最新Eviews实验报告.doc

精品资料Eviews实验报告.江西农业大学经济贸易学院学 生 实 验 报 告课程名称： 计量经济学 专业班级： 经济1201班姓 名： 学 号： 指导教师： 徐冬梅 职称： 讲师 实验日期： 2014.12.11 学生实验报告学生姓名学号组员：实验项目EVIEWS的使用必修 选修 演示性实验 验证性实验 操作性实验 综合性实验实验地点管理模拟实验室实验仪器台号指导教师实验日期及节次一、实验目的及要求1、目的会使用EVIEWS对计量经济模型进行分析2、内容及要求（1）对经典线形回归模型进行参数估计、参数的检验与区间估计，对模型总体进行显著性检验；（2）异方差的检验及其处理；（3）自相关的检验及其处理；（4）多重共线性检验及其处理；二、仪器用具仪器名称规格/型号数量备注计算机1无网络环境Eviews1三、实验方法与步骤(一)数据的输入、描述及其图形处理；(二)方程的估计；(三)参数的检验、违背经典假定的检验；(四)模型的处理与预测四、实验结果与数据处理实验一：中国城镇居民人均消费支出模型数据散点图：通过Eviews估计参数方程回归方程：Dependent Variable: YMethod: Least SquaresDate: 11/27/14 Time: 15:02Sample: 1 31Included observations: 31VariableCoefficientStd. Errort-StatisticProb. X1.3594770.04330231.395250.0000C-57.90655377.7595-0.1532890.8792R-squared0.971419 Mean dependent var11363.69Adjusted R-squared0.970433 S.D. dependent var3294.469S.E. of regression566.4812 Akaike info criterion15.57911Sum squared resid9306127. Schwarz criterion15.67162Log likelihood-239.4761 F-statistic985.6616Durbin-Watson stat1.294974 Prob(F-statistic)0.000000得出估计方程为：Y = 1.35947661442*X - 57.9065479515异方差检验1、 图示检验法 图形呈现离散趋势，大致判断存在异方差性。2、 Park检验Dependent Variable: LOG(E2)Method: Least SquaresDate: 11/27/14 Time: 16:16Sample: 1 31Included observations: 31VariableCoefficientStd. Errort-StatisticProb. C19.8256219.853590.9985910.3263LOG(X)-0.9564032.204080-0.4339240.6676R-squared0.006451 Mean dependent var11.21371Adjusted R-squared-0.027809 S.D. dependent var2.894595S.E. of regression2.934568 Akaike info criterion5.053338Sum squared resid249.7389 Schwarz criterion5.145854Log likelihood-76.32674 F-statistic0.188290Durbin-Watson stat2.456500 Prob(F-statistic)0.667555看到图中LOG(E2)中P值为0.6676 > 0.05，所以不存在异方差性3、 G-Q检验e1检验：Dependent Variable: XMethod: Least SquaresDate: 11/27/14 Time: 16:41Sample: 1 12Included observations: 12VariableCoefficientStd. Errort-StatisticProb. C4642.0282014.1832.3046710.0439Y0.2310460.2158241.0705300.3095R-squared0.102820 Mean dependent var6796.390Adjusted R-squared0.013102 S.D. dependent var293.2762S.E. of regression291.3486 Akaike info criterion14.33793Sum squared resid848840.2 Schwarz criterion14.41875Log likelihood-84.02758 F-statistic1.146034Durbin-Watson stat0.445146 Prob(F-statistic)0.309538e2检验：Dependent Variable: XMethod: Least SquaresDate: 11/27/14 Time: 16:42Sample: 20 31Included observations: 12VariableCoefficientStd. Errort-StatisticProb. C583.4526593.43700.9831750.3487Y0.6977480.04019617.358700.0000R-squared0.967879 Mean dependent var10586.89Adjusted R-squared0.964667 S.D. dependent var2610.864S.E. of regression490.7655 Akaike info criterion15.38082Sum squared resid2408507. Schwarz criterion15.46164Log likelihood-90.28493 F-statistic301.3245Durbin-Watson stat2.748144 Prob(F-statistic)0.000000第一个图中的残差平方和为848840.2第二个图中的残差平方和为2408507所以F值为2408507/848840.2 = 2.8374 < 2.97，所以不存在异方差性4、 White检验White Heteroskedasticity Test:F-statistic2.240402 Probability0.125152Obs*R-squared4.276524 Probability0.117860Test Equation:Dependent Variable: RESID2Method: Least SquaresDate: 11/27/14 Time: 16:50Sample: 1 31Included observations: 31VariableCoefficientStd. Errort-StatisticProb. C-2135113.1158576.-1.8428760.0760X503.7331242.20782.0797560.0468X2-0.0236090.011650-2.0265900.0523R-squared0.137952 Mean dependent var300197.6Adjusted R-squared0.076378 S.D. dependent var347663.4S.E. of regression334122.9 Akaike info criterion28.36817Sum squared resid3.13E+12 Schwarz criterion28.50694Log likelihood-436.7067 F-statistic2.240402Durbin-Watson stat1.871252 Prob(F-statistic)0.125152P值为0.11786 > 0.05，所以不存在异方差性通过四种不同的检验得知除了图示检验法得出异方差的结论，其他的检验的结论都是不存在异方差的。5、 WLS（加权最小二乘法）修正Dependent Variable: YMethod: Least SquaresDate: 11/27/14 Time: 17:14Sample: 1 31Included observations: 31Weighting series: E3VariableCoefficientStd. Errort-StatisticProb. C-85.6942624.15675-3.5474250.0013X1.3622210.002307590.56150.0000Weighted StatisticsR-squared1.000000 Mean dependent var13474.53Adjusted R-squared1.000000 S.D. dependent var61353.74S.E. of regression27.93264 Akaike info criterion9.559810Sum squared resid22626.73 Schwarz criterion9.652325Log likelihood-146.1770 F-statistic348762.9Durbin-Watson stat2.061818 Prob(F-statistic)0.000000Unweighted StatisticsR-squared0.971413 Mean dependent var11363.69Adjusted R-squared0.970427 S.D. dependent var3294.469S.E. of regression566.5415 Sum squared resid9308110.Durbin-Watson stat2.178992实验二：中国粮食生产函数1、回归方程Dependent Variable: LOG(Y)Method: Least SquaresDate: 12/11/14 Time: 15:06Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LOG(X1)0.3811450.0502427.5861820.0000LOG(X2)1.2222890.1351799.0420300.0000LOG(X3)-0.0811100.015304-5.3000240.0000LOG(X4)-0.0472290.044767-1.0549800.3047LOG(X5)-0.1011740.057687-1.7538530.0956C-4.1731741.923624-2.1694340.0429R-squared0.981597 Mean dependent var10.70905Adjusted R-squared0.976753 S.D. dependent var0.093396S.E. of regression0.014240 Akaike info criterion-5.459968Sum squared resid0.003853 Schwarz criterion-5.167438Log likelihood74.24960 F-statistic202.6826Durbin-Watson stat1.791427 Prob(F-statistic)0.000000得出回归方程为：LOG(Y) = 0.381144581612*LOG(X1) + 1.22228859801*LOG(X2) - 0.0811098881534*LOG(X3) - 0.04722870996*LOG(X4) - 0.101173736285*LOG(X5) - 4.17317444909通过检验结果可知 R2 较大且接近于1，而且F=202.6826 > F0.05(5,19) = 2.74，故认为粮食产量与上述变量之间总体线性关系显著。但是由于其中X4、X5前的参数估计值未通过t检验，且符号的经济意义不合理，故认为解释变量之间存在多重共线。2、相关系数表LNX1LNX2LNX3LNX4LNX5LNX1 1.000000-0.568744 0.451700 0.964357 0.440205LNX2-0.568744 1.000000-0.214097-0.697625-0.073270LNX3 0.451700-0.214097 1.000000 0.398780 0.411279LNX4 0.964357-0.697625 0.398780 1.000000 0.279528LNX5 0.440205-0.073270 0.411279 0.279528 1.000000由表可知LnX1与LnX2之间存在高度的线性相关性3、简单的回归形式LnY与LnX1Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:15Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX10.2240050.0255158.7792930.0000C8.9020080.20603443.206570.0000R-squared0.770175 Mean dependent var10.70905Adjusted R-squared0.760182 S.D. dependent var0.093396S.E. of regression0.045737 Akaike info criterion-3.255189Sum squared resid0.048114 Schwarz criterion-3.157679Log likelihood42.68986 F-statistic77.07599Durbin-Watson stat0.939435 Prob(F-statistic)0.000000LnY与LnX2Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:16Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX2-0.3834340.509669-0.7523210.4595C15.157485.9129712.5634290.0174R-squared0.024017 Mean dependent var10.70905Adjusted R-squared-0.018417 S.D. dependent var0.093396S.E. of regression0.094252 Akaike info criterion-1.809063Sum squared resid0.204321 Schwarz criterion-1.711553Log likelihood24.61329 F-statistic0.565986Durbin-Watson stat0.335219 Prob(F-statistic)0.459489LnY与LnX3Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:18Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX30.1080670.0852711.2673350.2177C9.6197220.85974411.189050.0000R-squared0.065274 Mean dependent var10.70905Adjusted R-squared0.024634 S.D. dependent var0.093396S.E. of regression0.092239 Akaike info criterion-1.852255Sum squared resid0.195684 Schwarz criterion-1.754745Log likelihood25.15319 F-statistic1.606139Durbin-Watson stat0.597749 Prob(F-statistic)0.217717LnY与LnX4Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:18Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX40.1669760.0282745.9056700.0000C8.9490900.29825530.004790.0000R-squared0.602605 Mean dependent var10.70905Adjusted R-squared0.585327 S.D. dependent var0.093396S.E. of regression0.060143 Akaike info criterion-2.707578Sum squared resid0.083194 Schwarz criterion-2.610068Log likelihood35.84472 F-statistic34.87693Durbin-Watson stat0.625528 Prob(F-statistic)0.000005LnY与LnX5Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:19Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX50.4887310.2346062.0831990.0485C5.6007492.4522072.2839620.0319R-squared0.158733 Mean dependent var10.70905Adjusted R-squared0.122156 S.D. dependent var0.093396S.E. of regression0.087506 Akaike info criterion-1.957599Sum squared resid0.176118 Schwarz criterion-1.860089Log likelihood26.46999 F-statistic4.339718Durbin-Watson stat0.327932 Prob(F-statistic)0.048538比较各个回归方程的R2可知Y与X1的R2最大，即粮食生产受农业化肥施用量最大，与经验相符，因此选为初始的回归方程。且初始化回归方程为：LOG(Y) = 0.224004867873*LOG(X1) + 8.90200821784R2 = 0.770175 D.W.= 0.9394354、逐步回归LnY与LnX1Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:28Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX10.2240050.0255158.7792930.0000C8.9020080.20603443.206570.0000R-squared0.770175 Mean dependent var10.70905Adjusted R-squared0.760182 S.D. dependent var0.093396S.E. of regression0.045737 Akaike info criterion-3.255189Sum squared resid0.048114 Schwarz criterion-3.157679Log likelihood42.68986 F-statistic77.07599Durbin-Watson stat0.939435 Prob(F-statistic)0.000000LnY与LnX1、LnX2Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:29Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX10.2978540.01548219.239290.0000LNX21.2586220.1500668.3871270.0000C-6.2956821.814941-3.4688090.0022R-squared0.945246 Mean dependent var10.70905Adjusted R-squared0.940269 S.D. dependent var0.093396S.E. of regression0.022826 Akaike info criterion-4.609666Sum squared resid0.011463 Schwarz criterion-4.463401Log likelihood60.62083 F-statistic189.9002Durbin-Watson stat1.595748 Prob(F-statistic)0.000000由输出结果可知R2有所提高，且各解释变量前得参数均通过t检验，符号也合理。D.W.检验也表明不存在一阶自相关。可以考虑再此模型上继续引入X3。LnY与LnX1、LnX2、LnX3Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:30Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX10.3233850.01086129.775520.0000LNX21.2907290.09615313.423650.0000LNX3-0.0867540.015155-5.7244840.0000C-5.9996381.162078-5.1628520.0000R-squared0.978616 Mean dependent var10.70905Adjusted R-squared0.975561 S.D. dependent var0.093396S.E. of regression0.014601 Akaike info criterion-5.469854Sum squared resid0.004477 Schwarz criterion-5.274834Log likelihood72.37318 F-statistic320.3438Durbin-Watson stat1.412883 Prob(F-statistic)0.000000由输出结果可知R2再次提高且参数符号合理，变量通过t检验。但是D.W.=1.419（dL=1.12、dU=1.66）落入无法判断的区域，且X4的参数没有通过t检验。LM检验Breusch-Godfrey Serial Correlation LM Test:F-statistic1.241319 Probability0.278428Obs*R-squared1.460972 Probability0.226776Test Equation:Dependent Variable: RESIDMethod: Least SquaresDate: 12/11/14 Time: 15:43VariableCoefficientStd. Errort-StatisticProb. LNX10.0024030.0110120.2182250.8295LNX20.0069520.0958090.0725610.9429LNX3-0.0054780.015850-0.3455890.7333C-0.0447291.156156-0.0386880.9695RESID(-1)0.2574590.2310821.1141450.2784R-squared0.058439 Mean dependent var1.07E-16Adjusted R-squared-0.129873 S.D. dependent var0.013658S.E. of regression0.014517 Akaike info criterion-5.450070Sum squared resid0.004215 Schwarz criterion-5.206295Log likelihood73.12588 F-statistic0.310330Durbin-Watson stat1.794969 Prob(F-statistic)0.867655LM检验显示不存在一阶自相关，继续引入X4。LnY与LnX1、LnX2、LnX3、LnX4Dependent Variable: LNYMethod: Least SquaresDate: 12/11/14 Time: 15:32Sample: 1983 2007Included observations: 25VariableCoefficientStd. Errort-StatisticProb. LNX10.3220610.0391618.2239570.0000LNX21.2940010.1353689.5591170.0000LNX3-0.0866650.015730-5.5095090.0000LNX40.0013030.0369720.0352510.9722C-6.0415541.682783-3.5902150.0018R-squared0.978617 Mean dependent var10.70905Adjusted R-squared0.974341 S.D. dependent var0.093396S.E. of regression0.014961 Akaike info criterion-5.389916Sum squared resid0.004476 Schwarz criterion-5.146141Log likelihood72.37395 F-statistic228.8316Durbin