Dear Group,
The much awaited results are in. Here are the average grades plus/minus one
standard deviation:
Sample A (Grand Espresso): 6.8 +/- 0.8
Sample B (Super Crema): 4.6 +/- 2.4
Sample C (Gold Selection): 6.3 +/- 2.0
Sample D (Top Class): 9.0 +/- infinity (that's what happens when you don't
vote)
Using Welch's t-test, the pairwise two-sided p-values are (excluding sample
D, of course):
p_AB = 0.11
p_AC = 0.58
p_BC = 0.23
As good social scientists, we will blindly set the significance cutoff at p
< 0.05. Given the data, therefore, we can't reject the null hypothesis that
all the coffees are equally preferred. The coffee purchasing decisions will
therefore chiefly be guided by maximizing the value to the shareholders
(read: some of the varieties are on sale).
Ivan
PS. My apologies to Leslie, who correctly claimed that there was a fifth
kind of Lavazza coffee that we did not sample in this experiment. It is
Pienaroma Espresso, which I believe we previously decided tasted bad.
Show replies by date