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

Re: st: Gllamm question

From   Maarten buis <>
Subject   Re: st: Gllamm question
Date   Fri, 30 Nov 2007 10:00:03 +0000 (GMT)

--- Barth Riley <> wrote:
> I am running a multilevel logistic model using gllamm. In addition to
> estimating fixed and random effects, I am using the geqs() option to
> fit a regression model to predict the second-level latent variable. I
> am encountering two problems: (1) even after 200+ iterations, gllamm
> fails to converge in estimating the fixed effects, even though the
> estimated coefficients remain unchanged throughout most of the 
> iteration cycle; and (2) I received the error "unable to estimate
> maxmimum likelihood" during the estimation of the geqs() model 

You are trying to do a lot in a single model. As a general modeling
strategy I would start simple and build the model by adding one
complication at a time. That way it is much easier to diagnose any
problem. I often don't practice what I preach, and I often start with
estimating a much too complicated model. However, whenever I run into
trouble (i.e. always) I diagnose the problem by going back to a simple
model and building the complicated model by adding one complication at
a time. 

Are you trying to estimate the fixed effects by adding dummies (e.g.
the variables d1_1-d17_1)? In that case you are getting inconsisten
estimates, see:

-- Maarten

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

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Yahoo! Answers - Got a question? Someone out there knows the answer. Try it
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index