[Limdep Nlogit List] random effects probit with clustering
Thornburg, Steven
sthornburg at wsu.edu
Sat Jul 31 07:03:09 EST 2010
Philip Grossman
It depends on how your data is arranged and whether you have balanced or
unbalanced panels
See R6.5 & E17.4
For balanced panels you need to specify the group size
;pds= panel specification
e.g.
;pds=10
Using your variable names from below
probit
;lhs=Z I=N, T=X
;rhs=one,independent variables
;panel
;pds=X
;random effects
;MARGINAL
$
For unbalanced panels you need to create a variable that indicates group
membership
e.g.
;pds=ti
The following is an example of the script I might use for creating a
group count variable for unbalanced panels
sample; all$
noskip$
sort; lhs=Xsessionindex; rhs=*$
create; Xsession=ndx(Xsessionindex,0)$
regress; lhs=one; rhs=one; str=Xsession; panel$
create; Sessionsize=_groupti$
probit
;lhs= lhs=Zti
;rhs=one,independent variables
;panel
;pds=Sessionsize
;random effects
;MARGINAL
$
-------------------------------------
Steven Thornburg, Ph.D. CPA
Assistant Professor
Todd 237F
Department of Accounting
Washington State University
Pullman, WA 99164
Voice: (509) 335-2222
Fax: (509) 335-4275
-----Original Message-----
From: limdep-bounces at limdep.itls.usyd.edu.au
[mailto:limdep-bounces at limdep.itls.usyd.edu.au] On Behalf Of Grossman,
Philip J.
Sent: Friday, July 30, 2010 11:57 AM
To: limdep at limdep.itls.usyd.edu.au
Subject: [Limdep Nlogit List] random effects probit with clustering
I have panel data comprised on X sessions with N subjects per session
each making Z yes/no decisions. I want to estimate a random effects,
probit model with clustering on session. Is this doable with Limdep and
if so how?
Thanks for any help provided.
Philip Grossman
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au
More information about the Limdep
mailing list