Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: a question about -gllamm- model

From   David Hoaglin <[email protected]>
To   [email protected]
Subject   Re: st: a question about -gllamm- model
Date   Sun, 25 Mar 2012 21:49:50 -0400


Have you looked at the documentation for -gllapred-?  I have the
impression that its options include what you are looking for.

To what extent do your data support changing that dichotomous
predictor from 0 to 1 while holding the other predictors constant?
You should check that you are not extrapolating beyond your data.  The
"held constant" interpretation generally oversimplifies what is going

David Hoaglin

2012/3/25 Huaandrew <[email protected]>:
> Hi Statalisters:
> Is there a way that I can estimate the changes in probability(the dependent variable) that result from changing a dichotomous independent variable from 0 to 1 (while holding all other independent variables constant) in a gllamm model? As far as I know, the gllamm model is not supported by Clarify. I believe that there is another way through Stata to complete this sort of simulation (under the estimated parameters of the model), but I do not know exactly how or what the exact command would be. Any insight you might have would be greatly appreciated!
> Andrew

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index