实验五异方差模型的检验

实验报告课程名称：计量经济学实验项目：实验五异方差模型的检验和处理实验类型：综合性□设计性□验证性专业班别：12国姓名：学号：412实验课室：厚德楼A404指导教师：实验日期：2015年5月28日广东商学院华商学院教务处制一、实验项目训练方案小组合作：是□否小组成员：无实验目的：掌握异方差模型的检验和处理方法实验场地及仪器、设备和材料实验室：普通配置的计算机，Eviews软件及常用办公软件。实验训练内容（包括实验原理和操作步骤）：【实验原理】异方差的检验：图形检验法、Goldfeld-Quanadt检验法、White检验法、Glejser检验法；异方差的处理：模型变换法、加权最小二乘法(WLS)。【实验步骤】本实验考虑三个模型：【1】广东省财政支出CZ对财政收入CS的回归模型；（数据见附表1：附表1-广东省数据）【2】广东省固定资产折旧ZJ对国内生产总值GDPS和时间T的二元回归模型；（数据见附表1：附表1-广东省数据）【3】广东省各市城镇居民消费支出Y对人均收入X的回归模型。（数据见附表2：附表2-广东省2005年数据）（一）异方差的检验1.图形检验法分别用相关分析图和残差散点图检验模型【1】、模型【2】和模型【3】是否存在异方差。注：①相关分析图是作应变量对自变量的散点图（亦可作模型残差对自变量的散点图）；②残差散点图是作残差的平方对自变量的散点图。③模型【2】中作图取自变量为GDPS来作图。模型【1】相关分析图04008001,2001,6002,0002,40005001,0001,5002,000CSCZ残差散点图04,0008,00012,00016,00020,00005001,0001,5002,000CSE1模型【2】相关分析图05001,0001,5002,0002,5003,0003,5004,00005,00010,00015,00020,00025,000GDPSZJ残差散点图05,00010,00015,00020,00025,00030,00005,00010,00015,00020,00025,000GDPSE2模型【3】相关分析图4,0008,00012,00016,00020,00024,0005,00010,00015,00020,00025,00030,000XY残差散点图02,000,0004,000,0006,000,0008,000,00010,000,00012,000,0005,00010,00015,00020,00025,00030,000XE3【思考】①相关分析图和残差散点图的不同点是什么？②*在模型【2】中，自变量有两个，有无其他处理方法？尝试做出来。（请对得到的图表进行处理，以上在一页内）2.Goldfeld-Quanadt检验法用Goldfeld-Quanadt检验法检验模型【3】是否存在异方差。注：Goldfeld-Quanadt检验法的步骤为：①排序：②删除观察值中间的约1/4的，并将剩下的数据分为两个部分。③构造F统计量：分别对上述两个部分的观察值求回归模型，由此得到的两个部分的残差平方为21ie和22ie。21ie为较大的残差平方和，22ie为较小的残差平方和。④算统计量)2)(,2)((~2221*kcnkcnFeeFii。⑤判断：给定显著性水平0.05，查F分布表得临界值)()2)(,2)((kcnkcnF。如果)()2)(,2)((*kcnkcnFF，则认为模型中的随机误差存在异方差。（详见课本135页）将实验中重要的结果摘录下来，附在本页。obsXY17021.944632.68999999999927220.446317.0337299.256463.3748241.2099999999996350.3858842.846757.0269214.67294.9379867.367669.84810097.27476.65910908.368113.641011944.088296.43111229.179505.661215762.7712651.951317680.114485.611418287.2414468.241518907.7314323.661621015.0318550.561722881.821767.781828665.2521188.84DependentVariable:YMethod:LeastSquaresDate:06/07/15Time:11:18Sample:17Includedobservations:7VariableCoefficientStd.Errort-StatisticProb.X0.7230770.2183863.3110030.0212C536.88741814.2540.2959270.7792R-squared0.686771Meandependentvar6497.894AdjustedR-squared0.624125S.D.dependentvar966.9988S.E.ofregression592.8541Akaikeinfocriterion15.84273Sumsquaredresid1757380.Schwarzcriterion15.82728Loglikelihood-53.44956Hannan-Quinncriter.15.65172F-statistic10.96274Durbin-Watsonstat1.761325Prob(F-statistic)0.021217DependentVariable:YMethod:LeastSquaresDate:06/07/15Time:11:20Sample:1218Includedobservations:7VariableCoefficientStd.Errort-StatisticProb.X0.7592910.1778984.2681250.0080C1243.7433707.2380.3354900.7509R-squared0.784640Meandependentvar16776.66AdjustedR-squared0.741567S.D.dependentvar3677.261S.E.ofregression1869.382Akaikeinfocriterion18.13956Sumsquaredresid17472943Schwarzcriterion18.12411Loglikelihood-61.48846Hannan-Quinncriter.17.94855F-statistic18.21689Durbin-Watsonstat2.037081Prob(F-statistic)0.007953有上图可知21ie=17472943，22ie=175738021ie/22ie=17472943/1757380=9.94260945.在0.05下，上式中分子、分母的自由度均为5，查F分布表得临界值F0.05（5,5）=5.05，因为F=9.94260945F0.05（5,5）=5.05，所以拒接原假设，说明模型存在异方差。（请对得到的图表进行处理，以上在一页内）3.White检验法分别用White检验法检验模型【1】、模型【2】和模型【3】是否存在异方差。Eviews操作：先做模型，选view/ResidualTests/HeteroskedasticityTests/White/(勾选crossterms)。摘录主要结果附在本页内。模型【1】HeteroskedasticityTest:WhiteF-statistic4.40866Prob.F(2,25)0.0156Obs*R-squared7.932189Prob.Chi-Square(2)0.0189ScaledexplainedSS14.57723Prob.Chi-Square(2)0.0007TestEquation:DependentVariable:RESID^2Method:LeastSquaresDate:06/07/15Time:12:44Sample:19782005Includedobservations:28VariableCoefficientStd.Errort-StatisticProb.C-879.85131125.376-0.7818290.4417CS12.937204.6513282.7813980.0101CS^2-0.0066200.002964-2.2335610.0347R-squared0.283292Meandependentvar1940.891AdjustedR-squared0.225956S.D.dependentvar4080.739S.E.ofregression3590.225Akaikeinfocriterion19.31077Sumsquaredresid3.22E+08Schwarzcriterion19.45351Loglikelihood-267.3508Hannan-Quinncriter.19.35441F-statistic4.940866Durbin-Watsonstat2.144291Prob(F-statistic)0.015552模型【2】HeteroskedasticityTest:WhiteF-statistic1.993171Prob.F(5,22)0.1195Obs*R-squared8.729438Prob.Chi-Square(5)0.1204ScaledexplainedSS14.67857Prob.Chi-Square(5)0.0118TestEquation:DependentVariable:RESID^2Method:LeastSquaresDate:06/07/15Time:12:47Sample:19782005Includedobservations:28VariableCoefficientStd.Errort-StatisticProb.C1837.8986243.7010.2943600.7712GDPS-3.3950935.407361-0.6278650.5366GDPS^2-9.08E-050.000185-0.4895370.6293GDPS*T0.1603000.3151760.5086040.6161T-491.56141982.891-0.2479010.8065T^249.08543152.98750.3208460.7514R-squared0.311766Meandependentvar3461.910AdjustedR-squared0.155349S.D.dependentvar7240.935S.E.ofregression6654.775Akaikeinfocriterion20.63147Sumsquaredresid9.74E+08Schwarzcriterion20.91694Loglikelihood-282.8405Hannan-Quinncriter.20.71874F-statistic1.993171Durbin-Watsonstat1.971537Prob(F-statistic)0.119510模型【3】HeteroskedasticityTest:WhiteF-statistic7.670826Prob.F(2,15)0.0051Obs*R-squared9.101341Prob.Chi-Square(2)0.0106ScaledexplainedSS14.09286Prob.Chi-Square(2)0.0009TestEquation:DependentVariable:RESID^2Method:LeastSquaresDate:06/07/15Time:12:51Sample:118Includ