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

st: alternatives to fixed effects for panel data

From   Lloyd Dumont <[email protected]>
To   [email protected]
Subject   st: alternatives to fixed effects for panel data
Date   Sun, 23 Dec 2007 07:42:02 -0800 (PST)

Hello.  My data structure presents what must be a
common problem.

I have quarterly outcomes data on 15 units.  The panel
is unbalanced.  The average number of observations per
unit is about 20.  But, just about all of the
explanatory variables that I have are time-constant. 
So, at the moment, I am doing something like...


I think a fixed-effects specification--xtreg, fe--is
called for, but it, of course, precludes my analysis
of the variables I care about.  This seems like it
must come up a lot in situations in which one has a
small number of units/clusters, but each with a
reasonable (.e., more than one) number of observations

Can anyone suggest an alternative, perhaps non- or
semi-parametric way to address change over time in
these data?

Thank you!  Lloyd.

Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index