor that preforms surgery with efficiency. As therapists we need to do our jobs and follow protocol, to get the body part functioning properly again. The patient needs to do their job and complete their home exercise program as instructed by the therapist. All these variables will assist in allowing the patient to have the best possible outcome.