[Limdep Nlogit List] Extracting latent class membership - follow-up question
William Greene
wgreene at stern.nyu.edu
Thu Oct 18 20:38:11 EST 2007
Anna-Kaisa. NLOGIT does not reorder the data. However, it is not
possible to make the class specific probabilities into a "panel"
the way you have described, because the class probabilities are
assumed by the estimator to be constant across periods - they are
rather like the random effects in a panel model. However, you can
achieve this effect by ignoring the panel nature of the data and
fitting it as if it were a large cross section, with class probabilities
modeled as functions of the variables you listed - use ;PDS=1.
(By specifying that the class probabilities vary across periods,
that is what you are assuming. Otherwise, it is only the class
probabilities that link one period to another in this model.)
Sincerely,
Bill Greene
----- Original Message -----
From: "Anna-Kaisa Kosenius" <anna-kaisa.kosenius at helsinki.fi>
To: "Limdep and Nlogit Mailing List" <limdep at limdep.itls.usyd.edu.au>
Sent: Thursday, October 18, 2007 2:49:09 AM (GMT-0500) America/Bogota
Subject: [Limdep Nlogit List] Extracting latent class membership - follow-up question
Dear all
can the matrix of case-specific class probabilities (one row per individual)
be transformed into panel data setting (where I have 18 rows per
individual) to further characterize the group membership with socioeconomic
and other factors?
Or is the only way to create a new data set for this analysis, and if yes,
is it so that Nlogit did not reorder the individuals when creating the
matrix?
-Anna-Kaisa
Anna-Kaisa Kosenius
PhD student
Department of Economics and Management
P.O.Box 27
FIN-00014 University of Helsinki
Tel. +358 9 191 58511
anna-kaisa.kosenius at helsinki.fi
_______________________________________________
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
More information about the Limdep
mailing list