.033135Р0.007760Р4.270053Р0.0004РX2*X3Р-97.11637Р96.87489Р-1.002493Р0.3281РX3Р55473498Р68538734Р0.809374Р0.4278РX3^2Р-283697.5Р290382.6Р-0.976978Р0.3403РR-squaredР0.937930Р Mean dependent varР77607780РAdjusted R-squaredР0.909998Р S.D. dependent varР1.80E+08РS.E. of regressionР54097636Р Akaike info criterionР38.71168РSum squared residР5.85E+16Р Schwarz criterionР39.17875РLog likelihoodР-570.6752Р Hannan-Quinn criter.Р38.86110РF-statisticР33.57944Р Durbin-Watson statР2.262413РProb(F-statistic)Р0.000000Р使用White检验法不论是否带有交叉项,所得的检验伴随概率均小于5%,均在5%的显著水平下拒绝方程不存在异方差性的原假设,认为模型具有比较严重的异方差性。需要对模型进行修正。Р②多重共线性检验:Р用逐步回归法检验如下Р以为被解释变量,逐个引入解释变量、、,构成回归模型,进行模型估计。Р表7: 被解释变量与最小二乘估计结果РDependent Variable: YРMethod: Least SquaresРDate: 12/29/11 Time: 18:32РSample: 1980 2009РIncluded observations: 30
全文预览
