异方差与序列相关性练习

《异方差与序列相关性练习》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、一、异方差检验与修正（一）建立初始回归模型相关命令：dataxyscatxylsycx模型一：DependentVariable:YMethod:LeastSquaresDate:10/23/14Time:10:46Sample:120Includedobservations:20VariableCoefficientStd.Errort-StatisticProb.  C272.3635159.67731.7057130.1053X0.7551250.02331632.386900.0000R-

2、squared0.983129    Meandependentvar5199.515AdjustedR-squared0.982192    S.D.dependentvar1625.275S.E.ofregression216.8900    Akaikeinfocriterion13.69130Sumsquaredresid846743.0    Schwarzcriterion13.79087Loglikelihood-134.9130    F-statistic1048.912Durb

3、in-Watsonstat1.301684    Prob(F-statistic)0.000000（二）异方差的四种检验方法及其分析右击resid选择ObjectCopy，输入e得到初始回归模型的残差序列；1.图示法：scatxe^22.模型检验法：lse^2cxDependentVariable:E^2Method:LeastSquaresDate:10/23/14Time:10:52Sample:120Includedobservations:20VariableCoefficientStd

4、.Errort-StatisticProb.  C-65281.6621544.58-3.0300730.0072X16.493443.1458955.2428430.0001R-squared0.604286    Meandependentvar42337.15AdjustedR-squared0.582302    S.D.dependentvar45279.67S.E.ofregression29264.05    Akaikeinfocriterion23.50075Sumsquared

5、resid1.54E+10    Schwarzcriterion23.60032Loglikelihood-233.0075    F-statistic27.48740Durbin-Watsonstat1.029463    Prob(F-statistic)0.0000553.GQ假设检验法首先，点击工具按钮proc选择sortcurrentpage，输入X，按升序排序；去掉中间约n/4个样本点，然后对前后两个子样本分别进行回归；子样本模型一：DependentVariable:YMetho

6、d:LeastSquaresDate:10/23/14Time:10:57Sample:18Includedobservations:8VariableCoefficientStd.Errort-StatisticProb.  C1277.1611540.6040.8290000.4388X0.5541260.3114321.7792870.1255R-squared0.345397    Meandependentvar4016.814AdjustedR-squared0.236296    S

7、.D.dependentvar166.1712S.E.ofregression145.2172    Akaikeinfocriterion13.00666Sumsquaredresid126528.3    Schwarzcriterion13.02652Loglikelihood-50.02663    F-statistic3.165861Durbin-Watsonstat3.004532    Prob(F-statistic)0.125501子样本模型二：DependentVariabl

8、e:YMethod:LeastSquaresDate:10/23/14Time:10:57Sample:1320Includedobservations:8VariableCoefficientStd.Errort-StatisticProb.  C212.2118530.88920.3997290.7032X0.7618930.06034812.625050.0000R-squared0.963723    Meandependentvar6760.477AdjustedR-sq