GLLAMM and XTMIXED are capable of estimating a wide-range of models, but in general, these models are in the random effects family and assume that c is orthogonal to x. I am not sure what you mean by "I have too few observations". There is nothing in particular about fixed effects models that require more observations than OLS or random effects (assuming that you have at least two time periods of data on many/most individuals/firms).
Steve
-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of Shuaizhang
Feng
Sent: Friday, October 14, 2005 8:17 AM
To: [email protected]
Subject: st: random effects, fixed effects and mixed effects models
Dear all:
I have a general question on what to choose with a
panel data model like this:
y(it)=x(it)b+c(i)+v(it)
If I use xtreg to estimate this, when using random
effects models (re), then I am assuming that c is
orthogonal to x, right? But that is some assumption I
am reluctant to make. I do not want to use fixed
effects estimates too because I have too few
observations.
So could I use GLLAMM or XTMIXED. I suppose in those c
is not assumed to be orthogonal to x. Is that true?
also, if both GLLAMM and XTMIXED work, which one
should I choose? I may have to use a two level mixed
effects model like this one.
y(ijt)=x(ijt)b+c1(ij)+c2(i)+v(ijt).
Thanks a lot.
SZ
__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
