Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
43 views

ISOM2500 Spring 2019 Assignment 4 Suggested Solution: Regression Statistics

This document provides a suggested solution to an assignment involving regression analysis. It includes regression models predicting the S&P 500 using the Dow Jones Industrial Average and predicting medical supply costs using hospital size. It finds significant relationships in both cases and provides confidence intervals, predictions, and other analyses of the regression models.

Uploaded by

Ching Yin Ho
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
43 views

ISOM2500 Spring 2019 Assignment 4 Suggested Solution: Regression Statistics

This document provides a suggested solution to an assignment involving regression analysis. It includes regression models predicting the S&P 500 using the Dow Jones Industrial Average and predicting medical supply costs using hospital size. It finds significant relationships in both cases and provides confidence intervals, predictions, and other analyses of the regression models.

Uploaded by

Ching Yin Ho
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

HKUST/ISOM2500/Spring2019

ISOM2500 Spring 2019


Assignment 4 Suggested Solution
1a)
1420
1400
1380
1360
S&P 500

1340
1320
1300
1280
1260
12200 12400 12600 12800 13000 13200 13400
DJIA

1b)
Regression Statistics
Multiple R 0.9740
R Square 0.9486
Adjusted R
Square 0.9446
Standard Error 9.6081
Observations 15

ANOVA
Significance
df SS MS F F
9.29184E-
Regression 1 22145.62926 22145.62926 239.8902 10
Residual 13 1200.1041 92.3157
Total 14 23345.73333

Standard
Coefficients Error t Stat P-value
Intercept -669.02124 130.7336 -5.1174 0.0002
9.29184E-
DJIA 0.15727 0.0102 15.4884 10

ŷ = -669.02124 + 0.15727 DJIA


1c)
Using the F test, the p-value corresponding to F = 239.8902 is .000. Because the p-
value < .05, we reject 𝐻! : 𝛽! = 0; there is a significant relationship.

1
HKUST/ISOM2500/Spring2019

1d)

The QQ plot suggests the residuals follow a normal distribution;


There is no obvious pattern or trend appear in residuals plot and hence the constant
variance assumption and linearity are valid;
The time plot of the residuals seems no obvious pattern and hence they are nearly
independent (it is a bit difficult to confirm independence assumption holds if the
pattern is not obvious and hypothesis testing may help here);

1e)
yˆ = −669.02124 + .15727(DJIA)= − 669.02124 + .15727(13,500) = 1454

1f) The DJIA is not that far beyond the range of the data. With the excellent fit
provided by the estimated regression equation, we should not be too concerned about
using the estimated regression equation to predict the S&P500.

2a) 𝜇!!!" = −4.28 + 0.254 30 = 3.34 days.


2b) 𝐻! : 𝛽 ≤ 0.2 𝑣𝑠 𝐻! : 𝛽 > 0.2
!.!"#$%!!.!
𝑡 = !.!"#$ = 1.887 > 𝑡!" 0.05 = 1.771 , so reject the null hypothesis.
2c) 0.25379 ± 2.16 0.0285 = (0.19223,0.31535)
! !".!"!!" !
2d) 3.34 ± 2.16 1.10807 !"
+ !"!!.!
= (2.55, 4.13)

2
HKUST/ISOM2500/Spring2019

3a)
Regression Analysis: Y versus X
The regression equation is
Y = 10.2 + 4.00 X

Predictor Coef SE Coef T P


Constant 10.2000 0.6633 15.38 0.000
X 4.0000 0.4690 8.53 0.000

S = 1.48324 R-Sq = 90.1% R-Sq(adj) = 88.9%

Analysis of Variance

Source DF SS MS F P
Regression 1 160.00 160.00 72.73 0.000
Residual Error 8 17.60 2.20
Total 9 177.60

Scatterplot of Y vs X
24

22

20

18

16
Y

14

12

10

0.0 0.5 1.0 1.5 2.0 2.5 3.0


X

b) 10.2+4(1)=14.2.

c) 4.

d) 𝑥 = 1 , 𝑦 = 14.2 ;
put x=1 into the regression line, then y=14.2, so the regression line passes through
𝑥, 𝑦 .

e) 95% C. I. ∶ 𝛽! ± t !!! n − 2 ∗ SE(β! ) = 4 ± 2.306 ∗ 0.469 = (2.918,5.082).


So we are 95% confidence that the slope is between 2.918 and 5.082.

3
HKUST/ISOM2500/Spring2019

f)
Predicted Values for New Observations
New Obs Fit SE Fit 99% CI 99% PI
1 18.200 0.663 (15.974, 20.426) (12.748, 23.652)
2 26.200 1.483 (21.223, 31.177) (19.162, 33.238)X
X denotes a point that is an outlier in the predictors.
Values of Predictors for New Observations
New Obs X
1 2.00
2 4.00
The interval (or standard error) of X=2 is much smaller than that of X=4. One
possible reason is that the original dataset does not have any information for X≥4 and
hence the estimation may introduce larger error.

g) 99% P.I. for X=2: (12.748, 23.652). The number of broken ampules falls within
12.748 and 23.652 with 99% probability/confidence.

h)
Analysis of Variance

Source DF SS MS F P
Regression 1 160.00 160.00 72.73 0.000
Residual Error 8 17.60 2.20
Total 9 177.60

𝐻! : 𝛽! = 0 𝑣𝑠 𝐻! : 𝛽! ≠ 0
Since F=72.73 and its p-value≈0 , so we reject the null hypothesis at 5% significant
level. Therefore, 𝛽! is significant different from 0.

i) 𝑅! = 0.901, 𝑟 = ± 𝑅! = 0.949 (take positive here because of positive slope). The


proportion of variation in Y is accounted for by introducing X into the regression
model is 90.1%.

You might also like