计量经济学第二 第五章答案.doc

5.2(1) 建立某地区消费额Y与收入X的回归模型 估计模型参数，结果为 Dependent Variable: Y Method: Least Squares Date: 05/16/12 Time: 20:21 Sample: 1 60 Included observations: 60 Variable Coefficient Std. Error t-Statistic Prob. C 9.347522 3.638437 2.569104 0.0128 X 0.637069 0.019903 32.00881 0.0000 R-squared 0.946423 Mean dependent var 119.6667 Adjusted R-squared 0.945500 S.D. dependent var 38.68984 S.E. of regression 9.032255 Akaike info criterion 7.272246 Sum squared resid 4731.735 Schwarz criterion 7.342058 Log likelihood -216.1674 F-statistic 1024.564 Durbin-Watson stat 1.790431 Prob(F-statistic) 0.000000 该模型样本回归估计式的书写格式为 （2）用Goldfeld-Quandt法进行检验 1） 将样本X按递增顺序排序，去掉中间1/4的样本，再分为两个部分的样本，即 在Sample菜单里，将区间定义为1—22，然后用OLS方法求的如下结果： Dependent Variable: Y Method: Least Squares Date: 05/16/12 Time: 20:26 Sample: 1 22 Included observations: 22 Variable Coefficient Std. Error t-Statistic Prob. C 12.53695 7.069578 1.773365 0.0914 X 0.605911 0.063910 9.480730 0.0000 R-squared 0.817990 Mean dependent var 78.63636 Adjusted R-squared 0.808890 S.D. dependent var 12.56050 S.E. of regression 5.490969 Akaike info criterion 6.330594 Sum squared resid 603.0148 Schwarz criterion 6.429780 Log likelihood -67.63654 F-statistic 89.88424 Durbin-Watson stat 1.136382 Prob(F-statistic) 0.000000 Dependent Variable: Y Method: Least Squares Date: 05/16/12 Time: 20:27 Sample: 39 60 Included observations: 22 Variable Coefficient Std. Error t-Statistic Prob. C -39.54393 27.08272 -1.460116 0.1598 X 0.841215 0.113266 7.426927 0.0000 R-squared 0.733898 Mean dependent var 160.8182 Adjusted R-squared 0.720593 S.D. dependent var 21.13367 S.E. of regression 11.17103 Akaike info criterion 7.751033 Sum squared resid 2495.840 Schwarz criterion 7.850219 Log likelihood -83.26137 F-statistic 55.15924 Durbin-Watson stat 0.610587 Prob(F-statistic) 0.000000 由表可以得到两个部分的残差平方和，即 2）求F统计量为