Funny Stuff (Chapter 14)
Once upon a time, a faculty member named Pat conducted a survey and gathered considerable amounts of data. Being quite busy with other projects, Pat duplicated the questionnaires and gave a complete set of the raw data to each of three students who wanted extra credit in Pat's statistics course. Pat's instruction to these students was short and simple: "Analyze these data on your own without talking with each other!" Well, as you might expect, the three students did not analyze the data in the same way. One student computed correlations among different questions. The second student compared men against women with a t-test. And the third student created confidence intervals around mean scores. After seeing what the students had done, Pat called the students together. He indicated that each one had successfully and accurately analyzed the data, and told each students that the desired extra credit had been earned. Pat had one question for the students, however. "I'm quite interested," said Pat, "in what accounts for the VARIANCE OF ANALYSIS."
What do statisticians get when they drink too much? (Answer: "Kurtosis of the liver.")
A husband and wife, both of whom were statisticians, had the misfortune of dying within a day of one another. They had always planned to be buried side by side. Unfortunately, the undertaker at the funeral home got them mixed up with another husband and wife who had expressed similar wishes.
This mistake became known as the first case of split-plot confounding.
Not all statisticians are the same. When asked what they want to drink, some of them choose t. Others prefer the hard stuff and drone on and on about box-of-whiskey plots.
Copyright © 2012
Schuyler W. Huck