Objective: Use experimental methods to design our own experiments Homework: Read section 6.1 and 6.2 (for vocab and examples) and then complete exercises: 6.19, 6.20, 6.21, 6.26, 6.27, 6.28, 6.33, 6.36 What is wrong with this experiment design? • A credit card bank wanted to test the sensitivity of the market to two factors: the annual fee charged for a card and the annual percentage rate charged. • The bank selected 100,000 people at random from a mailing list. It sent out 50000 offers with a low rate and no fee and 50000 offers with a higher rate and a$50 annual fee. The subjects signed up for the lower rate, no fee card at twice the rate of the other • The question the bank wanted to answer was: how much of the change was due to the rate and how much was due to the fee? • How could they have designed their experiment better? Lurking / Confounding? • Lurking variables can be thought of as a prior cause of both y and x that makes it appear that x may be causing y ( in observational studies and in regression). • A confounding variable is associated in a noncausal way with a factor and effects the response. We can’t tell whether the effect was caused by our factor or by the confounding variable ( in experiments). What is it? • There is found to be a positive association between ice cream sales and number of drownings. • This is a clear example of a lurking variable- the season (summer time) has both more ice cream sales and also more drownings. • A recent medical study found that among men who have had a heart attack, those with a higher level of a certain protein in their blood are at greater risk to have a second attack. • What can you conclude from this study? • Nothing much. Maybe the protein caused the second heart attack, maybe it was just something that occurred in both, or as a result of the heart attack. Smoking • A retrospective study that finds most people with lung cancer were smokers • A prospective study which follows a group of smokers and nonsmokers for several years • An experiment on 10 year olds… From Newsweek: • Of all the pre-college curricula, the highest level of mathematics one studies in secondary school has the strongest continuing influence on bachelor’s degree completion. Finishing a course beyond Algebra 2 more than doubles the odds that a student who enters post-secondary education will complete a bachelor’s degree. • Propose a study design to study that might enable you to draw this conclusion (at least a causal one) Ethical Concerns • Is it okay to use animals for experiments? Does it matter for what purpose? • What is “informed consent?” • Is it okay to drill placebo holes in patients’ skulls so that they and their doctors won’t know whether or not they received an experimental implantation of nerve cells as a possible treatment for Parkinson’s Disease? (this was actually done).