Appendices A and B in the text summarize descriptive and inferential statistics. Recently, NPR ran a story about this summer's extremely high temperatures. I love how it describes the standard deviation as a "unit of wierdness." And I also like how they describe rare events--statistically significant events--as "a lot of snake eyes in a row." Take a look:
http://www.npr.org/2012/08/06/158215252/are-recent-heat-waves-a-result-of-climate-change
This article is simply a great example of how to reason statistically.
While the artilce is commendable from the standpoint of discussing statistical concepts, I was concerned that the author equated the probability of being one standard deviation away from the mean with throwing snake eyes (two 1's with a pair of dice). The probability of snake eyes is 1/6 * 1/6 = 1/36. Doing that three times in a row is (1/36)*(1/36)*(1/36) or .00002 which is far more extreme than an observation being one standard deviation away from the mean. We should be careful about interpreting statictical results in a way that needlessly overstates the case.
Posted by: Don Robertson | 08/09/2012 at 09:50 AM