- 1、本文档共77页,可阅读全部内容。
- 2、原创力文档(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
- 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载。
- 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
经济计量课件2
周一讲座(第一讲)
Conceptual issues
Eight steps in econometric analysis
Revision of important distribution
Introduction to OLS technique
Classical assumptions (A1-A2)
第二讲
Classical assumptions (A3-A10)
估计参数值可靠性 (reliability)
估计参数特性(properties)
Goodness of fit
Three-variable models
2
A3: zero mean value (零均值假定) of
disturbance ?i
Given the value of Xi, the mean or
expected value of the disturbance term
?i is zero: E(?i |Xi ) = 0
3
A4: Equal variance (同方差假定) of ?i
(homoscedasticity )
?
Given the value of Xi, the variance of ?i
is the same for all observations.
2
That is, Var(?i) = E (?i - E(?i))
2 2
= E(?i ) = ?
4
PDF (?)
Y
X
X1 X2 X3
Homoscedasticity (equal variance).
(equal spread around the means)
5
PDF (?)
Y
X
X1 X2 X3
Heteroscedasticity (异方差 unequal
variance)
6
A5. (无自相关假定) No serial correlation or
autocorrelation between the disturbances
Given any two XiXj, Cov(?i ?j) =0 (i
? j). (covariance 协方差) That is, the
correlation between any two ?i and ?j (i
? j) is zero. Alternatively, ?i and ?j are
uncorrelated.
7
? ? ?
j j j
? ?
? ? ? ? ?
? ?
文档评论(0)