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, is already up and running.

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

Re: st: GLLAMM and intraclass correlation

From   Stas Kolenikov <>
Subject   Re: st: GLLAMM and intraclass correlation
Date   Thu, 13 Oct 2011 09:06:40 -0500

On Thu, Oct 13, 2011 at 8:07 AM, Gannaway Jack WWG HWWD > I have used
the GLLAMM command for a random intercept ordered logit
> model, which has then obviously produced a variance for the random
> effect.  Can anyone tell me how I get an intraclass correlation from
> this?  I know for a linear ML model with a random intercept it is the
> squared sd of the effect in question over the sum of the squared sds of
> all the error terms, but how do you do it for ordered logit?

Same way. Since the variance of logistic distribution is pi^2/3, your
ICC is [variance of the random intercept]/([variance of the random
intercept]+pi^2/3). You can obtain this by -display- or -nlcom-: the
latter will give you a confidence interval, although you'd probably
want to put rather little trust in it, as the sampling distributions
of ICC are highly skewed.

Stas Kolenikov, also found at
Small print: I use this email account for mailing lists only.

*   For searches and help try:

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