计量经济第三章.ppt
《计量经济第三章.ppt》由会员分享,可在线阅读,更多相关《计量经济第三章.ppt(38页珍藏版)》请在淘文阁 - 分享文档赚钱的网站上搜索。
1、Multiple Regression Analysis:Estimation(2)多元回归分析:估计(2)y=b0+b1x1+b2x2+.bkxk+u1Intermediate Econometrics,Yan ShenChapter Outline 本章大纲nMotivation for Multiple Regression使用多元回归的动因nMechanics and Interpretation of Ordinary Least Squares普通最小二乘法的操作和解释nThe Expected Values of the OLS Estimators OLS估计量的期望值估计量的
2、期望值nThe Variance of the OLS EstimatorsOLS估计量的方差估计量的方差nEfficiency of OLS:The Gauss-Markov TheoremOLS的有效性:高斯马尔科夫定理2Intermediate Econometrics,Yan ShenLecture Outline 课堂大纲nThe MLR.1 MLR.4 Assumptions假定假定MLR.1 MLR.4 nThe Unbiasedness of the OLS estimatesOLS估计值的无偏性估计值的无偏性nOver or Under specification of mo
3、dels模型设定不足或过度设定模型设定不足或过度设定nOmitted Variable Bias遗漏变量的偏误遗漏变量的偏误nSampling Variance of the OLS slope estimatesOLS斜率估计量的抽样方差斜率估计量的抽样方差3Intermediate Econometrics,Yan ShenThe expected value of the OLS estimatorsOLS估计量的期望值nWe now turn to the statistical properties of OLS for estimating the parameters in an
4、 underlying population model.我们现在转向我们现在转向OLS的统计特性,而我们知道的统计特性,而我们知道OLS是估是估计潜在的总体模型参数的。计潜在的总体模型参数的。nStatistical properties are the properties of estimators when random sampling is done repeatedly.We do not care about how an estimator does in a specific sample.统计特性是估计量在随机抽样不断重复时的性质。我们统计特性是估计量在随机抽样不断重复时
5、的性质。我们并不关心在某一特定样本中估计量如何并不关心在某一特定样本中估计量如何。4Intermediate Econometrics,Yan ShenAssumption MLR.1(Linear in Parameters)假定 MLR.1(对参数而言为线性)nIn the population model(or the true model),the dependent variable y is related to the independent variable x and the error u as在总体模型在总体模型(或称真实模型)中,因变量或称真实模型)中,因变量y与自变量
6、与自变量x和误差项和误差项u关系如下关系如下y=b b0+b b1x1+b b2x2+b bkxk+u (3.31)where b b1,b b2,b bk are the unknown parameters of interest,and u is an unobservable random error or random disturbance term.其中,其中,b b1,b b2,b bk 为所关心的未知参数,为所关心的未知参数,u为不可为不可观测的随机误差项或随机干扰项。观测的随机误差项或随机干扰项。5Intermediate Econometrics,Yan ShenAssu
7、mption MLR.2(Random Sampling)假定 MLR.2(随机抽样性)nWe can use a random sample of size n from the population,我们可以使用总体的一个容量为我们可以使用总体的一个容量为n的的随机随机样本样本 (xi1,xi2,xik;yi):i=1,n,where i denotes observation,and j=1,k denotes the jth regressor.其中其中i 代表观察,代表观察,j=1,k代表第代表第j个回归元个回归元nSometimes we write 有时我们将模型写为有时我们将模
8、型写为 yi=b b0+b b1xi1+b b2xi2+b bkxik+ui (3.32)6Intermediate Econometrics,Yan ShenAssumption MLR.3 假定MLR.3nMLR.3(No perfect collinearity)(不存在完全共线性)(不存在完全共线性):In the sample,none of the independent variables is constant,and there are no exact linear relationships among the independent variables.在在样样本本中中
9、,没没有有一一个个自自变变量量是常数,自变量之间也不存在严格的线性关系。是常数,自变量之间也不存在严格的线性关系。nWhen one regressor is an exact linear combination of the other regressor(s),we say the model suffers from perfect collinearity.当当一一个个自自变变量量是是其其它它解解释释变变量量的的严严格格线线性性组组合合时时,我我们说此模型有严格共线性。们说此模型有严格共线性。nExamples of perfect collinearity:完全共线性的例子:完全共
10、线性的例子:y=b b0+b b1x1+b b2x2+b b3x3+u,x2=3x3,y=b b0+b b1log(inc)+b b2log(inc2)+uy=b b0+b b1x1+b b2x2+b b3x3+b b4x4 u,x1+x2+x3+x4=1.7Intermediate Econometrics,Yan ShennPerfect collinearity also happens when y=b b0+b b1x1+b b2x2+b b3x3+u,n(k+1).当当y=b b0+b b1x1+b b2x2+b b3x3+u,n 0Corr(x1,x2)0Positive bia
11、s偏误为正偏误为正Negative bias偏误为负偏误为负b b2 0Negative bias偏误为负偏误为负Positive bias偏误为正偏误为正17Intermediate Econometrics,Yan ShenThe More General Case更一般的情形n Technically,it is more difficult to derive the sign of omitted variable bias with multiple regressors.从技术上讲,要推出多元回归下缺省一个变量时各个变量从技术上讲,要推出多元回归下缺省一个变量时各个变量的偏误方向
12、更加困难。的偏误方向更加困难。n But remember that if an omitted variable has partial effects on y and it is correlated with at least one of the regressors,then the OLS estimators of all coefficients will be biased.我们需要记住,若有一个对我们需要记住,若有一个对y有局部效应的变量被缺省,有局部效应的变量被缺省,且该变量至少和一个解释变量相关,那么且该变量至少和一个解释变量相关,那么所有所有系数的系数的OLS估计量
13、都有偏。估计量都有偏。18Intermediate Econometrics,Yan ShenThe More General Case更一般的情形(3.49-3.50)19Intermediate Econometrics,Yan ShenThe More General Case更一般的情形20Intermediate Econometrics,Yan ShenVariance of the OLS EstimatorsOLS估计量的方差 Now we know that the sampling distribution of our estimate is centered aroun
14、d the true parameter。现在我们知道估计值的样本分布是以真实参数现在我们知道估计值的样本分布是以真实参数为中心的。为中心的。Want to think about how spread out this distribution is我们还想知道这一分布的分散状况。我们还想知道这一分布的分散状况。Much easier to think about this variance under an additional assumption,so 在一个新增假设下,度量这个方差就容易多了,有:在一个新增假设下,度量这个方差就容易多了,有:21Intermediate Econom
15、etrics,Yan ShenAssumption MLR.5(Homoskedasticity)假定MLR.5(同方差性)Assume Homoskedasticity:同方差性假定:同方差性假定:Var(u|x1,x2,xk)=s s2.Means that the variance in the error term,u,conditional on the explanatory variables,is the same for all combinations of outcomes of explanatory variables.意思是,不管解释变量出现怎样的组合,误差项意思是
16、,不管解释变量出现怎样的组合,误差项u的的条件方差都是一样的。条件方差都是一样的。If the assumption fails,we say the model exhibits heteroskedasticity.如果这个假定不成立,我们说模型存在异方差性。如果这个假定不成立,我们说模型存在异方差性。22Intermediate Econometrics,Yan ShenVariance of OLS(cont)OLS估计量的方差(续)n Let x stand for(x1,x2,xk)用用x表示表示(x1,x2,xk)n Assuming that Var(u|x)=s s2 als
17、o implies that Var(y|x)=s s2 假定假定Var(u|x)=s s2,也就意味着也就意味着Var(y|x)=s s2n Assumption MLR.1-5 are collectively known as the Gauss-Markov assumptions.假定假定MLR.1-5共同被称为高斯马尔科夫假定共同被称为高斯马尔科夫假定23Intermediate Econometrics,Yan ShenTheorem 3.2(Sampling Variances of the OLS Slope Estimators)定理定理 3.2(OLS斜率估计量的抽样方差
18、)斜率估计量的抽样方差)24Intermediate Econometrics,Yan ShenInterpreting Theorem 3.2对定理对定理3.2的解释的解释n Theorem 3.2 shows that the variances of the estimated slope coefficients are influenced by three factors:定理3.2显示:估计斜率系数的方差受到三个因素的影响:nThe error variance误差项的方差n The total sample variation总的样本变异n Linear relationshi
19、ps among the independent variables解释变量之间的线性相关关系25Intermediate Econometrics,Yan ShenInterpreting Theorem 3.2:The Error Variance对定理对定理3.2的解释(的解释(1):误差项方差):误差项方差nA larger s s2 implies a larger variance for the OLS estimators.更大的更大的s s2意味着更大的意味着更大的OLS估计量方差。估计量方差。nA larger s s2 means more noises in the e
- 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 第三
限制150内