计量课后题上机回归分析.doc

《计量课后题上机回归分析.doc》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、第4章练习8下表列出了某年中国部分省市城镇居民家庭平均每个全年可支配收入（X）与消费性支出（Y）的统计数据。地区可支配收入（X）消费性支出（Y）地区可支配收入（X）消费性支出（Y）北京10349.698493.49浙江9279.167020.22天津8140.506121.04山东6489.975022.00河北5661.164348.47河南4766.263830.71山两4724.113941.87湖北5524.544644.5内蒙古5129.053927.75湖南6218.735218.79辽宁5357.794356.06广东9761.5780

2、16.91吉林4810.004020.87陕西5124.244276.67黑龙江4912.883824.44H'肃4916.254126.47I■-海11718.018868.19青海5169.964185.73江苏6800.235323.18新&5644.864422.93解：先作总体回归分析DependentVariable:YMethod:LeastSquaresDate:11/22/12Time:10:35Sample:120Ineludedobservations:20VariableCoefficientStd.Errort-Statis

3、ticProb.CX272.36350.755125159.67731.7057130.02331632.386900.10530.0000R-squaredAdjustedR-squaredS启.ofregressionSumsquaredresidLoglikelihoodDurbin-Watsonstat0.9831290.982192216.8900846743.0-134.91301.301684MeandependentvarS.D.dependentvarAkaikeinfocriterionSchwarzcriterionF-stat

4、isticProb(F-statistic)5199.5151625.27513.6913013.790871048.9120.000000可知冋归方程为:Y=272.3635389+0.7551249391*X一、异方差检验:1、图示检验法200000150000-CMLU100000-50000-OrOOO0-40006000800010000120002、怀特检验WhiteHeteroskedasticityTest:F-statistic14.63595Probability0.000201Obs*R-squared12.65213Proba

5、bility0.001789TestEquatiorrDependentVariable:RESIDA2Method:LeastSquaresDate:11/22/12Time:10:54Sample:120Ineludedobservationsi20VariableCoefficientStd.Errort-StatisticProb.c-180998.9103318.2-1.7518580.0978X49.4284628.939291.7080060.1058XA2-0.0021150.001847-1.1447420.2682R-square

6、d0.632606Meandependentvar42337.15AdjustedR-squared0.589384SDdependentvar45279.67S.E.ofregression29014.92Akaikeinfocriterion23.52649Sumsquaredresid1.43E+10Schwarzcriterion23.67585Loglikelihood-232.2649F-statistic14.63595Durbin-Watsonstat1.008103Prob(F-statistic)0.000201nR2=12.65

7、>5%置信水平下的卡方值5.99所以拒绝原假设，表明模型存在异方差。3、G・Q检验：先将X按生序排列，去掉中间4个样本观测值，得两个容量为8的子样本，再分别作普通最小二乘法回归得样本1DependentVariable:YMethod:LeastSquaresDate:11/22/12Time:11:10Sample:18Includedobservations:8VariableCoefficientStd.Errort-StatisticProb.c1277.1611540.6040.8290000.4388X0.5541260.3114321.

8、7792870.1255R-squared0.345397Meandependentvar4016.814A