计量经济学(英文).pptx
- 文档编号:18709954
- 上传时间:2023-10-13
- 格式:PPTX
- 页数:413
- 大小:1.61MB
计量经济学(英文).pptx
《计量经济学(英文).pptx》由会员分享,可在线阅读,更多相关《计量经济学(英文).pptx(413页珍藏版)》请在冰点文库上搜索。
LectureOne,MethodologyofEconometrics,1,立论?
结论?
立论:
要求给出求论的路径。
结论:
要求说明结论的来源。
自以为是的东西并不见得是真我们不是上帝!
2,我们的习惯是这样的吗?
结论来自感觉(象上帝)宏观思考(象战略家)习惯地提出政策建议(象顾问)得争取把一个个的大、小问题搞明白再说吧!
3,MainstreamAnalysisApproaches,NormativeAnalysisPositiveAnalysis(empiricalanalysis),4,TheWriterD.N.Gujarati,ProfessorofeconometricsattheMilitaryAcademyatWestPointMasterofCommerceMBAEditorialrefereeAuthorVisitingProfessor,5,WhatisEconometrics,EmpiricalsupporttothemodelsQuantitativeanalysisofactualeconomicphenomenaSocialscienceinwhichthetoolsofeconomictheory,mathematics,andstatisticalinferenceareappliedtotheanalysis.PositivehelpEconomictheory_measurements,6,MethodologyofEconometrics,StatementoftheoryorhypothesisObtainingthedataSpecificationofthemathematicalmodelSpecificationoftheeconometricmodelEstimationoftheparametersoftheeconometricmodelHypothesistestingForecastingorpredictionUsingthemodelforcontrolorpolicypurposes,7,StatementofTheoryorHypothesis,Postulate(givesomeexamples)StatementNote:
hypothesisisnotthesameasanassumption,8,ObtainingtheData,NatureSourcesLimitations,9,TypesofData,Timeseriesdata:
quantitative,qualitative(dummyvariable)(SATIONARY)Cross-sectionaldata:
(HETEROGENEITY)Pooleddata:
(Paneldata),10,Sources,11,AccuracyofData,Non-experimentalinnatureRound-offsandapproximationsNon-responseSelectivitybiasAggregatelevelConfidentialityTheresultsofresearchareonlyasgoodthequalityofthedata.,12,Specificationofthemathematicalmodel,Yi=b1+b2*Xi0b21YiconsumptionexpenditureXiincomeParametersb1b2,b1intercept,b2slopeConsumptionfunctionSingleequationmodelMultipleequationmodel,13,Terminology,YiDependentvariableExplainedvariablePredictandRegressandResponseEndogeousTargetvariable,XiIndependentvariableExplanatoryvariablePredictorRegressorStimulusvariableExogenousvariableControlvariable,14,Specificationoftheeconometricmodel,Yi=b1+b2*Xi+uiUidisturbance,errorterm,random(stochasticvariable)EconometricmodelLinearregressionmodel,15,SignificanceofDisturbance:
ui,SurrogateforallomittedvariablesVaguenessoftheoryUnavailabilityofdataCorevariablesVS.peripheralvariablesIntrinsicrandomnessinhumanbehaviorPoorproxyvariablesPrincipleofparsimonyWrongfunctionalform,16,LinearRelationship,LinearityinthevariablesE(y|xi)=b1+b2*xi*xiLinearityintheparametersE(y|xi)=b1+root(b2)*xi*xiLRM:
linearregressionmodelNLRM:
non-linearregressionmodel,17,RegressionRelationships,RegressionanalysisisconcernedwiththestudyofthedependenceofYiwithoneormoreXi.Xiisknownorfixed,predictingthemeanofYioftotal.Statisticalvs.deterministicrelationshipsRegressionvs.causationRegressionvs.correlation,18,Estimationoftheeconometricmodel,NumericalestimatesoftheparametersRegressionanalysisisthemaintoolusedtoobtaintheestimatesThehatonYindicatesthatitisanestimate,19,HypothesisTesting,Sample-sampleparameter-estimate-estimatordistribution-populationparameter-populationcharacteristicsConfirmationorrefutationofeconomictheoriesonthebasisofsampleevidenceThebasementisstatisticalinference(Hypothesistesting),20,ForecastingorPrediction,HypothesisortheorybeconfirmedKnownorpredictorvariableXPredictthefuturevaluesofthedependent,21,UseoftheModelforControlorPolicyPurposes,ControlvariableXTargetvariableYYi=b1+b2*XiManipulatethecontrolvariableXtoproducethedesiredlevelofthetargetvariableY,22,AnatomyofClassicalEconometricModeling,EconomictheoryMathematicalmodeloftheoryEconometricmodeloftheoryDataEstimationofeconometricmodelHypothesistestingForecastingorpredictionUsingthemodelforcontrolorpolicypurposes,23,24,PleaseGiveSomeSuggestions,Z3-W163.COM027-62082852Thankyou.,25,AReviewofSomeStatisticalConcepts,LectureTwo,26,Samplespace、Samplepoints、Events,Populationisthesetofallpossibleoutcomesofrandomexperiment(samplespace)SamplepointistheeachmemberofthissamplespaceEventisasubsetofthesamplespace,27,ProbabilityandRandomVariables,P(A)probability(pr;p;pro)Xrandomvariable(rv)Xthevalueofarandomvariable0=P(A)=1P(A+B+)=1A+B+C+=samplespaceDiscretevariableContinuousvariable,28,ProbabilityDensityFunction,Discreterv:
f(x)=P(X=xi)(i=1,2,)=0(x=/xi)(PDF)Sum(f(x)=1Continuousrv:
f(x)=0,29,CumulativeDistributionFunction,CDFF(X)=P(X=x)(discrete)=(continuous),30,CDFofDiscreteVR,31,32,CDFofContinuousVR,33,34,MaincontentsExpectedvalue(propertiesofexpectedvalues)Variance(propertiesofvariance)Covariance(propertiesofcovariance)CorrelationcoefficientMomentSkewnessandKurtosis,CharacteristicsofProbabilityDensity,35,ExpectedValue,Expectedvalue(populationmeanvalue)(discretervX,f(x)isthediscretePDFofX;continuousrvV,f(x)isthecontinuousPDFofX)(someexamples),36,PropertiesofExpectedValues,1.Theexpectedvalueofaconstantistheconstantitself.E(b)=b2.Ifaandbareconstants,E(aX+b)=aE(X)+b3.IfXandYareindependentrv,thenE(XY)=E(X)*E(Y)4.IfXisrvwithPDFf(x),g(X)isanyfunctionofX,then,ifXisdiscreteifXiscontinuous,37,Variance,LetXbearvandletE(X)=u,thedistribution,orspread,oftheXvaluesaroundtheexpectedvaluecanbemeasuredbythevariance.isthestandarddeviationofX,38,PropertiesofVariance,1.2.Thevarianceofaconstantiszero.3.Ifaandbareconstants,then4.IfXandYareindependentrv,then5.IfXandYareindependentrv,then,39,Covariance,40,Correlationcoefficient,Thepopulationcorrelationcoefficientrhoisdefinedasisameasureoflinearassociationbetweentwovariablesandliesbetween-1and+1,-1indicatingperfectnegativeassociation+1indicatingperfectpositiveassociation,41,VariancesofCorrelatedVariables,Var(X+Y)=Var(X)+Var(Y)+2Cov(X,Y)Var(X-Y)=Var(X)+Var(Y)-2Cov(X,Y)Var(X+Y+Z)=Var(X)+Var(Y)+Var(Z)+2Cov(X,Y)+2Cov(X,Z)+2Cov(Z,Y),42,Moment,Mean,variance,andcovariancearethemostfrequentlyusedsummarymeasuresofunivariateandmultivariatePDFs.MomentrthmomentE(X-)nE(X)nr=1r=2r=3r=4,43,SkewnessandKurtosis,Skewnesslacksymmetry.(rightS0,symmetricalS=0orleftS0skewed.)Kurtosistallnessorflatness.Leptokurtic:
slim,orlongtailed.Klessthan3Mesokurtic:
normaldistribution.Kis3Platykurtic:
fatorshort-tailed.Kmorethan3,44,Sampleaverageandvariance,45,Samplecovariance,46,Samplecorrelation,47,Sampleskewnessandkurtosis,3thcentermoment4thcentermoment,48,Probabilitydistributions,NormalDistributionStandardNormalDistributiont-DistributionChi-DistributionF-Distribution,49,LectureThreeProbabilityDistribution,NormalDistributionChi-squareDistributioTDistributionFDistribution,50,NormalDistribution,A(continuous)randomvariableXissaidtobenormallydistributedifitsPDFhasthefollowingform:
whereand2,knownastheparametersofthedistribution,are,respectively,themeanandthevarianceofthedistribution.XN(,2),51,ThePropertiesofThisDistribution,Itissymmetricalarounditsmeanvalue.68%oftheareaunderthenormalcurveliesbetweenthevaluesof;95%ofthearealiesbetween2;99.7%ofthearealiesbetween3.,52,StandardizedNormalVariable,weconvertthegivennormallydistributedvariableXwithmeanand2intoastandardizednormalvariableZbythefollowingtransformationAnimportantpropertyofanystandardizedvariableisthatitsmeanvalueiszeroanditsvarianceisunity.XN(0,1),53,AnExample,AssumethatXN(8.4).WhatistheprobabilitythatXwillassumeavaluebetweenX1=4andX2=12?
ComputetheZvaluesFromTableweobservethatPr(0Z2)=0.4772.Then,bysymmetry,Pr(-2Z0)=0.4772.therequireprobabilityis0.4772+0.4772=0.9544.,54,AnExample,WhatistheprobabilitythatintheprecedingexampleXexceeds12?
TheprobabilitythatXexceeds12isthesameasthatZexceeds2.FromTable,itisobviousthatthisprobabilityis0.50.4772)=0.0228.,55,PropertiesofND,LetX1N(1,21)andX2N(2,22)andtheyareindependent.thelinearcombinationY=aX1+bX2,a,bareconstants.ThenitcanbeshownthatYN(a1+b2),(a221+b222),56,CentralLimitTheorem,LetX1,X2,Xn.denotenindependentrandomvariables,allofwhichhavethesamePDFwithmean=andvariance=2.Let=Xi/n(i.e.,thesamplemean).(i.e.,n),57,Thatis,approachesthenormaldistributionwithmeanandvariance2/n.NoticethatthisresultholdstrueregardlessoftheformofthePDF.Asaresult,itfollowsthatZisastandardizednormalvariable.,58,Moments,Thethirdandfourthmomentsofthenormaldistributionaroundthemeanvalueareasoffollows:
E(X-)3=0E(X-)4=34Note:
Allodd-poweredmomentsaboutthemeanvalueofanormallydistributedvariablearezero.,59,SkewnessandKurtosis,ForanormalPDF,Skewness=0andKurtosis=3;Anormaldistributionissymmetricandmesokurtic.AsimpletestofnormalityisJarque-Bera(JB)testofnormality,60,Jarque-Bera(JB)TestofNormality,WhereSstandsforskewnessandKforkurtosis.Underthenullhypothesisofnormality,JBisdistributedasaChi-squarestatisticwith2df.,61,The(Chi-square)Distribution,LetZ1,Z2,ZKbeindependentstandardizednormalvariable.distributionwithkdegreesoffreedom(df),dfmeansthenumberofindependentquantitiesintheprevioussum,Achi-square-distributedvariableisdenotedby,wherethesubscriptkindicatesthedf.,62,PropertiesoftheDistribution,thedistributionisskewed,thedegreeoftheskewnessdependingonthedf.Forcomparativelyfewdf,thedistributionishighlyskewedtotheright;Asthenumberofdfincreases,thedistributionbecomesincreasinglysymmetrical.,63,Themeanofthechi-squaredistributionisk,anditsvarianceis2k,kisthedf.IfZ1andZ2aretwoindependentchi-squarevariableswithk1andk2df,thenthesumZ1+Z2isalsoachi-squarevariablewithdf=-k1+k2.,MeanandVariance,64,AnExample,Whatistheprobabilityofobtainingax2valueof40orgreater,giventhedf
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 计量 经济学 英文