Eviews作业.docx

2.12（1）得到Y与GDP的散点图建立税收随国内生产总值GDP变化的一元线性回归方程用OLS进行回归，的以下结果：Dependent Variable: YMethod: Least SquaresDate: 12/09/16 Time: 10:43Sample: 1 31Included observations: 31VariableCoefficientStd. Errort-StatisticProb.??C-10.6296386.06992-0.1235000.9026GDP0.0710470.0074079.5912450.0000R-squared0.760315????Mean dependent var621.0548Adjusted R-squared0.752050????S.D. dependent var619.5803S.E. of regression308.5176????Akaike info criterion14.36378Sum squared resid2760310.????Schwarz criterion14.45629Log likelihood-220.6385????Hannan-Quinn criter.14.39393F-statistic91.99198????Durbin-Watson stat1.570523Prob(F-statistic)0.000000Y=-10.6296+0.071GDP(-0.1235) ( 9.5912)RSS=2760310经济意义：由估计的模型可知，斜率项系数为0.0710，表明2007年国内生产总值GDP没增加1亿元，各地区税收平均增长0.071亿元。（2）模型检验由（1）中结果，可以看出，方程的调整后的可决系数为0.7521，即国内生产总值GDP可解释各地区税收的79.51%,仍有20.49%须由其他因素来解释，拟合效果一般。F值大于显著性水平为5%、自由度为（1,29）的F临界值，对应P值0.05,说明方程是显著的。GDP的系数对应t值大于显著性水平为5%、自由度为29的相应的临界值，对应的P值小于0.05，即GDP系数显著。（3）模型预测根据上述回归模型，当国内生产总值为8500亿元时，得到税收收入预测结果为593.2667亿元。借助“scalar AGDP=@MEAN(GDP) ,scalar VGDP=@VAR(GDP)”命令，出GDP的均值8891.1258，及方差5595781.5961，离差平方和为由RSS可知5981697.5682，于是预测的的标准差为439.8697,即预测区间为[153.3968,1033.1366]。3.13（1）首先对设定模型进行OLS回归：Dependent Variable: LOG(Y)Method: Least SquaresDate: 12/09/16 Time: 09:27Sample: 1 31Included observations: 31VariableCoefficientStd. Errort-StatisticProb.??C1.1539940.7276111.5860040.1240LOG(K)0.6092360.1763783.4541490.0018LOG(L)0.3607960.2015911.7897410.0843R-squared0.809925????Mean dependent var7.493997Adjusted R-squared0.796348????S.D. dependent var0.942960S.E. of regression0.425538????Akaike info criterion1.220839Sum squared resid5.070303????Schwarz criterion1.359612Log likelihood-15.92300????Hannan-Quinn criter.1.266075F-statistic59.65501????Durbin-Watson stat0.793209Prob(F-statistic)0.000000lnY=1.1540+0.6092lnK+0.3608lnL(1.5860) (3.4541) (1.7897)F=59.6550 RSS=5.0703 DW=0.7932可以看出，方程的调整后的可决系数为0.7964，即资产总值与职工数量联合起来可解释工业总产值的79.64%,仍有20.36%须由其他因素来解释。F值大于显著性水平为5%的F临界值，对应P值0.05,即整个方程是显著的。LnK的系数对应t值大于显著性水平为0.05的相应的临界值，对应