4Р0.002233Р-2.070782Р0.0509РX2*X4Р-0.003038Р0.001238Р-2.453421Р0.0230РX3Р40.11611Р26.50572Р1.513489Р0.1451РX3^2Р-0.000320Р0.000245Р-1.304830Р0.2061РX3*X4Р0.002580Р0.000754Р3.423015Р0.0026РX4Р-174.5875Р51.71831Р-3.375738Р0.0029РX4^2Р0.000409Р0.000166Р2.456311Р0.0228РR-squaredР0.924848Р Mean dependent varР66956.09РAdjusted R-squaredР0.892640Р S.D. dependent varР98811.72РS.E. of regressionР32376.52Р Akaike info criterionР23.86395РSum squared residР2.20E+10Р Schwarz criterionР24.32653РLog likelihoodР-359.8912Р Hannan-Quinn criter.Р24.01474РF-statisticР28.71478Р Durbin-Watson statР1.700188РProb(F-statistic)Р0.000000Р从表可看出nR2 =28.67028,而在5%的显著性水平下,查表得临界值Χ2(9)=16.919。因为nR2 =28.67028>Χ2(9)=16.919,所以,表明模型存在异方差。Р 从上述几种方法检验结果可以看出模型存在异方差性,那么我们将对其进行修正。Р6.2异方差的修正Р运用加权最小二乘法估计过程中,我们分别使用权
全文预览
