[Limdep Nlogit List] Sample selection with ordered probit

[Abdulbaki Bilgic]

 Abdulbaki Bilgic, PhD.,
Associate Professor
*****************************************************
*****************************************************
Current Address:
Harran Üniversitesi, Ziraat Fakültesi,
Tarim Ekonomisi Bölümü, Osmanbey Kampusu,
63040 Sanliurfa/Turkey
Phone :+90 414 312 0020/1527
Webpage:http://bilgic.webs.com
*****************************************************
***************************************************** 

Barry,
Put a semicolon (;) between 'selection' and 'margin' in your procedure. I see there is a dot between them and i think this is a reason not getting marginal effects.
Best wishes

----- Original Message ----
From: Barry Quinn <b.quinn at qub.ac.uk>
To: "limdep at limdep.itls.usyd.edu.au" <limdep at limdep.itls.usyd.edu.au>
Sent: Wed, January 20, 2010 9:47:42 AM
Subject: [Limdep Nlogit List] Sample selection with ordered probit

Hi all,

I am new to DCM estimation so excuse the ignorant question.  I am running the following command and cannot seem to get marginal effects for my selection model, is this due to the nature of the estimation process?.. any suggestions are much appreciated??

namelist    ;x=one,lta,lnta,divers,urban,ilcu,occup,roa,ci,socio
            ;z=one,bband,emp,compedu,prof,geny,female$
create      ;y=web08$
reject      ;web08<1$
create      ;tec=tecscore$
recode      ;tec;1=0;2=1;3=2$
sample      ;All$
Probit      ;lhs=y;rhs=x;hold$
ordered    ;lhs=tec;rhs=z;selection,margin$


regards


Barry Quinn
Queens University Belfast


_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au