《计量经济学 案例分析 Eviewsxct.docx》由会员分享,可在线阅读,更多相关《计量经济学 案例分析 Eviewsxct.docx(15页珍藏版)》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、一、 研究课题:通过对1984420033年某国GDDP和出口的的分析,研究究GDP和出出口量的相关关关系并对参参数估计值进进行检验。二、 模型及数据来源源:GDP为因变量量,出口量为为自变量。选选择模型是一一元线性回归归模型y=cc0+c1x+u(yy代表GDPP,x代表出出口量,u表表示残差项)数据来自计量经济学软件eviews的使用135页表12.1。提取其进口和国内生产总值两列数据:annualexportgdp1984580.571711985808.98964.419861082.110202.221987147011962.5519881766.714928.3319891956
2、16909.2219902985.818547.9919913827.121617.8819924676.326638.1119935284.834634.44199410421.8846759.44199512451.8858478.11199612576.4467884.66199715160.7774462.66199815233.6678345.22199916159.8882067.55200020634.4489468.11200122024.4497314.88200226947.44105172.3200336287.99117251.9三、 作业1、 根据表格得到曲曲线图、散
3、点点图、X-YY曲线图:曲线图散点图X-Y曲线图2、数据描述统统计分析3、简单的回归归估计Dependeent Vaariablle: GDDPMethod: Leasst SquuaresDate: 006/14/09 Time: 16:338Sample: 19844 20033Includeed obsservattions: 20VariablleCoefficcientStd. Errrort-StatiisticProb. C11772.7772862.41194.11287730.0007EXPORT3.54779900.197911917.9254480.0000R-squar
4、red0.9469553 Meaan deppendennt varr49439.002Adjusteed R-ssquareed0.9440006 S.DD. deppendennt varr36735.119S.E. off regrressioon8692.6556 Akaaike iinfo ccriterrion21.072998Sum squuared residd1.36E+009 Schhwarz criteerion21.172556Log likkelihoood-208.72298 F-sstatisstic321.32229Durbin-Watsoon staat0.6
5、049771 Proob(F-sstatisstic)0.0000000根据输出结果,写写出OLS估估计式,并进进行分析说明明:yt=-117772.777+3.5477790xt R2=0.9466953 df=188检验回归系数显显著性的原假假设和备择假假设是(给定定a = 0.05)H0:c1= 0; H11:c1 0。因为t = 117.925548 tt0.0255 (18) = 2,所所以检验结果果是拒绝c1=0,即认认为进口额和GDDP之间存在在回归关系,二二者正方向变变化。上述模模型的经济解解释是,对于于出口量每增加加1亿元,GGDP将平均均增加3.554779亿亿元。拟合优优
6、度为0.99469533说明上式的的拟合情况较较好。GDPP变动的94.7%可以由出口量量的变动解释释。 4、自相关及其其解决残差:残差序列列图717113832.225806552-6661.2258065521| . * | . |8964.414642.557332886-5678.1173328857| . * | . |10202.2215611.882958993-5409.6629589929| . * | . |11962.5516988.001737667-5025.5517376674| . * | . |14928.3318040.664670553-3112.33467
7、05528| . * | . |16909.2218712.224337449-1803.0043374495| . *| . |18547.9922365.775764003-3817.8857640028| . * | . |21617.8825350.55134688-3732.7713468804| . * | . |26638.1128363.229683777-1725.1196837774| . *| . |34634.4430522.1127125564112.2772874338| . | * . |46759.4448747.112497008-1987.772497008
8、4| . *| . |58478.1155949.113891442528.9661086004| . |* . |67884.6656391.11935622911493.440643771| . | .* |74462.6665559.7747569948902.8552430661| . | * |78345.2265818.33814699112526.881853009| . | .* |82067.5569104.3344678812963.11553222| . | . * |89468.1184979.2286347794488.8113652006| . | * . |973
9、14.8889910.7714614447404.0885385558| . | *. |105172.3107376.4853774-2204.1185374401| . *| . |117251.9140514.6189888-23262.71898877|* . | . |由图看出残差具具有较明显的的自相关趋势势,同时由简简单回归估计计的D-W值值0.6044971,远远小于2,也也可推出模型型存在自相关关可能。AR(1)模型型的估计消除自相关的回回归分析Dependeent Vaariablle: GDDPMethod: Leasst SquuaresDate: 006/14/09 T
10、ime: 15:446Sample(adjussted): 19855 20033Includeed obsservattions: 19 aafter adjussting endpoointsConverggence achieeved aafter 30 itteratiionsVariablleCoefficcientStd. Errrort-StatiisticProb. C-80417.9387670.331-0.91722770.3726EXPORT0.79691100.34276642.32495550.0336AR(1)1.03613310.028188436.7636690
11、.0000R-squarred0.9947880 Meaan deppendennt varr51663.665Adjusteed R-ssquareed0.9941227 S.DD. deppendennt varr36331.334S.E. off regrressioon2784.2559 Akaaike iinfo ccriterrion18.845229Sum squuared residd1.24E+008 Schhwarz criteerion18.994441Log likkelihoood-176.03303 F-sstatisstic1524.4449Durbin-Wats
12、oon staat0.5602994 Proob(F-sstatisstic)0.0000000Inverteed AR Rootss 1.04Estimatted ARR proccess iis nonnstatiionaryy经过GLS处理理以后,我们们可以看出GG-W的值由由原来的0.6049771改进为00.5602294,基本本上消除了自自相关性。GDP = -804177.931881 + 00.796991049337*EXPPORT + AR(1)=1.03613305915、异方差性及及其修正先看散点图:由图可以看出,残残差随着GDDP的增大其其分散程度也也增大,这是是存在
13、异方差差性的初步经经验证据。怀特检验:White HHeterooskedaasticiity Teest:F-statiistic14.975332 Proobabillity0.0001778Obs*R-ssquareed12.758335 Proobabillity0.0016997Test Eqquatioon:Dependeent Vaariablle: REESID22Method: Leasst SquuaresDate: 006/14/09 Time: 23:222Sample: 19844 20033Includeed obsservattions: 20VariablleC
14、oefficcientStd. Errrort-StatiisticProb. C4195547773179760011.31945540.2045EXPORT-7517.77835261.1445-1.42899250.1711EXPORT20.50613300.16177753.12861120.0061R-squarred0.6379118 Meaan deppendennt varr680060339Adjusteed R-ssquareed0.5953220 S.DD. deppendennt varr1.23E+008S.E. off regrressioon780134771 A
15、kaaike iinfo ccriterrion39.320114Sum squuared residd1.03E+117 Schhwarz criteerion39.469550Log likkelihoood-390.20014 F-sstatisstic14.975332Durbin-Watsoon staat1.7370444 Proob(F-sstatisstic)0.0001778辅助回归模型中中,取显著性性水平a=00.05,由由于Obs*R-squuared=12.755835Xa/20.055(10)=18.331,所以函函数不存在异异方差性。由由输出结果的的概率值(PP值)
16、可以看看出,函数不不存在异方差差性。因为存在异方差差性,OLSS所估计出来来的参数标准准有误,我们们采用怀特法法重新估计参参数解决这一一问题。Dependeent Vaariablle: GDDPMethod: Leasst SquuaresDate: 006/14/09 Time: 20:008Sample: 19844 20033Includeed obsservattions: 20White HHeterooskedaasticiity-Coonsisttent SStandaard Errrors & CovvariannceVariablleCoefficcientStd. Err
17、rort-StatiisticProb. C11772.7772757.30044.26966660.0005EXPORT3.54779900.342644110.3542260.0000R-squarred0.9469553 Meaan deppendennt varr49439.002Adjusteed R-ssquareed0.9440006 S.DD. deppendennt varr36735.119S.E. off regrressioon8692.6556 Akaaike iinfo ccriterrion21.072998Sum squuared residd1.36E+009 Schhwarz criteerion21.172556Log likkelihoood-208.72298 F-sstatisstic321.32229Durbin-Watsoon staat0.6049771 Proob(F-sstatisstic)0.0000000GDP = 111772.7659 + 3.5547790012*EXXPORT6、总结研究发现,出口口量与国民生生产总值(GGDP)存在在正相关关系系,即可以通通过各种优惠惠政策刺激出出口来有效地地促进GDPP的增长。同同时出口量存存在明显的自自相关性,也也就是说当年年的出口量受受前几年出口量的的影响很大。