计量经济学多重线性的检验及修正

《计量经济学多重线性的检验及修正》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、多重线性的检验及修正成员:【实验目的】运用Eviews软件掌握多重共线性的检验及修正方法，加深对理论知识的理解和运用。【实验内容】经分析，影响财政收入（丫）的主要因素有：农业增加值（么）、工业增加值（x2）、建筑业增加值（x3）、总人口（X4）、受灾面积（x5）。拟建模型为:Y=+X1+x2+X3+X4+x5+Wo相关数据如下：财政收入丫农业增加值XIT、lk增加值X2建筑业増加值X3总人口X4受灾面积X51132.31027.51607138.296259507901146.61270.21769.7143.8975423

2、93701159.91371.41996.5195.598705445261175.81559.52048.5207.1100072397901212.31777.42162.3220.71016543313013671978.52375.8270.6103008347101642.52316.12789316.7104357318902004.62564.33448.5417.91058514436521222788.73987.5525.7107507471702199.432334565.9665.8109300420

3、902357.63865.45062810111026508702664.550628087.3794112704469912937.45342.310284.5859.4114333384743149.485866.8141881015.1115823554723483.486963.619480.51415117171513334348.959572.719480.42266.5118517488295218.112315.724950.72964.7119850550436242.214015.829447.63728

4、.8121121458217407.9914441.832921.44387.4122389468988615.1414917.634018.44985.8123626534299875.9514944.5400365172.11247615914511444.0815871.843580.65522.31257864998113395.231653747431.65913.71267435468816386.0417381.854945.56465.51276275221518903.6421412.7652107490.

5、81284534711921715.252242076912.68694.31292275450626396.472122487632.48967.81299883710631649.492242089834.510133.81307563881838760.224040.991310.911851.11314484109151321.4528095107367.214014.113212948992【实验过程及分析】(一)普通最小二乘法估计模型DependentVariable:YMethod:LeastSquaresDa

6、te:10/24/11Time:22:49Sample:19782007Includedobservations:30VariableCoefficientStd.Errort-StatisticProb.C-6734.39411259.37-0.5981150.5554X1•1.6786110.328371-5.1119370.0000X20.0710780.0811710.8756660.3899X35.6991990.7455917.6438700.0000X40.1014810.1142440.8882770.383

7、2X5•0.0109220.057578-0.1896910.8511R-squared0.983660Meandependentvar10047.83AdjustedR-squared0.980255S.D.dependentvar12585.61S.E.ofregression1768.473Akaikeinfocriterion17.97048Sumsquaredresid75059958Schwarzcriterion18.25072Loglikelihood-263.5572F-statistic288.9512D

8、urbin-Watsonstat0.898668Prob(F-statistic)0.000000由此可见，该模型尺2=0.983660，M2=0.980255可决系数很高，F检验值为288.9512,明显显著。但是，当汉=0.05时，f//2(n-k)=“()25(30-6)=2.064