Date:         Sun, 4 Mar 2012 17:28:30 -0500
Reply-To:     Shaojie Cui <shaojiec@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Shaojie Cui <shaojiec@GMAIL.COM>
Subject:      Panel data analysis random effect and binomial distribution
Content-Type: text/plain; charset=ISO-8859-1

Hi there,

I am conducting panel data analysis with random effects. My data consist
of information of 65 individuals over 15 years. I want to do the same
analysis shown below in proc panel. But because the dependent variable is
count data with overdispersion, I need to use a negative binomial
distribution. Can anyone tell me which proc I should use? I couldn’t find
a way to specify a distribution in proc panel. I also need to get Hausmann
test for selecting random effects.

PROC PANEL DATA=inno;
id id year;
MODEL y= x1 x2 x1*x2 /ranone bp ;run;

Thank you very much in advance!

Anna