計(jì)量實(shí)驗(yàn)報(bào)告

第6頁共11頁《計(jì)量經(jīng)濟(jì)學(xué)》實(shí)驗(yàn)報(bào)告實(shí)驗(yàn)時(shí)間：2012.系別：經(jīng)濟(jì)管理系專業(yè)班級：09級國際經(jīng)濟(jì)與貿(mào)易學(xué)號(hào)：200901901053姓名：楊文夏成績：【實(shí)驗(yàn)名稱】計(jì)量經(jīng)濟(jì)學(xué)【實(shí)驗(yàn)?zāi)康摹酷槍?shí)際問題建立、估計(jì)、檢驗(yàn)和應(yīng)用計(jì)量經(jīng)濟(jì)學(xué)模型的方法，掌握EVIEWS的使用，提高應(yīng)用計(jì)量經(jīng)濟(jì)學(xué)建立模型解決問題的實(shí)踐動(dòng)手能力。通過實(shí)驗(yàn)使我們深入、直觀地理解和掌握計(jì)量經(jīng)濟(jì)學(xué)的模型及方法?！局R(shí)準(zhǔn)備】計(jì)量經(jīng)濟(jì)學(xué)中一元回歸分析、多元回歸分析、非線性最小二乘法進(jìn)行估計(jì)、異方差分析、自相關(guān)分析、多重共線性分析、工具變量法、格蘭杰因果檢驗(yàn)的模型及方法?！緦?shí)驗(yàn)軟件】EVIEWS軟件【實(shí)驗(yàn)要求】要求我們能對一般的實(shí)際經(jīng)濟(jì)問題運(yùn)用計(jì)量經(jīng)濟(jì)學(xué)的方法進(jìn)行分析研究，掌握計(jì)量經(jīng)濟(jì)學(xué)軟件EVIEWS?！緦?shí)驗(yàn)方案與進(jìn)度】【實(shí)驗(yàn)過程】一．p54例子2.6DependentVariable:YMethod:LeastSquaresDate:04/08/12Time:09:55Sample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C283.7574715.16260.3967730.6944X1.3380070.08197816.321550.0000R-squared0.901826Meandependentvar11524.98AdjustedR-squared0.898440S.D.dependentvar3365.226S.E.ofregression1072.445Akaikeinfocriterion16.85561Sumsquaredresid33353986Schwarzcriterion16.94813Loglikelihood-259.2620F-statistic266.3928Durbin-Watsonstat1.864727Prob(F-statistic)0.000000Yi=283.7574+1.338007Xi(0.396773)(16.32155)=0.901826S.E.=1072.445F=266.3928二．p56例2.6.2一元回歸分析DependentVariable:YMethod:LeastSquaresDate:04/08/12Time:10:02Sample:129Includedobservations:29VariableCoefficientStd.Errort-StatisticProb.C2091.645335.29406.2382420.0000X0.4375410.00930647.015370.0000R-squared0.987933Meandependentvar14855.72AdjustedR-squared0.987486S.D.dependentvar9472.076S.E.ofregression1059.615Akaikeinfocriterion16.83567Sumsquaredresid30315159Schwarzcriterion16.92997Loglikelihood-242.1172F-statistic2210.445Durbin-Watsonstat0.276696Prob(F-statistic)0.000000Yi=2091.645+0.437541X三．P61習(xí)題DependentVariable:YMethod:LeastSquaresDate:04/08/12Time:10:09Sample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C2244.650970.93272.3118490.0281X10.702471.1158629.5912090.0000R-squared0.760313Meandependentvar8891.126AdjustedR-squared0.752048S.D.dependentvar7604.152S.E.ofregression3786.469Akaikeinfocriterion19.37860Sumsquaredresid4.16E+08Schwarzcriterion19.47111Loglikelihood-298.3682F-statistic91.99130Durbin-Watsonstat1.562204Prob(F-statistic)0.000000Yi=2244.650+10.70247Xi四．p85-87例3.5DependentVariable:LOG(Q)Method:LeastSquaresDate:04/08/12Time:10:40Sample:122Includedobservations:22VariableCoefficientStd.Errort-StatisticProb.C5.4483110.07904168.930420.0000LOG(X)0.5217010.03716114.038860.0000LOG(P0)0.0338140.1043500.3240460.7496LOG(P1)-0.5244280.104549-5.0160990.0001R-squared0.975002Meandependentvar7.493909AdjustedR-squared0.970835S.D.dependentvar0.193147S.E.ofregression0.032985Akaikeinfocriterion-3.822567Sumsquaredresid0.019584Schwarzcriterion-3.624196Loglikelihood46.04824F-statistic234.0170Durbin-Watsonstat0.540020Prob(F-statistic)0.000000LnQ=5.448311+0.521701lnX-0.524428lnP1+0.033814lnP0(0.033814)(14.03886)(-5.016099)(0.324046)五．p91例3.5.DependentVariable:QMethod:LeastSquaresDate:04/14/12Time:08:40Sample:122Includedobservations:22Convergenceachievedafter11iterationsQ=EXP(C(1))*X^C(2)*P1^C(3)*P0^C(4)CoefficientStd.Errort-StatisticProb.C(1)5.4505730.07740870.413620.0000C(2)0.5445210.03328616.358720.0000C(3)-0.4705420.110634-4.2531380.0005C(4)-0.0639810.109935-0.5819920.5678R-squared0.978405Meandependentvar1830.000AdjustedR-squared0.974806S.D.dependentvar365.1392S.E.ofregression57.95725Akaikeinfocriterion11.12025Sumsquaredresid60462.76Schwarzcriterion11.31863Loglikelihood-118.3228Durbin-Watsonstat0.620897六．p116例4.1.4異方差分析DependentVariable:LOG(Y)Method:LeastSquaresDate:05/09/12Time:14:45Sample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C3.3762351.0407093.2441700.0030LOG(X1)0.1392410.1092801.2741750.2131LOG(X2)0.4735290.0513599.2199540.0000R-squared0.777710Meandependentvar7.928613AdjustedR-squared0.761832S.D.dependentvar0.355750S.E.ofregression0.173615Akaikeinfocriterion-0.572190Sumsquaredresid0.843978Schwarzcriterion-0.433417Loglikelihood11.86895F-statistic48.98065Durbin-Watsonstat1.769958Prob(F-statistic)0.000000e2=（resid^2殘差平方對解釋變量X的散點(diǎn)圖主要分布在圖形中的下三角部分，大致看出殘差平方隨的變動(dòng)呈增大的趨勢，因此，模型很可能存在異方差但是否確實(shí)存在異方差還應(yīng)通過更進(jìn)一步的檢驗(yàn)Goldfeld-Quanadt檢驗(yàn)：子樣本1：DependentVariable:LOG(Y)Method:LeastSquaresDate:04/14/12Time:10:03Sample:112Includedobservations:12VariableCoefficientStd.Errort-StatisticProb.C3.4023341.0497123.2412080.0101LOG(X1)0.3985360.0762115.2293610.0005LOG(X2)0.1968720.1041681.8899540.0913R-squared0.752446Meandependentvar7.700532AdjustedR-squared0.697434S.D.dependentvar0.156574S.E.ofregression0.086125Akaikeinfocriterion-1.853715Sumsquaredresid0.066758Schwarzcriterion-1.732489Loglikelihood14.12229F-statistic13.67785Durbin-Watsonstat1.247545Prob(F-statistic)0.001869子樣本2：DependentVariable:LOG(Y)Method:LeastSquaresDate:04/14/12Time:10:05Sample:2031Includedobservations:12VariableCoefficientStd.Errort-StatisticProb.C3.9936441.8840542.1197080.0631LOG(X1)-0.1137660.159962-0.7112050.4950LOG(X2)0.6201680.1116545.5543800.0004R-squared0.876931Meandependentvar8.239746AdjustedR-squared0.849582S.D.dependentvar0.375812S.E.ofregression0.145754Akaikeinfocriterion-0.801478Sumsquaredresid0.191197Schwarzcriterion-0.680251Loglikelihood7.808868F-statistic32.06485Durbin-Watsonstat2.565362Prob(F-statistic)0.000080F=0.191197/0.0667582=2.864031在5%與10%的顯著性水平下，自由度為（9，9）的F分布的臨界值分別為F0.05=3.18與F0.10=2.44因此5%顯著性水平下不拒絕兩組子樣方差相同的假設(shè)，但在10%的顯著性水平下拒絕。White檢驗(yàn)：WhiteHeteroskedasticityTest:F-statistic9.489121Probability0.000036Obs*R-squared20.30232Probability0.001097TestEquation:DependentVariable:RESID^2Method:LeastSquaresDate:04/14/12Time:11:02Sample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C10.143055.4960311.8455230.0768LOG(X1)-2.2322371.119801-1.9934230.0572(LOG(X1))^20.1420910.0580582.4473930.0218(LOG(X1))*(LOG(X2))0.0201420.0415470.4848030.6320LOG(X2)-0.5224450.450305-1.1602030.2569(LOG(X2))^20.0248560.0129651.9172430.0667R-squared0.654914Meandependentvar0.027225AdjustedR-squared0.585896S.D.dependentvar0.041648S.E.ofregression0.026801Akaikeinfocriterion-4.228793Sumsquaredresid0.017957Schwarzcriterion-3.951247Loglikelihood71.54629F-statistic9.489121Durbin-Watsonstat2.064152Prob(F-statistic)0.000036R^2=0.654914懷特統(tǒng)計(jì)量nR^2=31*0.654914=20.302334該值大于5%顯著性水平下，自由度為5的分布的相應(yīng)臨界值0.05=11.07,因此拒絕同方差的原假設(shè)。加權(quán)最小二乘法（WLS）：DependentVariable:LOG(E2)Method:LeastSquaresDate:04/14/12Time:14:45Sample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C95.3071638.573752.4707770.0198LOG(X2)-26.4883510.10050-2.6224790.0140(LOG(X2))^21.7312660.6570502.6349080.0136R-squared0.199293Meandependentvar-5.118537AdjustedR-squared0.142099S.D.dependentvar2.245038S.E.ofregression2.079420Akaikeinfocriterion4.393821Sumsquaredresid121.0717Schwarzcriterion4.532594Loglikelihood-65.10422F-statistic3.484537Durbin-Watsonstat1.500950Prob(F-statistic)0.044528Ln(e2)=95.30716-26.48835lnx2+1.731266(ln(x2))^2于是wi=1/sqr(fi)=1/spr(exp(95.30716-26.48835lnxi2+1.731266(ln(xi2))^2))DependentVariable:LOG(Y)Method:LeastSquaresDate:04/14/12Time:15:11Sample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C7.9431890.079097100.42370.0000W2-0.0005790.001798-0.3220800.7497R-squared0.003564Meandependentvar7.928613AdjustedR-squared-0.030796S.D.dependentvar0.355750S.E.ofregression0.361186Akaikeinfocriterion0.863493Sumsquaredresid3.783203Schwarzcriterion0.956009Loglikelihood-11.38415F-statistic0.103735Durbin-Watsonstat0.411334Prob(F-statistic)0.749703七、p132-134例4.2.1自相關(guān)分析DependentVariable:YMethod:LeastSquaresDate:04/08/12Time:10:02Sample:129Includedobservations:29VariableCoefficientStd.Errort-StatisticProb.C2091.645335.29406.2382420.0000X0.4375410.00930647.015370.0000R-squared0.987933Meandependentvar14855.72AdjustedR-squared0.987486S.D.dependentvar9472.076S.E.ofregression1059.615Akaikeinfocriterion16.83567Sumsquaredresid30315159Schwarzcriterion16.92997Loglikelihood-242.1172F-statistic2210.445Durbin-Watsonstat0.276696Prob(F-statistic)0.000000殘差圖：DependentVariable:EMethod:LeastSquaresDate:04/14/12Time:09:17Sample:129Includedobservations:29VariableCoefficientStd.Errort-StatisticProb.E1.0000002.14E-174.68E+160.0000-18.91E-282.18E-144.08E-141.0000R-squared1.000000Meandependentvar-5.39E-12AdjustedR-squared1.000000S.D.dependentvar1040.521S.E.ofregression1.18E-13Sumsquaredresid3.74E-25Durbin-Watsonstat0.276687八、p140-143例4.3.1多重共線性分析DependentVariable:LOG(Y)Method:LeastSquaresDate:04/14/12Time:15:33Sample:125Includedobservations:25VariableCoefficientStd.Errort-StatisticProb.C-3.4652152.160307-1.6040380.1252LOG(X1)0.3909710.0568446.8779190.0000LOG(X2)1.1945850.1548337.7153090.0000LOG(X3)-0.0800520.017336-4.6177190.0002LOG(X4)-0.0496300.051358-0.9663440.3460LOG(X5)-0.1442080.064034-2.2520420.0363R-squared0.976386Meandependentvar10.70905AdjustedR-squared0.970172S.D.dependentvar0.093396S.E.ofregression0.016130Akaikeinfocriterion-5.210674Sumsquaredresid0.004944Schwarzcriterion-4.918143Loglikelihood71.13342F-statistic157.1223Durbin-Watsonstat1.511523Prob(F-statistic)0.000000Y=-3.465215+0.390971lnX1+1.194585lnX2-0.080052lnX3-0.049630lnX4-0.144208lnX5由于R^2=0.976386較大且接近于1，而且F=157.1223>F0.05(5,19)=2.74故認(rèn)為糧食生產(chǎn)與上述解釋變量間總體線性關(guān)系顯著。但由于其中X4,X5前參數(shù)估計(jì)值未能通過t檢驗(yàn)，而且符號(hào)的經(jīng)濟(jì)意義也不合理，故認(rèn)為解釋變量間存在多重共線性。九、p151-153例4.4.1十、p176例5.2.4格蘭杰因果檢驗(yàn)PairwiseGrangerCausalityTestsDate:05/09/12Time:11:23Sample:129Lags:1NullHypothesis:ObsF-StatisticProbabilityYdoesnotGrangerCauseX286.327410.01868XdoesnotGrangerCauseY15.03410.00068PairwiseGrangerCausalityTestsDate:0