Question: Professor Quagmire tells one of his graduate students that he must “prove” that students in his college (“Own”) have higher IQ’s than those in the neighboring college of one of the professor’s archenemies (“Other”). The graduate student manages to get IQ values for 10 students in each college and calculates a T test.

a. Has he succeeded in meeting the professor’s demand?

He then goes out and gets 400 IQ’s for each college and calculates a T test.

b. Has he met the professor’s demand?

c. What do you think accounts for this difference (from the previous analysis)?

d. Do you think the IQ difference between the 2 colleges is “practically” significant?

e. Do you think that “statistically significant” always means “practically significant”?

f. How does this problem relate to the commonly held (at least for statisticians) belief that you can make anything statistically significant if you try hard enough? How could this be done?