Re: st: How to model a positive continuous dependent variable with many zeros?

[Maarten Buis <maartenlbuis@gmail.com>]

On Tue, May 31, 2011 at 9:24 AM, Adriaan Hoogendoorn wrote:
> I try to run regression models for two dependent variables that
> concern the seclusions of psychiatric patients:
> y1 = the number of seclusion incidents and y2 = the seclusion duration.
> Fortunately (at least from the patients perspective), there are many zeroes.
> I successfully applied a Poisson model (xtpoisson) to model the number
> of seclusion incidents
> for patients (level 1) in different clinics (level 2)
<snip>
>
> I am running into problems when I try modeling the duration of seclusions.
<snip>

My first reaction would be to consider the time spent in seclusion as
a survival model(*). This is not quite the model you asked for as it
will model the time till end of seclusion given that someone started a
spell of seclusion.  This may not be a big problem, since you also
modeled the zeros in your first Poisson model. I can imagine a result
like institution A is sending patients more often to seclusion, but
they stay there for shorter periods of time than institution B.
(Interpretation has to be a bit careful, as institution B may be
getting the "harder" patients than institution A. That is where you
need information from local experts.)

Hope this helps,
Maarten

(*) You can read more on survival models here:
<http://www.maartenbuis.nl/wp/survival.html>. This page contains my
own introduction to survival analysis but also links to a number of
other useful sites.

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/