Dealing with heteroscedasticity additional material 2
Download as DOCX, PDF0 likes408 views
The document provides steps to test for heteroscedasticity in regression analysis. It describes estimating a regression model using GDP, investment, and unemployment. Residual plots and tests, including Breusch-Pagan Godfrey and White tests, are used to check if the variance of the error terms increases with predicted values, which would indicate heteroscedasticity. Both tests find significant p-values, rejecting the null hypothesis of homoscedasticity and concluding heteroscedasticity is present.
Dealing with heteroscedasticity additional material 2
- 1. Dealing with heteroscedasticity Click Quick and select Estimate Equation. A dialog box will appear as below:- Type dependent variable (gdp) followed by constant (c) and independent variables (invest and unemp) in the space provided. Click ok and the results will be as below:-
- 2. Click View and select Actual, Fitted Residual and direct to Residual Graph The illustration of residual graph will be as below:-(Actual GDP-Predicted GDP)
- 3. Create a new series, residual3 that can capture the values of residuals Genr residual3=resid Click Quick and select Graph. A dialog box will appear as below: Type gdp followed by residuals3 in the provided space Click ok .A dialog box will appear as below: Select Scatter in the specific box and choose Regression Line as a type of Fit Line, and then click ok.
- 4. The illustration of scatter plot will be shown as below:- Click View and select Residual Tests, and then direct to Heteroskedasticity Tests. A dialog box on Heteroskedasticity Tests will appear as below:-
- 5. Select Breusch-Pagan Godfrey.Click ok and the results will be as below:- Null hypothesis : There is no heteroscedasticity (Homoscedasticity) Alternative hypothesis: There is a heteroscedasticity p-value=0.0020, p-value<0.05, Existence of heteroscedasticity Click View and select Residual Tests, and then direct to Heteroskedasticity Tests. A dialog box on Heteroskedasticity Tests will appear
- 6. Select White. Click Ok and the results will be as below:- Null hypothesis : There is no heteroscedasticity (Homoscedasticity) Alternative hypothesis: There is a heteroscedasticity p-value=0.00, p-value<0.05, Existence of heteroscedasticity