I just used this example in class, and I realized that I never shared it on my blog. I really love this example of Type II error (and some other stuff, too). So here it goes. http://www.achievement.org/autodoc/page/mar1int-1

A chocolate study seems very appropriate for the day after Easter. Messerli's study found a strong and positive correlation between a nation's per capita chocolate consumption and the number of Nobel prizes won by that nation (see graph below). The research article is a pretty straight forward: The only statistical analysis conducted was a correlation, the journal article is very short, and it used archival data. As such, you can use this example to illustrate correlation and archival data as well as the dread "third variable" problem (by asking students to generate variables that may increase chocolate consumption as well as top-notch research/writing/peace/etc.). Property of Messerli/New England Journal of Medicine