[Limdep Nlogit List] Simulations for latent class model

[William Greene]

Manos. It is not possible to do the analysis for the classes
separately because the classes are not known. (They are
"LATENT" after all.)  The probabilities used in the simulations
are the poseriors based on averaging the estimates across the
class estimates - probabilistically.  Since there is only a 
single market (data set), the simulation results give the program's
best estimate of what happens in that market when you make the
changes specified in your command.
/B. Greene

----- Original Message -----
From: "Emmanouil Mentzakis" <e.mentzakis at abdn.ac.uk>
To: limdep at limdep.itls.usyd.edu.au
Sent: Thursday, April 24, 2008 4:27:09 PM (GMT-0500) America/New_York
Subject: [Limdep Nlogit List] Simulations for latent class model

Dear list, 

I am estimating a three class Latent Class Model (for a discrete choice
experiment). Following the estimation I would like to do some
simulations. 

However, the output of the simulations does not distinguish among
classes but it produces only one result (instead of three). 

Is it possible to have the results for all three classes separately? If
not, what exactly does this single output indicate?

The command I use is similar to this: 

nlogit
;lhs=choice2
;choices=alt1, alt2, alt3
;Rhs= cons,x1,x2,x3,x4,x5 
;LCM=z1,z2
;Pts=3; Pds=8;Par
;Simulation
;Scenario: x2(alt1,alt2)=10 & x2(alt1,alt2)=35
$

Thank you in advance
Manos


The University of Aberdeen is a charity registered in Scotland, No SC013683.
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au


-- 
Professor William Greene
Department of Economics
Stern School of Business
New York University
44 West 4th St., Rm. 7-78
New York, NY   10012
http://www.stern.nyu.edu/~wgreene
212.998.0876