Hey stat 201:
I hope you get an A, so I am wondering more about your project. What is your hypothesis? It seems to me that you might be able to design a better study for your class than just asking a bunch of people their prs and mileage. The sample you get is going to have lots of confounding variables such as age, gender, and "talent." If you are trying to prove anything beyond that the correlation between mileage and time is weak, I suggest you retool your data set. For example, if your hypothesis is that higher mileage works, mileage suffers from diminishing returns, or theres a nonlinear relationship, I would think a time series would be better. I would get data on people SEASON bests and SEASON mileage, that way you negate the variables of gender and "talent." If you confine the data to people that are past maturity but not over the hill, you will go along way to limiting the effect of age as a variable. I think an ideal study would be to get the season bests and season mileage of runner 20-30 and then compare correlation. This would be an example:
Me:
age 18--14:40, 65 mpw
age 19--14:26, 75 mpw
age 20--14:06, 90 mpw
age 21--14:31, 70 mpw
age 22--14:15, 85 mpw
age 23--15:10, 45 mpw
age 24--15:50, 20 mpw
this is a very strong correlation and demonstrates the unremarkable position that more training and faster times are highly correlated.
Good luck.