T | S | T*S | T2 | |
---|---|---|---|---|
2.6 | 95 | 247 | 6.76 | |
3.7 | 140 | 518 | 13.69 | |
2.4 | 85 | 204 | 5.76 | |
4.5 | 180 | 810 | 20.25 | |
2.6 | 100 | 260 | 6.76 | |
5.0 | 195 | 975 | 25.00 | |
2.8 | 115 | 322 | 7.84 | |
3.0 | 136 | 408 | 9.00 | |
4.0 | 175 | 700 | 16.00 | |
3.4 | 150 | 510 | 11.56 | |
Total | 34.0 | 1371 | 4954 | 122.6 |
b= (4954 - 10 * 3.4 * 137.1)/(122.6 - 10 * 3.42) =
41.6809
a= 137.1 - 41.6809 * 3.4 = - 4.6151
The estimated regression line is then S = -4.6151 + 41.6809 * T
The easiest way to calculate the coefficient of determination is to
use the computational formula:
All the quantities needed in this formula with the exception of the "sum of squared S" is already on hand from the solution of 21. Sum of squared-S (first square the S values then sum) is 201,121.
r2 = (-4.6151*1371 + 41.6809 * 4954 - 10 * 137.1 * 137.1)/(201,121 - 10 * 137.1* 137.1) = 0.92695
The Sample Correlation Coefficien is the square-root or r =
0.96278
To Top
t = (b - B) / sb = 1.12/.0470606 = 2.37991
Critical t = 2.069 ( two-tail with 23 df)
Thus reject the Null
Alternatively p-value = .02 (From Excel function tdist (2.37991,
23, 2) )
leading to the same conclusion
Click for
Excel output for part a) and part b
To Top