Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: (fixed effect) sigma_u & s.d. of u


From   Man Jia <jia.m@neu.edu>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: (fixed effect) sigma_u & s.d. of u
Date   Fri, 5 Sep 2008 23:56:55 -0400 (EDT)

Hi Mr. Baum,

Thanks very much for your wonderful example! I find I totally  forget 
the step  "egen oneu=tag(company)". Thanks for pointing out this point. 
Now, I understand why I got sd of u is not the same as sigma_u.

Do you mind helping me with a related question? In your example, I 
tried to predict e(i,t):
.predict e, e
.sum e

The result is 
 
    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
         e |       200   -3.27e-08    51.28882  -184.0086   250.7097

while sigma_e =  52.767964

Could I think 51.28882  approximately equals 52.767964?

Thanks very much! I really parreaite it.

Man Jia
------------------------------------------------------------------------
--

Kit Baum wrote:


>< >
>Man Jia said
>Sorry to bother you with a very simple question. I just can't figure
>out why in my xtreg,fe output is value for sigma_u and the value for
>sd. of predicted u is not the same. I'm thinking I must have got lost
>with some simple thing. Could anyone give some hint? Thanks in advance!
>
>At the bottom of the result from --xtreg,fe  stata reports sigma_u. I'm
>thinking it means the standard deviation of u(fixed effect). Please
>correct me if I'm wrong here. Then, I use --predict u,u to get variable
>u. But the standard deviation of u is not the same as sigma_u. I think
>either I misunderstand sigma_u's meaning or predicted u's meaning. Any
>comments is highly appreciated.
>
>
>webuse grunfeld
>xtreg invest mvalue kstock,fe
>predict u, u
>egen oneu = tag(company)
>su u if oneu
>
>There are only ten unique values of u_i. Their s.d. is indeed the same  
>as that reported by sigma_u.
>
>Kit Baum, Boston College Economics and DIW Berlin
>http://ideas.repec.org/e/pba1.html
>An Introduction to Modern Econometrics Using Stata:
>http://www.stata-press.com/books/imeus.html
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index