作者bcs (= ="frailty..gggg XD)
看板Economics
标题[请益] 计量的causality
时间Sat Oct 17 14:44:22 2009
出处:
http://tinyurl.com/ykbawo6
reg lines from economic data often cannot be given a causal interpretation.
The reason being that in the relation of interest between observables
and unobservables we might expect they are correlated, whereas in a
^^^^^^^^^^^^
reg model regressors and unobservables are uncorrelated by construction.
请问黄字文意为何?
y=xb+e
回归假设之一： E[xe]=0 为了得到b的认定.
不懂的是为何 E[xe]!=0 会是expect的条件为了casuality的关系？
谢谢^^"
--
※ 发信站: 批踢踢实业坊(ptt.cc)
◆ From: 140.112.86.136
※ bcs:转录至看板 Statistics 10/17 14:45
1F：推 JamesChen:Non stochastic 也可以解释 10/17 15:06
2F：推 ichibond3:我记得greene的附录c有证明 10/19 20:14