资源描述：
《计量期末复习笔记》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。
1、Stata部分(30^409单选&多选：Stata哪个命令正确？填空：做某件事需要哪个命令？写小程序或者解释小程序：解释某个命令是干什么的计量原理部分(1045，)大题：证明OLS估计的某个性质，写出证明过程其他课上重点讲解的部分课本部分PartlThelinearregressionmodel第一章Anoverview•Threetypesofdata:-TimeseriesdataXtAsetofobservationsthatavariabletakesatdifferenttimes,suchasdaily(e.g.,stockprices),weekly(e.g
2、.,moneysupply),monthly(e.g.,theunemploymentrate),quarterly(e.g.,GDP),annually(e.g.zgovernmentbudgets),quinqueniallyoreveryfiveyears(e.g.zthecensusofmanufactures),ordecenniallyoreverytenyears(e.g.,thecensusofpopulation).TheproblemofautocorrelationAtimeseriesdatasetisstationaryifitsmeanandv
3、arianeedonotvarysystematicallyovertime・-Cross-sectionaldataXfCross-sectionaldataaredataononeormorevariablescollectedatthesamepointintime.ExamplesarethecensusofpopulationconductedbytheCensusBureauevery10years,opinionpollsconductedbyvariouspollingorganizations,andtemperatureatagiventimeinse
4、veralplaces.Theproblemofheterogeneity-Panel,Iongitudinalormicro-paneldataX/tPaneldatacombinesfeaturesofbothcross-sectionandtimeseriesdata.•EightassumptionstheClassicalLinearRegressionModel(CLRM)A・l:Linearintheparameters;itmayormaynotbelinearinthevariablesYandtheXs.系数线'性A-2:Theregressorsar
5、eassumedtobefixedornonstochasticinthesensethattheirvaluesarefixedinrepeatedsampling.回归元X假定为不变的或者非随机的，即其值在重复抽样中是固定的。A-3:GiventhevaluesoftheXvariables,theexpected,orthemean,valueoftheerrortermiszero.Thatis,E(uf
6、X)=0给定X变量的值，其误差项的条件期望为零(知道X就一定知道Y)A-4:Homoscedastic,orconstant,varianceofu^orvar
7、(ujX)=，.给定X的值，每一个血的方差都是不变的，或者说它们具有同方差性(变化有规律)A-5:Noautocorrelation,Thereisnocorrelationbetweentwoerrorterms.两个i吴差项之间不相关，不存在自相关性cov(i^,u；
8、X)=0ijA-6:Nomulticollinearity,ornoperfectlinearrelationshipsamongtheXvariables.A-7:Nospecificationbias.模型设计正确，观察值的数口n要比待估参数的数目多A-8:尽管不是CLRM的一部分,我们还是假定误差
9、项服从零均值和(不变)方差，的正态分布均〜N(0,，)若再增加假设8,在正态性假设下CLRM称为正态经典线性回归模型NCLRM•在A1-7这些假设条件卜,OLS估计量是最佳线性无偏估计量(BLUE:bestlinearunbiasedestimators)(1)EstimatorsarelinearfunctionsofthedependentvariableY.(2)Theestimatorsareunbiased;inrepeatedapplicationsofthemethod,theestimatorsapproach
