Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

st: Re: xt commands for parallel growth curves


From   "Joseph Coveney" <jcoveney@bigplanet.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Re: xt commands for parallel growth curves
Date   Wed, 25 Jul 2012 23:50:55 +0900

Alan Acock wrote:

I have parallel growth curves. One is a growth curve for a variable called secd
which is a count variable and the other growth curve is for negative behavior,
also a count. I can estimate either growth curve readilly using xtmepoisson, but
I'm interested in estimating the relationship between them. That is, as secd
goes up, does negative behavior go down. I would like to estimate them
simultaneously as you can do with Mplus. I also have this data clustered by
school. So I have 8 waves of data for each child and 15 schools.

I could use Stata's sem to estimate parallel growth curves and see how they are
related, but it is restricted to continuous variables.

--------------------------------------------------------------------------------

It seems as if you should be able to fit two parallel growth curves (random 
intercept and random slope for each of two variables at the child level and the
same at the school level) using -xtmepoisson-.  

What's your Mplus set-up look like?  Is it anything like the following?

DATA:
	FILE IS DATA.DAT;
VARIABLE:
	NAMES ARE SCHOOL CHILD SECD1-SECD8
		NBEH1-NBEH8;
	USEVARIABLES ARE SCHOOL SECD1-NBEH8;
	COUNT ARE SECD1-NBEH8;
	CLUSTER IS SCHOOL;
ANALYSIS:
	TYPE IS TWOLEVEL;
	INTEGRATION = 5;
MODEL:
	%WITHIN%
	iw1 sw1 | SECD1@0 SECD2@1 SECD3@2 SECD4@3
		SECD5@4 SECD6@5 SECD7@6 SECD8@7;
	iw2 sw2 | NBEH1@0 NBEH2@1 NBEH3@2 NBEH4@3
		NBEH5@4 NBEH6@5 NBEH7@6 NBEH8@7;
	%BETWEEN%
	ib1 sb1 | SECD1@0 SECD2@1 SECD3@2 SECD4@3
		SECD5@4 SECD6@5 SECD7@6 SECD8@7;
	ib2 sb2 | NBEH1@0 NBEH2@1 NBEH3@2 NBEH4@3
		NBEH5@4 NBEH6@5 NBEH7@6 NBEH8@7;

I don't know if you're intending to fit eight random effects, but if so, then
that's bound to be too many for -gllamm-.

Joseph Coveney


*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index