Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: HLM random effects estimates


From   Derek Darves <darves@darkwing.uoregon.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: HLM random effects estimates
Date   Thu, 5 Dec 2002 12:11:54 -0800

Hello,

I have a correction to my post yesterday, and would like to repost the question now that I have learned little more about my problem: using gllamm, if I specify the "fam(bin)" option, Stata does not print the level 1 random effects covariance estimate, as shown in example 1 below. But, if I do *not* specify family(bin), it appears to report a level 1random effects convariance estimate, as shown in the second example below. Is this a bug or is there something about specifying the binomial family option that precludes the ability to get the level 1 random effects covariance estimate?

Example 1

. gllamm infor, i(site9) fam(bin) link(logit)

number of level 1 units = 8897
number of level 2 units = 152

Condition Number = 1.8886851

gllamm model

log likelihood = -4529.9736

------------------------------------------------------------------------ ------
informal | Coef. Std. Err. z P>|z| [95% Conf. Interval]
------------- +----------------------------------------------------------------
_cons | -.8308119 .0492316 -16.88 0.000 -.9273041 -.7343197
------------------------------------------------------------------------ ------


Variances and covariances of random effects
------------------------------------------------------------------------ ------


***level 2 (site9)

var(1): .598095 (.06353876)
------------------------------------------------------------------------ ------


Example 2
(Now, without the "fam(bin)" option, note the "Variance at Level 1" section below)

. gllamm infor, i(site9) link(logit)
number of level 1 units = 8897
number of level 2 units = 152

Condition Number = 6.969789

gllamm model

log likelihood = -4593.7317

------------------------------------------------------------------------ ------
informal | Coef. Std. Err. z P>|z| [95% Conf. Interval]
------------- +----------------------------------------------------------------
_cons | -.8180495 .0476116 -17.18 0.000 -.9113665 -.7247326
------------------------------------------------------------------------ ------

Variance at level 1
------------------------------------------------------------------------ ------

.15948543 (.00240677)

Variances and covariances of random effects
------------------------------------------------------------------------ ------


***level 2 (site9)

var(1): .63212927 (.06542581)
------------------------------------------------------------------------ ------

Thanks,

Derek Darves

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* 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   |   What's new   |   Site index