Sometimes it is instructive to generate random data that have a given relationship and then use regression on the generated data to see if the estimated regression is basically the same as that used to generate the data in the first place.

a. Let X be normally distributed with mean 100 and standard deviation 10, and let Y equal 100 + 5X plus a normally distributed random residual with mean 0 and standard deviation 40. Generate 50 ­observations.

b. Copy the random data and paste as values in another range of the worksheet. Then run the regression on the copy (the frozen values).

c. Repeat part b several times to see if you get basically the same regression results each time. That is, each run should be on a different set of frozen data.

d. Write a short report of your method and your findings.