计量经济学实验报告-多重共线性检验(共8页).doc

《计量经济学实验报告-多重共线性检验(共8页).doc》由会员分享，可在线阅读，更多相关《计量经济学实验报告-多重共线性检验(共8页).doc（8页珍藏版）》请在得力文库 - 分享文档赚钱的网站上搜索。

1、精选优质文档-倾情为你奉上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内旅游和入境旅游两大市场，入境旅游外汇收入年均增长22.6%，与此同时国内旅游也迅速增长。改革开放20多年来，特别是进入90年代后，中国的国内旅游收入年均增长14.4%，远高于同期GDP 9.76%的增长率。 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， Yt第t年全国旅游收入 X2国内旅游人数（万人） X3城镇居民人均旅游支出 （元） X4农村居民人均旅游支

2、出 （元） X5公路里程（万公里） X6铁路里程（万公里）Y = 0.*X2 + 0.*X3 + 5.*X4 - 3.*X5 - 53.*X6 - 2220.数据来源中国统计局网站样本区间 19942009实验过程及结果（一）实证结果Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 15:49Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.X20.0.8.0.0000X30.1.0.0.8768

3、X45.1.2.0.0204X5-3.2.-1.0.1886X6-53.38584434.6829-0.0.9047C-2220.1512210.044-1.0.3388R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression252.1678Akaike info criterion14.17806Sum squared resid.0Schwarz criterion14.46778Log likelihood-107.4245F-statisti

4、c347.2644Durbin-Watson stat1.Prob(F-statistic)0.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（二）检查各解释变量之间的相关性X2X3X4X5X6X21.0.0.0.0.X30.1.0.0.0.X40.0.1.0.0.X50.0.0.1.0.X60.0.0.0.1.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：变量x2x3x4x5x6参数估计值0.17.7860328.798712

5、4.211453751.241t统计量27.922416.4.10.8558812.83403R-squared0.982360.0.0.0.按R-squared大小排序为：X2、X6、X5、X3、X4以X2为基础，分别加入X3、X4、X5、X6加入X3，不显著，排除。Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:11Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-3012.810

6、549.3740-5.0.0001X20.0.14.959910.0000X32.1.2.0.0555R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression335.2951Akaike info criterion14.63526Sum squared resid.Schwarz criterion14.78012Log likelihood-114.0821F-statistic487.3775Durbin-Watson stat0.Prob(F

7、-statistic)0.加入X4，显著，保留Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:12Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2360.596180.9037-13.048910.0000X20.0.27.598450.0000X45.1.4.0.0006R-squared0.Mean dependent var4270.119Adjusted R-squared

8、0.S.D. dependent var2720.860S.E. of regression242.2577Akaike info criterion13.98524Sum squared resid.5Schwarz criterion14.13010Log likelihood-108.8819F-statistic939.5591Durbin-Watson stat0.Prob(F-statistic)0.加入X5，不显著，排除Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:12Sample: 1994

9、2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2052.457252.8897-8.0.0000X20.0.8.0.0000X5-4.3.-1.0.2549R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression368.5792Akaike info criterion14.82455Sum squared resid.Schwarz crite

10、rion14.96941Log likelihood-115.5964F-statistic402.2068Durbin-Watson stat0.Prob(F-statistic)0.加入X6，显著，保留Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:12Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-6813.6822061.836-3.0.0057X20.0.8.0.0000X

11、6867.2159365.80032.0.0339R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression324.3425Akaike info criterion14.56884Sum squared resid.Schwarz criterion14.71370Log likelihood-113.5507F-statistic521.2956Durbin-Watson stat0.Prob(F-statistic)0.保留了X4和X6分别以X2

12、和X4为基础，加入其他变量加入X3不显著，排除Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:35Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2323.426462.3572-5.0.0003X20.0.20.013290.0000X45.1.3.0.0061X3-0.1.-0.0.9313R-squared0.Mean dependent var4270.119Adjusted

13、 R-squared0.S.D. dependent var2720.860S.E. of regression252.0685Akaike info criterion14.10960Sum squared resid.3Schwarz criterion14.30274Log likelihood-108.8768F-statistic578.5661Durbin-Watson stat0.Prob(F-statistic)0.加入X5不显著，排除Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:37Samp

14、le: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2432.172178.3609-13.636240.0000X20.0.11.558080.0000X45.1.4.0.0006X5-3.2.-1.0.1508R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression230.5411Akaike info criterion13.9

15、3105Sum squared resid.2Schwarz criterion14.12420Log likelihood-107.4484F-statistic692.4432Durbin-Watson stat1.Prob(F-statistic)0.加入X6不显著，排除Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:25Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2291

16、.5332172.352-1.0.3123X20.0.10.999930.0000X45.1.3.0.0095X6-12.85589402.8593-0.0.9751R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression252.1391Akaike info criterion14.11016Sum squared resid.7Schwarz criterion14.30330Log likelihood-108.8813F-statistic5

17、78.2396Durbin-Watson stat0.Prob(F-statistic)0.以X2、X6为基础，加入其他变量加入X3 ，不显著，排除Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:38Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-5877.5502399.623-2.0.0306X20.0.8.0.0000X6606.4675495.01671.0.2440X31.

18、1.0.0.4416R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression329.0154Akaike info criterion14.64240Sum squared resid.Schwarz criterion14.83555Log likelihood-113.1392F-statistic337.9400Durbin-Watson stat0.Prob(F-statistic)0.加入X4不显著，排除Dependent Variable

19、: YMethod: Least SquaresDate: 04/06/11 Time: 16:38Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2291.5332172.352-1.0.3123X20.0.10.999930.0000X6-12.85589402.8593-0.0.9751X45.1.3.0.0095R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent va

20、r2720.860S.E. of regression252.1391Akaike info criterion14.11016Sum squared resid.7Schwarz criterion14.30330Log likelihood-108.8813F-statistic578.2396Durbin-Watson stat0.Prob(F-statistic)0.加入X5不显著，排除Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:39Sample: 1994 2009Included observa

21、tions: 16VariableCoefficientStd. Errort-StatisticProb.C-6722.3952029.031-3.0.0062X20.0.6.0.0000X6835.0915360.72052.0.0391X5-3.2.-1.0.2529R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression318.9580Akaike info criterion14.58031Sum squared resid.Schwarz

22、 criterion14.77346Log likelihood-112.6425F-statistic359.8440Durbin-Watson stat0.Prob(F-statistic)0.最后证实结果X2和X4Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:12Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-2360.596180.9037-13.048910.0000X2

23、0.0.27.598450.0000X45.1.4.0.0006R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var2720.860S.E. of regression242.2577Akaike info criterion13.98524Sum squared resid.5Schwarz criterion14.13010Log likelihood-108.8819F-statistic939.5591Durbin-Watson stat0.Prob(F-statistic)0.X2和X6

24、Dependent Variable: YMethod: Least SquaresDate: 04/06/11 Time: 16:12Sample: 1994 2009Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C-6813.6822061.836-3.0.0057X20.0.8.0.0000X6867.2159365.80032.0.0339R-squared0.Mean dependent var4270.119Adjusted R-squared0.S.D. dependent var272

25、0.860S.E. of regression324.3425Akaike info criterion14.56884Sum squared resid.Schwarz criterion14.71370Log likelihood-113.5507F-statistic521.2956Durbin-Watson stat0.Prob(F-statistic)0.结果分析：在其他因素不变的情况下，当国内旅游人数X2增加1万人和农村人均旅游支出增加1元，国内旅游收入将分别0.056亿元和增长5.45亿元。在其他因素不变的情况下，作为旅游设施的代表，铁路里程X6 每增加1万公里时, 国内旅游收入将增长867.21亿元。专心-专注-专业