MOM'S SECRET ROOM

[ONEBGG]

A naive researcher approached a statistician one day about analyzing some data.
Researcher: "How do I test the difference between four treatment group means?"

Statistician: "Perform an Analysis of Variance."

Researcher: "But I don't want to test the difference in the group variances!"

Statistician: "You aren't! You are comparing the ratio of the variation between the group means to the combined variation within the groups to see if it is beyond chance."

Researcher: "You simply don't understand. You persist in talking about variation which does not interest me in the least!"

Statistician(Exasperated and Angry): "O.K. I have an alternative for you which is called the Interocular Test. Just examine any difference in the means and if it STRIKES YOU RIGHT BETWEEN THE EYES, declare it significant!!!"

*Isn't it rather ironic that the significance of the differences between a set of means can be tested by the ratio of two variances? Sir Ronald Fisher was very cagey when he perfected this seemingly contradictory procedure. This little story is my own so you know where to shoot the barbs.