Hi, My question is regarding to the last part of this chapter. I've seen in some textbooks that level of significance is divided by number of different possible pairing of two samples to compensate for performing multiple tests instead of one test. Wonder why this isn't done in the notes?
I suspect it's for simplicity. Also you'll find that different statisticians do things in different ways and there isn't always agreement. For example there are five definitions for calculating the quartiles...
