Scribd
Upload
81 views5 pages

TSA Assignment Number 14

1) The residuals from fitting a regression model to the time series data show strong positive autocorrelation at lag 1, indicating the presence of autocorrelation. 2) The Durbin-Watson statistic of 0.0504316 is less than the lower limit value of 1.20, confirming the presence of positive autocorrelation in the residuals. 3) Estimating the regression using generalized differences to account for autocorrelation produces a regression coefficient estimate of β1 = 0.00303 with a smaller standard error compared to the original model, showing it is a more accurate approach.

Uploaded by

Musa Khan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
81 views5 pages

TSA Assignment Number 14

1) The residuals from fitting a regression model to the time series data show strong positive autocorrelation at lag 1, indicating the presence of autocorrelation. 2) The Durbin-Watson statistic of 0.0504316 is less than the lower limit value of 1.20, confirming the presence of positive autocorrelation in the residuals. 3) Estimating the regression using generalized differences to account for autocorrelation produces a regression coefficient estimate of β1 = 0.00303 with a smaller standard error compared to the original model, showing it is a more accurate approach.

Uploaded by

Musa Khan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

Lahore School of Economics

BBA III

Time Series Analysis and Forecasting Spring 2021

Name:____M.Nauman Ahmed____ Section:___F________

Date:______________________ Score:________________

Assignment No. 14

The data given below show seasonally adjusted quarterly sales for Dickson Corporation (Y) and for the entire industry (X)
for 20 quarters.

a. Fit a regression model, and store the residuals. Plot the residuals against time, and obtain the residual
autocorrelation. What do you find?
b. Calculate the Durbin-Watson statistic and determine whether autocorrelation exists.
c. Estimate the regression coefficient, 1, using generalized differences. (Estimate with the lag 1 residual
autocorrelation.)
d. Compare the standard errors of the two estimates of 1 obtained using the original and the generalized differences.
Which estimate is more accurate? Explain.
e. Convert both X and Y to simple differences. That is, create the numbers

Y ,t =Y t −Y t−1 and X t, = X t− X t−1

Fit a simple regression model to the differenced data. Compare your results to the results
obtained by the method of generalized differences. Did you expect them to be much different?

Dickson Industry
sales Sales
(1,000) (millions)

Y X
83.8 31.8
85.6 32.5
87.8 33.2
86.1 32.4
89.6 33.8
91 34.3
93.9 35.3
94.6 35.7
96.4 364
96 36.3
98.2 37.1
97.2 36.6
100.1 37.6
102.6 38.3
105.4 39.3
107.9 40.2
110.1 41.1
111.1 41.4
110.1 41.1
111.1 41.4
Answers:

(a)

Regression Equation

Y = 97.90
+ 0.0005 X

Model Summary

S R-sq R-sq(adj) R-sq(pred)

9.3407 0.00% 0.00% 0.00%


0

Analysis of Variance

Source DF Adj SS Adj MS F-Value P-Value

Regression 1 0.03 0.0265 0.00 0.986

 X 1 0.03 0.0265 0.00 0.986

Error 18 1570.4 87.2486    


8

  Lack-of- 16 1570.4 98.1547 * *


Fit 8

  Pure Error 2 0.00 0.0000    

Total 19 1570.5      
0

Durbin-Watson Statistic

Durbin-Watson 0.0504316
Statistic =
Autocorrelations

Lag ACF T LBQ

1 0.85604 3.83 16.97


4

2 0.71924 2.05 29.62


1

3 0.56359 1.35 37.84


1

4 0.38469 0.85 41.91


6

5 0.23332 0.50 43.50


7
From the regression equation we can see that the value of the regression coefficients are β 0 = 97.90 and β1 = 0.0005. Also
looking at the autocorrelation plot we can see that the first lag has a spike in it meaning that it is significantly different
from 0 with r1 = 0.856044.

(b)

N=20, k=1, α = 0.05

DL = 1.20, DU = 1.41

Since DL value of 1.20 is greater than output value of Durbin-Watson statistic = 0.0504316 we can conclude that there is
positive autocorrelation among the residuals at the 5% level.

(c)

Regression Equation

GD(y = 15.410
) + 0.00303 GD(X)

By this regression equation we can see that the regression coefficient β 1 = 0.00303.

(d)

From the minitab output, the standard error of estimate of β 1 of general model is 0.0293 and for transformed model β1 is
0.00428. Hence, we can see that the standard error of estimate of the transformed model is less than that of the general
model. Therefore, the transformed model is by using generalized differences is the better model.

(e)

Regression Equation

Dy = 1.435
+ 0.00342 Dx
Durbin-Watson Statistic

Durbin-Watson 2.38887
Statistic =

From this we can see that the durbin Watson statistic from the minitab output comes out to be 2.38887 which is greater
than the DL value and is also greater than 2 meaning that negative autocorrelation is present in this model.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy