回归结果为表4-9所示。Р 表4-9 运行结果(三)РDependent Variable: YРMethod: Least SquaresРDate: 05/02/12 Time: 15:53РSample: 1978 2008РIncluded observations: 31РVariableРCoefficientРStd.ErrorРt-StatisticРProb. РCР-1.529508Р1.099218Р-1.391451Р0.1759РX1Р0.083140Р0.035970Р2.311348Р0.0290РX2Р-0.057462Р0.012569Р-4.571708Р0.0001РX5Р-0.009149Р0.004241Р-2.157424Р0.0404РX6Р0.077022Р0.021178Р3.636922Р0.0012РR-squaredР0.793687Р Mean dependent varР2.063871РAdjusted R-squaredР0.761946Р S.D. dependent varР0.361512РS.E. of regressionР0.176385Р Akaikeinfo criterionР-0.485609РSum squared residР0.808900Р Schwarz criterionР-0.254321РLog likelihoodР12.52694Р F-statisticР25.00546РDurbin-Watson statР0.738860Р Prob(F-statistic)Р0.000000Р表4-9可见,样本决定系数=0.794,且=25.005,显著性值≈0.000,通过检验,拟合优度良好。同时,由表7可知所有系数的估计值都通过了检验且经济意义均合理,最后得到最优的回归方程:
全文预览
