【正文】 og likelihoodAkaike info criterionSum squared resid89040158. dependent var. of regressionMean dependent varAdjusted RsquaredCX4RsquaredDurbinWatson statProb(Fstatistic)圖8Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 23:13Sample: 1983 1995Included observations: 13VariableCoefficientStd. ErrortStatisticProb.HannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid+08. dependent var. of regressionMean dependent varAdjusted RsquaredCX3RsquaredDurbinWatson statProb(Fstatistic)圖7Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 23:13Sample: 1983 1995Included observations: 13VariableCoefficientStd. ErrortStatisticProb.HannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid13369621. dependent var. of regressionMean dependent varAdjusted RsquaredCX2RsquaredDurbinWatson statProb(Fstatistic)圖6Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 23:12Sample: 1983 1995Included observations: 13VariableCoefficientStd. ErrortStatisticProb.HannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid6540962.. dependent var. of regressionMean dependent varAdjusted RsquaredCX2X3X4X5X6RsquaredDurbinWatson statProb(Fstatistic)圖5Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 22:28Sample: 1983 1995Included observations: 13VariableCoefficientStd. ErrortStatisticProb.HannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid. dependent var. of regressionMean dependent varAdjusted RsquaredCZRsquaredDurbinWatson statProb(Fstatistic)圖4Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 21:52Sample: 1 14Included observations: 14VariableCoefficientStd. ErrortStatisticProb.HannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid. dependent var. of regressionMean dependent varAdjusted RsquaredCX2X3X4RsquaredProb(Fstatistic)圖3Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 21:39Sample: 1 14Included observations: 14VariableCoefficientStd. ErrortStatisticProb.FstatisticDurbinWatson statSchwarz criterionLog likelihoodAkaike info criterionSum squared resid. dependent var. of regressionMean dependent varAdjusted RsquaredCX2RsquaredProb(Fstatistic)圖2 Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 20:40Sample: 1 10Included observations: 10VariableCoefficientStd. ErrortStatisticProb.FstatisticDurbinWatson statSchwarz criterionLog likelihoodAkaike info criterionSum squared resid. dependent var. of regressionMean dependent varAdjusted RsquaredCX2X3Rsquared附錄 圖1Dependent Variable: YMethod: Least SquaresDate: 12/20/09 Time: 20:25Sample: 1 10Included observations: 10VariableCoefficientStd. ErrortStatisticProb.修正得到如下結果（輸出結果見附錄圖34）： t值 結果分析：雖然加權之后回歸擬合效果提高了3%，但是，必須看到，這里做的修正是在降低顯著性水平條件下進行的，即只是一個練習操作而已，沒有實質意義。這時如果要修正模型，可采用WLS法。DurbinWatson statProb(Fstatistic)結果分析，兩種檢驗，（95%置信水平），均不能拒絕無異方差性的假設。HannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid. dependent var. of regressionMean dependent varAdjusted RsquaredCX3^2X3*X4X3*X5X4^2X4*X5X5^2RsquaredProb. ChiSquare(6)Test Equation:Dependent Variable: RESID^2Method: Least SquaresDate: 12/21/09 Time: 15:47Sample: 1 12Included observations: 12VariableCoefficientStd. ErrortStatisticProb.Prob. ChiSquare(6)Scaled explained SSProb. F(6,5)Obs*RsquaredDurbinWatson statProb(Fstatistic)圖9 white檢驗結果Heteroskedasticity Test: WhiteFstatisticHannanQuinn criter.FstatisticSchwarz criterionLog likelihoodAkaike info criterionSum squared resid. dependent var. of regressionMean dependent varAdjusted RsquaredCX3X4X5