多重共线性案例分析.ppt

《多重共线性案例分析.ppt》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、第十二讲多重共线性案例分析多重共线性的侦查与处理研究美国每人的子鸡消费量令：获取数据196027.80000397.500042.2000050.7000078.3000065.80000196129.90000413.300038.1000052.0000079.2000066.90000196229.80000439.200040.3000054.0000079.2000067.80000196330.80000459.700039.5000055.3000079.2000069.60000196431.20000492.900037.300

2、0054.7000077.4000068.70000196533.30000528.600038.1000063.7000080.2000073.60000196635.60000560.300039.3000069.8000080.4000076.30000196736.40000624.600037.8000065.9000083.9000077.20000196836.70000666.400038.4000064.5000085.5000078.10000196938.40000717.800040.1000070.0000093.700

3、0084.70000197040.40000768.200038.6000073.20000106.100093.30000197140.30000843.300039.8000067.80000104.800089.70000197241.80000911.600039.7000079.10000114.0000100.7000197340.40000931.100052.1000095.40000124.1000113.5000197440.700001021.50048.9000094.20000127.6000115.3000197540

4、.100001165.90058.30000123.5000142.9000136.7000197642.700001349.60057.90000129.9000143.6000139.2000197744.100001449.40056.50000117.6000139.2000132.0000197846.700001575.50063.70000130.9000165.5000132.1000197950.600001759.10061.60000129.8000203.3000154.4000198050.100001994.20058

5、.90000128.0000219.6000174.9000198151.700002258.10066.40000141.0000221.6000180.8000198252.900002478.70070.40000168.2000232.6000189.4000估计方程及报告结果DependentVariable:LOG(Y)Method:LeastSquaresDate:10/11/10Time:21:40Sample:19601982Includedobservations:23VariableCoefficientStd.Errort

6、-StatisticProb.C2.1897920.15571514.062830.0000LOG(X1)0.3425550.0832664.1139700.0007LOG(X2)-0.5045920.110894-4.5502120.0002LOG(X3)0.1485450.0996731.4903340.1535LOG(X4)0.0911050.1007160.9045680.3776R-squared0.982313    Meandependentvar3.663887AdjustedR-squared0.978383    S.D.de

7、pendentvar0.187659S.E.ofregression0.027591Akaikeinfocriterion-4.152987Sumsquaredresid0.013703    Schwarzcriterion-3.906140Loglikelihood52.75935Hannan-Quinncriter.-4.090906F-statistic249.9282    Durbin-Watsonstat1.826069Prob(F-statistic)0.000000通过总体显著性检验检验：、通过显著性检验；、未通过显著性检验很大

8、多重共线性的侦查（1）值高且不显著的值四个中占了两个，有理由怀疑该模型中存在着较为严重的多重共线性。LOG(X1)LOG(X2)LOG(