[Limdep Nlogit List] LCRPLOGIT question

Sat Nov 25 08:29:14 AEDT 2023

One suggestion
You could treat each LC as a decision process like we do with ANA by allowing a parameter for any climate related variables to be non zero in all but one class and see what you get. We no longer call these latent classes because you behaviorally define a class and hence decision processes. This is explained in ACA 2015 book.

David
I am on leave until early December.

Sent from my iPhone
+61 (0) 418 433 057
Professor David A Hensher AM, PhD, FASSA

Choice modellers prefer Nlogit: See https://protect-au.mimecast.com/s/mbkxCoV1kpfDYwRBYu1P-ec?domain=limdep.com
ITLS celebrating 33 years

Note:
David.hensher at bigpond.com
David.hensher at sydney.edu.au

I am no longer using hgroup at hensher.com.au
and any correspondence to this email needs to be replaced with Hgroup<David.hensher at bigpond.com>
These email addresses are linked so use one only

On 25 Nov 2023, at 7:14 am, Kolady, Deepthi Elizabeth P via Limdep <limdep at mailman.sydney.edu.au> wrote:

Dear Professor Greene,

Thank you very much for the quick response and the clarification on LCRPLOGIT and RH2 command. In addition to demography, I want to include climate change perceptions as another choice invariant variable.

I I want to estimate class-wise WTP estimates for various carbon market attributes in one step. If I understand your response and the manual correctly, LCRPLOGIT allows me to do that. Am I right? Sorry for not communicating it clearly earlier.

Thanks again for your timely help.

Best regards,
Deepthi

-----Original Message-----
From: Limdep <limdep-bounces at mailman.sydney.edu.au> On Behalf Of William Greene via Limdep
Sent: Friday, November 24, 2023 2:50 PM
To: Limdep and Nlogit Mailing List <limdep at mailman.sydney.edu.au>
Cc: William Greene <wgreene at stern.nyu.edu>
Subject: Re: [Limdep Nlogit List] LCRPLOGIT question

[You don't often get email from limdep at mailman.sydney.edu.au. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]

Deepthi. LCRPLOGIT is for latent classes and random parameters. Your description sounds like you only want LCLOGIT - latent classes.  Note that you can't separate observations into classes, by construction the classes are latent.  You can, post estimation, make your best guess as to who is in which class - please see the NLOGIT manual for details.  RH2 is not the latent class determinants.
RH2 is for choice invariant variables such as Age or Income.  To make the class probabilities conditional, you will use ;LCM = list ... where the variables should be choice invariant as well.
Regards
B. Greene

On Fri, Nov 24, 2023 at 3:24 PM Kolady, Deepthi Elizabeth P via Limdep < limdep at mailman.sydney.edu.au> wrote:

Hello Professor Greene,

I am interested in examining how climate change perceptions and
adverse weather experiences influence farmers' willingness to
accept(WTA) carbon payments. I have discrete choice data to estimate
WTA estimates. However, I would like to classify farmers into latent
classes based on climate change perception and adverse weather
experience and then estimate utility preferences and WTA per class.

My first question is can I use LCRPLOGIT command to do it one step
instead of doing it separately (i.e. form classes and then estimate).
I am thinking of the following command

LCRPLOGIT ; Lhs = dependent variable

; Choices = the names of the J alternatives

; Rhs = list of choice specific attributes

; Rh2 = list of choice invariant individual characteristics

; Fcn = definition of the random parameters part

; Pts = the number of classes $

My second question is what is Rh2?  I assume it is the latent class
determinants (i.e. climate change perception, adverse weather
experience etc).

Thanks,
Deepthi

_______________________________________________
Limdep site list
Limdep at mailman.sydney.edu.au
http://limde/
p.itls.usyd.edu.au%2F&data=05%7C01%7Cdeepthi.kolady%40okstate.edu%7C66
eb2145f9554c7b877608dbed2f075d%7C2a69c91de8494e34a230cdf8b27e1964%7C0%
7C0%7C638364558478225530%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiL
CJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=71
E%2FeTAiph6qxL%2FKOOzMMQpQ7Qvzf%2FFNeKSXibI%2By58%3D&reserved=0

--
William Greene
Department of Economics, emeritus
Stern School of Business, New York University
URL: https://protect-au.mimecast.com/s/gOZSCp81lrt9owWvosD1lCf?domain=people.stern.nyu.edu
Email: wgreene at stern.nyu.edu
Editor in Chief: Foundations and Trends in Econometrics Associate Editor: Economics Letters Associate Editor: Journal of Business and Economic Statistics _______________________________________________
Limdep site list
Limdep at mailman.sydney.edu.au
http://limdep.itls.usyd.edu.au/

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