计量经济学实验报告 多重共线性检验.docx

1、计量经济学实验报告 多重共线性检验计量经济学实验报告 多重共线性检验计量经济学实验报告 计量经济学上机实验报告 多重共线性检验 实验背景 近年来，中国旅游业一直保持高速发展，旅游业作为国民经济新的增长点，在整个社会经济发展中的作用日益显现。中国的旅游业分为国为了规划中国未来旅游产业的发展，需要定量地分析影响中国旅游市场发展的主要因素。 模型 其中， 第t年全国旅游收入 X2国 Time: 15:49 Sample: 1994 2009 Included Ytobservations: 16 Variable X2 X3 X4 Coefficient 0.063969 0.209819 5.28

2、3347 Std. Error 0.007714 1.319292 1.918838 1 t-Statistic 8.292875 0.159039 2.753409 Prob. 0.0000 0.8768 0.0204 计量经济学实验报告 X5 X6 C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat -3.352907 -53.38584 -2220.151 2.376484 434.6829 2210.044 -1.410869 -0.1

3、22816 -1.004573 0.1886 0.9047 0.3388 4270.119 2720.860 14.17806 14.46778 347.2644 0.000000 0.994274 Mean dependent var 0.991411 S.D. dependent var 252.1678 Akaike info criterion 635886.0 Schwarz criterion -107.4245 F-statistic 1.224560 Prob(F-statistic) R2很高，F显著，但x3、x5、x6不显著，X5、X6的符号甚至是负的。可能存在多重共线性

4、(二) 检查各解释变量之间的相关性 X2 X3 X4 X5 X6 各解释变量相互之间的相关系数较高，证实确实存在严重多重共线性。 (三) 进一步检验和消除多重共线性，采用逐步回归法 分别作Y对X2、X3、X4、X5、X6的一元回归，结果如下: X2 1.000000 0.843453 0.700655 0.964909 0.944409 X3 0.843453 1.000000 0.843991 0.822201 0.913464 X4 0.700655 0.843991 1.000000 0.661270 0.827867 X5 0.964909 0.822201 0.661270 1.00

5、0000 0.904867 X6 0.944409 0.913464 0.827867 0.904867 1.000000 以X2为基础，分别加入X3、X4、X5、X6 加入X3，不显著，排除。 Dependent Variable: Y Method: Least Squares Date: 04/06/11 Time: 16:11 Sample: 1994 2009 Included observations: 16 Variable C Coefficient -3012.810 Std. Error 549.3740 t-Statistic -5.484078 Prob. 0.0001

6、 按R-squared大小排序为:X2、X6、X5、X3、X4 2 计量经济学实验报告 X2 X3 R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 0.055411 2.541150 0.003704 1.208186 14.95991 2.103276 0.0000 0.0555 4270.119 2720.860 14.63526 14.78012 487.3775 0.000000 0.986839 Mean dependent var

7、0.984814 S.D. dependent var 335.2951 Akaike info criterion 1461496. Schwarz criterion -114.0821 F-statistic 0.702762 Prob(F-statistic) 加入X4，显著，保留 Dependent Variable: Y Method: Least Squares Date: 04/06/11 Time: 16:12 Sample: 1994 2009 Included observations: 16 Variable C X2 X4 R-squared Adjusted R-s

8、quared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat Coefficient -2360.596 0.055609 5.450434 Std. Error 180.9037 0.002015 1.207441 t-Statistic -13.04891 27.59845 4.514038 Prob. 0.0000 0.0000 0.0006 4270.119 2720.860 13.98524 14.13010 939.5591 0.000000 0.993129 Mean dependent

9、 var 0.992072 S.D. dependent var 242.2577 Akaike info criterion 762954.5 Schwarz criterion -108.8819 F-statistic 0.905401 Prob(F-statistic) 加入X5，不显著，排除 Dependent Variable: Y Method: Least Squares Date: 04/06/11 Time: 16:12 Sample: 1994 2009 Included observations: 16 Variable C X2 X5 R-squared Coeffi

10、cient -2052.457 0.071556 -4.063442 Std. Error 252.8897 0.008330 3.411178 t-Statistic -8.116016 8.590340 -1.191214 Prob. 0.0000 0.0000 0.2549 4270.119 0.984096 Mean dependent var 3 计量经济学实验报告 Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 0.981649 S.D. depend

11、ent var 368.5792 Akaike info criterion 1766058. Schwarz criterion -115.5964 F-statistic 0.718757 Prob(F-statistic) 2720.860 14.82455 14.96941 402.2068 0.000000 加入X6，显著，保留 Dependent Variable: Y Method: Least Squares Date: 04/06/11 Time: 16:12 Sample: 1994 2009 Included observations: 16 Variable C X2

12、X6 R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 保留了X4和X6 加入X3 不显著，排除 Dependent Variable: Y Method: Least Squares Date: 04/06/11 Time: 16:35 Sample: 1994 2009 Included observations: 16 Variable C X2 X4 X3 R-squared Coefficient -2323.426 0.055771

13、5.547612 -0.106429 Std. Error 462.3572 0.002787 1.672536 1.209188 t-Statistic -5.025174 20.01329 3.316886 -0.088017 Prob. 0.0003 0.0000 0.0061 0.9313 4270.119 Coefficient -6813.682 0.048874 867.2159 Std. Error 2061.836 0.005854 365.8003 t-Statistic -3.304667 8.348242 2.370736 Prob. 0.0057 0.0000 0.0

14、339 4270.119 2720.860 14.56884 14.71370 521.2956 0.000000 0.987685 Mean dependent var 0.985790 S.D. dependent var 324.3425 Akaike info criterion 1367575. Schwarz criterion -113.5507 F-statistic 0.733651 Prob(F-statistic) 分别以X2和X4为基础，加入其他变量 0.993134 Mean dependent var 4 计量经济学实验报告 Adjusted R-squared S

15、.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 加入X5 不显著，排除 Dependent Variable: Y Method: Least Squares 0.991417 S.D. dependent var 252.0685 Akaike info criterion 762462.3 Schwarz criterion -108.8768 F-statistic 0.923342 Prob(F-statistic) 2720.860 14.10960 14.30274 578.5661 0.000000 Std. Error 178.3609 0.005495 1.152645 2.140330 t-Statistic -13.63624 11.55808 4.607421 -1.534588 Coefficient -2432.172 0.063511 5.310720 -3.284526 Prob. 0.0000 0.0000 0.0006 0.1508 4270.119 2720.860 13.93105 14.12420 692.4432 0.000000 Date: 04/06/11 Time: 16:37