ABSTRACT
Just like you have many different ways to estimate parameters, you also have many different ways to test hypotheses. For example, to test whether a student’s age is independent of their seat preference in a classroom you could use the median age of students in the front rows, minus the median age of students in the back rows as your test statistic. And if you think about it, there are lots of other test statistics you could use as well. Why not compare the maximum age in each group? And further, why does the F-statistic in Chapter 16, Equation 16.14, look like it does? Who decides these things anyway? How can you know what is the best test statistic to use?
