In: Statistics and Probability
Date | DOW | RUSSELL | |||||
8-Apr-20 | 23,719.37 | 1,246.73 | |||||
1-Apr-20 | 22,653.86 | 1,139.17 | |||||
25-Mar-20 | 21,917.16 | 1,153.10 | |||||
18-Mar-20 | 20,704.91 | 1,096.54 | |||||
11-Mar-20 | 21,237.38 | 1,106.51 | |||||
4-Mar-20 | 25,018.16 | 1,350.90 | |||||
26-Feb-20 | 25,917.41 | 1,486.08 | |||||
19-Feb-20 | 27,081.36 | 1,571.90 | |||||
12-Feb-20 | 29,232.19 | 1,683.52 | |||||
5-Feb-20 | 29,276.34 | 1,677.51 | |||||
29-Jan-20 | 28,807.63 | 1,656.77 | |||||
22-Jan-20 | 28,722.85 | 1,658.31 | |||||
15-Jan-20 | 29,196.04 | 1,685.90 | |||||
8-Jan-20 | 28,939.67 | 1,675.74 | |||||
1-Jan-20 | 28,583.68 | 1,658.31 |
a. Create a scatter plot. Say what you see.
b. Use the data to develop an estimated regression
equation showing how your team data is related to DOW, the Dow
Jones industrial average. What is the estimated regression model
(y-mx+b, slope & intercept)?
Let x represent the DOW indexes.
c. How much of the variation in the sample values of
your team data does the model estimated in part (b) explain?
Round your answer to two decimal places.
d. Suppose that the closing price for the DOW is 29,000. Estimate the closing price for your data
e. Preform a hypothesis test for the model (F test) with an significance of 0.05. State your conclusion
Ʃx = | 391008.01 |
Ʃy = | 21846.99 |
Ʃxy = | 580536532.1717 |
Ʃx² = | 10340354614.3419 |
Ʃy² = | 32656195.8231 |
Sample size, n = | 15 |
x̅ = Ʃx/n = 391008.01/15 = | 26067.20067 |
y̅ = Ʃy/n = 21846.99/15 = | 1456.466 |
SSxx = Ʃx² - (Ʃx)²/n = 10340354614.3419 - (391008.01)²/15 = | 147870355.4 |
SSyy = Ʃy² - (Ʃy)²/n = 32656195.8231 - (21846.99)²/15 = | 836797.6858 |
SSxy = Ʃxy - (Ʃx)(Ʃy)/n = 580536532.1717 - (391008.01)(21846.99)/15 = | 11046659.88 |
a) Scatter plot:
b)
Slope, b = SSxy/SSxx = 11046659.87904/147870355.3979 = 0.074705
y-intercept, a = y̅ -b* x̅ = 1456.466 - (0.07471)*26067.20067 = -490.8851
Regression equation :
ŷ = (0.0747) x - 490.8851
c)
Coefficient of determination, r² = (SSxy)²/(SSxx*SSyy)
= (11046659.87904)²/(147870355.3979*836797.68576) = 0.9862
98.62% variation in y is explained by the least squares model.
d)
Predicted value of y at x = 29000
ŷ = -490.8851 + (0.0747) * 29000 = 1675.5609
e)
Null and alternative hypothesis:
Ho: β₁ = 0 ; Ha: β₁ ≠ 0
SSE = SSyy -SSxy²/SSxx = 836797.68576 - (11046659.87904)²/147870355.3979 = 11556.5875
SSR = SSxy²/SSxx = (11046659.87904)²/147870355.3979 = 825241.0982
Test statistic:
F = SSR/(SSE/(n-2)) = 825241.0982/(11556.5875/13) = 928.3133
Critical value F crit = F.INV.RT(0.05, 1, 13) = 4.667
p-value = F.DIST.RT(928.3133, 1, 13) = 0.0000
Conclusion:
p-value < α Reject the null hypothesis.