In: Statistics and Probability
A Ministry of Transportation investigation on driving speed and gasoline consumption for midsize vehicles (in miles per gallon), resulted in the following data:
Speed (Miles per Hour) 55 50 25 60 `25 30 55 40 50 30
Miles per Gallon 25 26 35 21 32 30 23 25 25 28
a) Covariance { {55,50,25,60,25,30,55,40,50,30} , {25,26,35,21,32,30,23,25,25,28} }
Let,
Speed (Miles per Hour) = x
Miles per Gallon = y
x |
y |
( x- x̄ ) = ( x – 42) |
(y - ȳ )= (y – 27) |
( x- x̄ ) (y - ȳ ) |
( x- x̄ )² |
(y - ȳ )² |
55 |
25 |
13 |
-2 |
-26 |
169 |
4 |
50 |
26 |
8 |
-1 |
-8 |
64 |
1 |
25 |
35 |
-17 |
8 |
-136 |
289 |
64 |
60 |
21 |
18 |
-6 |
-108 |
324 |
36 |
25 |
32 |
-17 |
5 |
-85 |
289 |
25 |
30 |
30 |
-12 |
3 |
-36 |
144 |
9 |
55 |
23 |
13 |
-4 |
-52 |
169 |
16 |
40 |
25 |
-2 |
-2 |
4 |
4 |
4 |
50 |
25 |
8 |
-2 |
-16 |
64 |
4 |
30 |
28 |
-12 |
1 |
-12 |
144 |
1 |
∑ x = 420 |
∑y = 270 |
∑(x- x̄ ) = 0 |
∑( y - ȳ) = 0 |
∑( x- x̄ ) (y - ȳ ) = - 475 |
∑ (x- x̄ )² = 1660 |
∑( y - ȳ)² = 164 |
Mean x̄ = ∑ x / n
= 420 / 10
x̄ = 42
Mean ȳ = ∑y / n
= 270 / 10
ȳ = 27
Population Cov (x, y) = ∑( x- x̄ ) (y - ȳ ) / n
= - 475 / 10
= -47.5
Population Cov (x, y) = -47.5
Sample Cov (x, y) = ∑( x- x̄ ) (y - ȳ ) / ( n – 1)
= - 475 / 9
= - 52.7778
Sample Cov (x, y) = - 52.7778
b. Correlation Coefficient r:
r = ∑( x- x̄ ) (y - ȳ ) / ( √ ∑ (x- x̄ )²) ( √∑( y - ȳ)² )
r = ( - 475 ) / (√1660)(√164)
r = - 0.9104
c. Since the value of r = - 0.9104, there is A strong negative linear relationship between Speed (Miles per Hour) and Miles per Gallon