Professional Documents
Culture Documents
Eviews 4
Eviews 4
1
Variables (Relevant):
Consumption, investment, imports, stock building, or unemployment)
2
Database: Paul Turner’s Online Database
http://runbasichosting.com:8021/seaside/go/runbasicpersonal?app=database
3
Choose Variable & Country , and paste on
Excel
4
Log Transformation
Model Type Equation What is beta Interpretation of beta
Linear-Linear Unit change If X goes up by one unit Y goes up by units
Log-Log Elasticity If X goes up by one percent, Y goes up by
percent
Log-Linear Semi-elasticity If X goes up by one unit, Y goes up by
percent
Linear-Log Semi-elasticity If X goes up by one percent, Y goes up by
unit
5
Calculating rates of change i.e. inflation
rate
6
Assumption of the Classical Linear Regression
(BLUE)
1. The model is linear in its parameters and it is
correctly specified
2. Errors are not correlated with each other (i.e.
no autocorrelation)
3. The errors are homoscedastic (i.e. variance is
constant)
4. All independent variables are uncorrelated
with the error term
5. The error terms are normally distributed and
have a mean of zero
7
Diagnostic Tests
Assumption Test(s) Consequences of failure
Linear Functional Form Ramsey Reset Test Coefficient estimates biased
Errors have constant variance 1.White Test 1.Estimates are unbiased but inefficient
2.ARCH test 2.The standard errors of the coefficients are biased
Errors are serially independent 1.Durbin=Watson test 1.Estimates are unbiased but inefficient
2.Durbin’s h-test 2.The standard errors of the coefficients are biased
3.Breusch-Godfrey test
8
Ramsey Reset Test (Linear Functional
Form)
• To test for Functional form
• After run the regression
• View>stability
diagnostics>Ramsey Reset
test
• =Linear regression has
correct functional form
• P-value< 0.05 indicates
specification errors (for
any given statistics)
Book: Applied Econometrics with Eviews Applications, Research Gate
9
Ramsey Reset Output
How to deal:
10
Testing for Heteroscedasticity (White test)
• View-Residual Test-White Heteroscedasticity (no cross terms). If • Dealing with heteroscedasticity
you follow these steps then your equation window should look
like this: • Click on Estimate in the equation window and then
choose the Options button. You should see a
window like that below in which you should check
the box marked Heteroskedasticity consistent
coefficient covariance and choose the White option:
11
Testing for Heteroscedasticity (ARCH test)
12
Testing for non-normal residuals (Jarque-Bera test statistic )
13
Dealing with non-normality
• One very obvious possible cause for non-normality
is the large negative outlier in the distribution of
the residuals. If we graph the residuals then we can
see that this corresponds to observation number 32.
To see
• We can include a dummy variable to deal with this
outlying observation. First use the command line to
create a new variable:
• series d32 = 0
• Now edit d32 by typing data d32 on the command
line.
• If you now include this as an extra regressor in
your equation you should obtain better results- to
check this do repeat the test.
14
When running a linear regression:
When assessing the statistical fit of the model you should look for:
3. The significance of the estimated coefficients (t-ratios/ or p-vals)
4. The joint significance of the variables in the model (the F-statistic)
5. The proportion of the variance of the endogoneous variable which is explained by the model (the statistics)
15
Summary
1-Estimate the model.
Test the null hypothesis that the slope coefficient is equal to zero against the alternative that it is not equal to
zero
16