fference between the tests respectively with and without ‘cross terms’ in this case because there is only one regressor in the CIP model.
Click View/Residual Tests/White Heteroskedasticity (cross terms) on the equation toolbar:
White Heteroskedasticity Test:F-statistic3.970368 Prob. F(2,1144)0.019127Obs*R-squared7.906678 Prob. Chi-Square(2)0.019191Test Equation:Dependent Variable: RESID^2Method: Least SquaresDate: 02/04/07 Time: 21:27Sample: 5/10/2001 9/30/2005Included observations: 1147VariableCoefficientStd. Errort-StatisticProb. C-9.53E-067.48E-06-1.273280.2032UK_3MTBILLS-US_3MTBILLS0.0017870.0006352.8135870.005(UK_3MTBILLS-US_3MTBILLS)^2-0.0360160.013064-2.7569960.0059R-squared0.006893 Mean dependent var1.14E-05Adjusted R-squared0.005157 S.D. dependent var1.94E-05S.E. of regression1.94E-05 Akaike info criterion-18.86498Sum squared resid4.28E-07 Schwarz criterion-18.85178Log likelihood10822.06 F-statistic3.970368Durbin-Watson stat1.907598 Prob(F-statistic)0.019127
Q: What do you infer from the results of this test?
Q: What are the consequences of your findings for your previous OLS estimates of CIP?
ii.) Breusch-Godfrey LM test for autocorrelation
Click View/Residual Tests/Serial Correlation LM Test on the equation toolbar:
Breusch-Godfrey Serial Correlation LM Test:F-statistic80.44297 Prob. F(2,1143)0Obs*R-squared141.5279 Prob. Chi-Square(20Test Equation:Dependent Variable: RESIDMethod: Least SquaresDate: 02/04/07 Time: 21:29Sample: 5/10/2001 9/30/2005Included observations: 1147Presample missing value lagged residuals set to zero.VariableCoefficientStd. Errort-StatisticProb. C4.14E-060.0003960.0104780.9916UK_3MTBILLS-US_3MTBILLS-0.0001620.015795-0.0102370.9918RESID(-1)0.2077660.0287687.2221370RESID(-2)0.2325290.0287698.0826230R-squared0.12339 Mean dependent va-1.99E-18Adjusted R-squared0.121089 S.D. dependent var0.003373S.E. of regression0.003162 Akaike info criterion-8.671714Sum squared resid0.011428 Schwarz criterion-8.654121Log likelihood4977.228 F-statistic53.62865Durbin-Watson stat2.114047 Prob(F-statistic)0
Q: What do you infer from the results of this test?
Q: What are the consequences of your findings for your previous OLS estimates of CIP?
iii.) Ramsey’s RESET test for incorrect functional form
Click View/Stability Tests/Ramsey RESET Test on the equation toolbar:
Number of Fitted Terms: 2
Ramsey RESET Test:F-statistic31.97472 Prob. F(2,1143)0Log likelihood ratio62.44229 Prob. Chi-Square(20Test Equation:Dependent Variable: FP_3MMethod: Least SquaresDate: 02/04/07 Time: 21:32Sample: 5/10/2001 9/30/2005Included observations: 1147VariableCoefficientStd. Errort-StatisticProb. C-0.0012620.003222-0.3916010.6954UK_3MTBILLS-US_3MTBILLS0.5928630.2729552.1720190.0301FITTED^21.96327921.043780.0932950.9257FITTED^3389.9774463.14320.8420230.4R-squared0.748988 Mean dependent va0.01585Adjusted R-squared0.74833 S.D. dependent var0.006551S.E. of regression0.003287 Akaike info criterion-8.594461Sum squared resid0.012346 Schwarz criterion-8.576868Log likelihood4932.924 F-statistic1136.858Durbin-Watson stat1.537418 Prob(F-statistic)0
Including 2 fitted terms will test the null of a linear functional form against the alternative of a cubic functional form
The Eviews version of this test regresses the dependent variable on the original regressors and the fitted terms. Sometimes the test is carried out by regressing the model residuals on the original regressors and fitted terms. The two versions of the
本论文由英语论文网提供整理,提供论文代写,英语论文代写,代写论文,代写英语论文,代写留学生论文,代写英文论文,留学生论文代写相关核心关键词搜索。