|Date: ||Tue, 16 Feb 2010 16:49:26 -0500|
|Reply-To: ||Anna Cui <cuis@RPI.EDU>|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||Anna Cui <cuis@RPI.EDU>|
|Subject: ||Unbalanced panel regression|
I am trying to analyze a dataset with hundrads of firms and three years.
But for some firms there is only data for one time point (year). I know
proc tscsreg only deals with balanced panel, and proc panel requires at
least two time periods for each firm. what procedure can I use to do fixed
and random effects?
Thank you very much in advance.