An engineer wanted to determine how the weight of a car affects gas mileage. The accompanying data represent the weights of various domestic cars and their gas mileages in the city for a certain model year. Suppose that we add Car 12 to the original data. Car 12 weighs 3,05 pounds and gets 19 miles per gallon.
Complete parts (a) through (e).

Car Weight (lb) Miles per Gallon
1 4724 17
2 4006 18
3 3097 22
4 3555 19
5 4029 19
6 3934 19
7 3242 24
8 2960 26
9 3530 19
10 3823 18
11 2752 26

(a) Draw the scatter diagram with Car 12 included.
(b) Compute the linear correlation coefficient with Car 12 included.
(c) Compare the linear correlation coefficient of part (b) with the linear correlation coefficient for the original data.
Why are the results here? reasonable?

i. The correlation coefficient changed significantly when Car 12 was added. This is reasonable since Car 12 does not follow the pattern of the original data.
ii. The correlation coefficients both indicate a strong negative correlation. This is reasonable since Car 12 does not follow the pattern of the original data.
iii. The correlation coefficients both indicate a strong negative correlation. This is reasonable since Car 12 roughly follows the pattern of the original data.

(d) Now suppose that we add Car 13 (a hybrid car) to the original data (remove Car 12). Car 13 weighs 2,890 pounds and gets 60 miles per gallon. Draw the scatter diagram with Car 13 included.
(e) Compute the linear correlation coefficient with Car 13 included.