Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: markov transition model

From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: markov transition model
Date   Mon, 17 Apr 2006 21:27:24 +0100 (BST)

Thomas B�ttner already pointed out that estimating such a model is quite simple if you don't want
to control for unobserved heterogeneity. Say you have 3 states. Than you have to make three
variables: Choice after state1, choice after state 2, and choice after state 3. Each variable is
separately modeled as a -mlogit-. One thing to wory about is that one person can start in one
state  multiple times, which would mean he enters the dataset multiple times. So you might want to
use the cluster option to cluster by person id. 

In short, if you don't want to do anything fancy with unobserved heterogeneity than the most
dificult job about estimating this model is to prepare your data and you don't have to write any
new program.


--- Carol Bao <[email protected]> wrote:
> There have been quite a few postings regarding
> estimating a multiple state (more than 2) markov
> transition model, including the two postings in the
> following.  I am wondering if anyone has suggestions
> about this.  Also, there is this paper titled "Using
> Stata's -ml method d2- to estimate a multi-state
> Markov transition model" by Thomas B�ttner
> ([email protected]).  I am wondering if anyone has
> used this method and what is the converging speed and
> results, since -ml- with -d2- is no easy job.  It
> seems that I cannot get hold of Mr. B�ttner via the
> no-longer valid email address.

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z214

+31 20 5986715

Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail
*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index