The Statistics Textbook Statistics

I have never actually read my statistics textbook, last year I definitely didn't read it. It's only this year I decided to study from it because I need to do better on the theory portion of the exam. I have learned some interesting things from the book, especially in regards to the chapter on correlation:

The following passages are taken from the book:
1) Figure 9.3 (pictured below) presents data from a 1991 article in Newsweek on the relationship between breast cancer and sunshine. For people like myself who love the sun, it is encouraging to find that there may be at least some benefit from additional sunlight - though that is probably a short-sighted interpretation of the data. Notice that as the amount of solar radiation increases, the incidence of deaths from breast cancer decreases. (There has been considerable research on this topic in recent years, and the reduction in rates of certain kinds of cancer is thought to be related to the body's production of vitamin D, which is increased by sunlight). This is a good illustration of a negative relationship, the correlation here is -.76.
- Fundamental Statistics for the Behavioral Sciences, David C. Howell


The solid blue line is the regression line, but for our purposes here it just indicates the relationship of the correlation. In this example the relationship is negative, which means the more amount of sunlight you are exposed to, the lower the breast cancer rate.

Personally, I don't know why he's so excited about this fact. He's not a chick, so he's not going to get breast cancer.

2)This next example is presented in Figure 9.4 (pictured below), and shows the relationship between the incidence of death due to heart attacks in various European countries and the consumption of wine.

Notice that for this example deaths due to heart disease actually decline with an increase in consumption of wine. This was originally a controversial finding, but there is now general agreement that it is a real (though not necessarily causal) effect. This might be taken by some as a license for university students to drink even more. However, heart disease is only rarely a problem for younger people, and no protection is needed. In addition alcohol has many negative effects that are not examined here. It doesn't seem sensible to increase the very real danger of alcohol abuse to ward off a problem that is highly unlikely to arise. Killing yourself by driving into a tree or developing liver problems is not a good way to reduce your risk of dying of heart disease.
- Fundamental Statistics for the Behavioral Sciences, David C. Howell


This is also another negative relationship, which means the more amount of wine consumed, the less incidence of heart disease.

You know what he feels like? A father lecturing you about good vs. bad conduct by using the simplest of logic. Also, he's pretty funny sometimes. It makes reading stats not so bad.


(bolded are sections which I laughed out loud while reading)