Effects of an Unusual Point. You are analyzing a data set of size \(n=100\). You have just performed a regression analysis using one predictor variable and notice that the residual for the 10th observation is unusually large.

a. Suppose that, in fact, it turns out that \(e_{10}=8 s\). What percentage of the error sum of squares, Error \(S S\), is due to the 10th observation?

b. Suppose that \(e_{10}=4 s\). What percentage of the error sum of squares, Error \(S S\), is due to the 10 th observation?

c. Suppose that you reduce the dataset to size \(n=20\). After running the regression, it turns out that we still have \(10=4 s\). What percentage of the error sum of squares, Error \(S S\), is due to the 10th observation?