Scribd
Upload
77 views

Miss-Specification: Assignment No # 5

This document contains the results of several statistical tests performed to check the specification of an econometric model. The Jarque-Bera test indicates the residuals are not normally distributed, so the null hypothesis of normality is rejected. The Ramsey RESET test shows the model is misspecified. The Box-Cox test compares a log and linear functional form, finding the log form fits the data slightly better with a lower sum of squared residuals. Overall, the tests suggest the original model specification is incorrect.

Uploaded by

Anonymous Hx7MTi
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
77 views

Miss-Specification: Assignment No # 5

This document contains the results of several statistical tests performed to check the specification of an econometric model. The Jarque-Bera test indicates the residuals are not normally distributed, so the null hypothesis of normality is rejected. The Ramsey RESET test shows the model is misspecified. The Box-Cox test compares a log and linear functional form, finding the log form fits the data slightly better with a lower sum of squared residuals. Overall, the tests suggest the original model specification is incorrect.

Uploaded by

Anonymous Hx7MTi
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/ 4

MISS-SPECIFICATION

Assignment No # 5

ZOYA EHSAN
BECF14M013
Jarque-Berra Normality Test:
Graphically:
6
Series: RESID
Sample 1985Q1 1994Q2
5
Observations 38

4 Mean 1.28e-15
Median -0.001917
3 Maximum 0.106230
Minimum -0.067365
Std. Dev. 0.044987
2
Skewness 0.419918
Kurtosis 2.617587
1
Jarque-Bera 1.348308
0 Probability 0.509587
-0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10

From the graph of histogram it is see that residual is not normally distributed. We calculate the
chi_crit i-e 5.99 and Jarque-Bera vale i-e 1.35, which is less than chi_crit. So we reject H0 and
accept H1, it means that our model is miss-specified.

Ramsey RESET Test:

Dependent Variable: LCONS


Method: Least Squares
Date: 04/24/19 Time: 14:53
Sample: 1985Q1 1994Q2
Included observations: 38

Variable Coefficient Std. Error t-Statistic Prob.

C 2.717238 0.576652 4.712091 0.0000


LDISP 0.414366 0.126279 3.281340 0.0023

R-squared 0.230230 Mean dependent var 4.609274


Adjusted R-squared 0.208847 S.D. dependent var 0.051415
S.E. of regression 0.045732 Akaike info criterion -3.280845
Sum squared resid 0.075291 Schwarz criterion -3.194656
Log likelihood 64.33606 Hannan-Quinn criter. -3.250180
F-statistic 10.76719 Durbin-Watson stat 0.412845
Prob(F-statistic) 0.002301

Ramsey RESET Test


Equation: UNTITLED
Specification: LCONS C LDISP
Omitted Variables: Squares of fitted values

Value df Probability
t-statistic 4.663918 35 0.0000
F-statistic 21.75213 (1, 35) 0.0000
Likelihood ratio 18.36711 1 0.0000

F-test summary:
Mean
Sum of Sq. df Squares
Test SSR 0.028858 1 0.028858
Restricted SSR 0.075291 36 0.002091
Unrestricted SSR 0.046433 35 0.001327

LR test summary:
Value
Restricted LogL 64.33606
Unrestricted LogL 73.51961

Unrestricted Test Equation:


Dependent Variable: LCONS
Method: Least Squares
Date: 04/24/19 Time: 14:54
Sample: 1985Q1 1994Q2
Included observations: 38

Variable Coefficient Std. Error t-Statistic Prob.

C -204.0134 44.32789 -4.602370 0.0001


LDISP -204.4012 43.91503 -4.654471 0.0000
FITTED^2 53.74844 11.52431 4.663919 0.0000

R-squared 0.525270 Mean dependent var 4.609274


Adjusted R-squared 0.498142 S.D. dependent var 0.051415
S.E. of regression 0.036423 Akaike info criterion -3.711559
Sum squared resid 0.046433 Schwarz criterion -3.582275
Log likelihood 73.51961 Hannan-Quinn criter. -3.665561
F-statistic 19.36302 Durbin-Watson stat 0.795597
Prob(F-statistic) 0.000002

Here F-Statistics prob value is significant i-e is less than 5%. It means that we reject H0 and
accept H1. So our model is miss-specified.

Box-Cox Test:

Dependent Variable: CONSTAR


Method: Least Squares
Date: 04/24/19 Time: 16:14
Sample: 1985Q1 1994Q2
Included observations: 38

Variable Coefficient Std. Error t-Statistic Prob.

C 0.039783 0.000101 393.1375 0.0000


INCREAL 1.48E-05 9.88E-07 15.00171 0.0000

R-squared 0.862096 Mean dependent var 0.041296


Adjusted R-squared 0.858265 S.D. dependent var 0.000129
S.E. of regression 4.84E-05 Akaike info criterion -16.98119
Sum squared resid 8.45E-08 Schwarz criterion -16.89500
Log likelihood 324.6426 Hannan-Quinn criter. -16.95052
F-statistic 225.0513 Durbin-Watson stat 0.260483
Prob(F-statistic) 0.000000

Dependent Variable: CONSTAR


Method: Least Squares
Date: 04/24/19 Time: 16:16
Sample: 1985Q1 1994Q2
Included observations: 38

Variable Coefficient Std. Error t-Statistic Prob.

C 0.034367 0.000499 68.82024 0.0000


LINCR 0.001499 0.000108 13.87706 0.0000

R-squared 0.842501 Mean dependent var 0.041296


Adjusted R-squared 0.838126 S.D. dependent var 0.000129
S.E. of regression 5.18E-05 Akaike info criterion -16.84833
Sum squared resid 9.65E-08 Schwarz criterion -16.76214
Log likelihood 322.1182 Hannan-Quinn criter. -16.81766
F-statistic 192.5729 Durbin-Watson stat 0.236019
Prob(F-statistic) 0.000000

From the above results we can conclude both functional form are well, with R-squared .84-
84% and RSS1 is 8.45E-08 and RSS2 is 9.65E-08 with minor difference. The RSS1 is lower
than RSS2 so we can say log form is better than linear form.

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