In: Statistics and Probability
A publisher of books has produced seven comparable Statistical Management books with the following costs.
Quantity produced (000) |
1 | 2 | 4 | 5 | 7 | 9 | 13 |
Manufacturing Cost(000£) | 5 | 5.9 | 6.5 | 7.5 | 8 | 9.5 | 10.8 |
a) (5 pts) Construct the regression line for predicting manufacturing costs from quality produced.
b) (5 pts) Calculate the predicted values and the residuals of this data.
c) (5 pts) Construct a 99% two-sided confidence interval for the intercept.
d) (5 pts) Construct a 99% two-sided confidence interval for the slope.
e) (5 pts) Test for significance of regression by ANOVA and interpret your conclusion.
No | x | y | (x-xbar)^2 | (y-ybar)^2 | (x-xbar)*(y-ybar) |
1 | 1 | 5 | 23.591837 | 6.76 | 12.62857143 |
2 | 2 | 5.9 | 14.877551 | 2.89 | 6.557142857 |
3 | 4 | 6.5 | 3.4489796 | 1.21 | 2.042857143 |
4 | 5 | 7.5 | 0.7346939 | 0.01 | 0.085714286 |
5 | 7 | 8 | 1.3061224 | 0.16 | 0.457142857 |
6 | 9 | 9.5 | 9.877551 | 3.61 | 5.971428571 |
7 | 13 | 10.8 | 51.020408 | 10.24 | 22.85714286 |
sum | 41 | 53.2 | 104.85714 | 24.88 | 50.6 |
mean | 5.857142857 | 7.6 | sxx | syy | sxy |
calculation below using excell
slope = b1 = sxy/sxx = 0.482561308
intercept = b0 = ybar-(slope*xbar) = 4.773569482
SST = SYY = 24.88
SSR = sxy^2/sxx = 24.41760218
SSE = syy-sxy^2/sxx = 0.46239782
r^2 = SSR/SST = 0.981414879
r = sxy/sqrt(sxx*syy) = 0.990663858
error variance s^2 = SSE/(n-2) = 0.092479564
S^2b1 = s^2/sxx = 0.000881958
standard error b1 = se(b1) = sqrt(s^2b1) = 0.029697773
test statistics = b1/se(b1) = 16.24907413
For (1-alpha)% CI value of t = 0.99 = 4.032142984
Lower Confidence bound for slope = estimated slope -t*se(b1) =
0.362815643
Upper Confidence bound for slope = estimated slope +t*se(b1) =
0.602306973
standard error b0 = se(b0) = s*sqrt(1/n+ Xbar^2/Sxx) =
0.208489604
Lower Confidence bound for Intercept = estimated intercept
-t*se(b0) = 3.932909588
Lower Confidence bound for Intercept = estimated intercept
+t*se(b0) = 5.614229376
a) (5 pts) Construct the regression line for predicting manufacturing costs from quality produced.
slope = b1 = sxy/sxx = 0.482561308 =
intercept = b0 = ybar-(slope*xbar) = 4.773569482 =
y= 4.7736 + 0.4826*x
y=Manufacturing Cost
x=Quantity produced
Regression Model is
Manufacturing Cost =4.7736 + 0.4826*Quantity produced
=======================================================
b)
Observation | x | y | Predicted y | Residuals |
1 | 1 | 5 | 5.256131 | -0.25613 |
2 | 2 | 5.9 | 5.738692 | 0.161308 |
3 | 4 | 6.5 | 6.703815 | -0.20381 |
4 | 5 | 7.5 | 7.186376 | 0.313624 |
5 | 7 | 8 | 8.151499 | -0.1515 |
6 | 9 | 9.5 | 9.116621 | 0.383379 |
7 | 13 | 10.8 | 11.04687 | -0.24687 |
=======================================================
c) 99% two-sided confidence interval for the intercept.
intercept = b0 = ybar-(slope*xbar) = 4.7736
For (1-alpha)% CI value of t (1-alpha=0.99) = 4.0321
standard error b0 = se(b0) = s*sqrt(1/n+ Xbar^2/Sxx) =
0.208489604
Lower Confidence bound for Intercept = estimated intercept
-t*se(b0) = 3.9329
Lower Confidence bound for Intercept = estimated intercept
+t*se(b0) = 5.6142
99% two-sided confidence interval for the Intercept=(3.9329 ,5.6142)
=======================================================
d)99% two-sided confidence interval for the slope
slope = b1 = 0.4826
standard error b1 = se(b1) = sqrt(s^2b1) = 0.029697773
For (1-alpha)% CI value of t (1-alpha=0.99) = 4.0321
Lower Confidence bound for slope = estimated slope -t*se(b1) =
0.3628
Upper Confidence bound for slope = estimated slope +t*se(b1) =
0.6023
99% two-sided confidence interval for the slope =(0.3628 , 0.6023)
===============================================
e)
Test for significance of regression by ANOVA and interpret your conclusion.
ANOVA | |||||
df | SS | MS | F | Significance F | |
Regression | 1 | 24.41760218 | 24.4176 | 264.0324 | 1.61E-05 |
Residual | 5 | 0.46239782 | 0.09248 | ||
Total | 6 | 24.88 |
Null and alternative Hypothesis
H0: Model is not significant
H1:Overall model is significant
test statistics= F0= MS_Regression/MS_Residual =264.0324
p-value = p(F > F0)=0
Decision:P-value is less close to 0 hence we reject H0
Conclusion : Overall Model is significant
ANOVA | |||||
df | SS | MS | F | Significance F | |
Regression | 1 | 24.41760218 | 24.4176 | 264.0324 | 1.61E-05 |
Residual | 5 | 0.46239782 | 0.09248 | ||
Total | 6 | 24.88 |
================================================================
If you have any doubt please let me know through comment
Please give positive vote if you find this
solution helpful. Thank you