The history of statistical hypothesis testing really began with a tea-tasting experiment (Fisher, 1935), so it seems fitting for this book to end with one. (This is also a fitting ending because Fisher was one of the first to advocate randomization procedures, though he did not have the facilities to use them for larger problems.) The owner of a small tearoom doesn’t think people really can tell the difference between the first cup made with a given tea bag and the second and third cups made with the same bag (which is why it is still a small tearoom). He chooses eight different brands of tea bags, makes three cups of tea with each, and then has a group of customers rate each cup on a 20-point scale (without knowing which cup is which). The data are shown here, with higher ratings indicating better tea:

I frequently ask you to use SPSS or R to solve a problem. You should have no problem with SPSS if you stick with the “Legacy” analyses. You should be able to use R by modifying the code that I give on the Web page for this chapter: http://www.uvm .edu/~dhowell/fundamentals9/Supplements/Chapter20R.html

Transcribed Image Text:

Cup Tea Brands First Second Third | 8 3 2 23 15 14 4 16 17 12 4 7 5 5 9 4 3 6 6 8 9 4 7 10 3 4 8 12 10 2 Using a randomization test, draw the appropriate conclusions.