Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: differences between loneway, xtreg, and xtmixed in calculating ICC
From
Caroline Wilson <[email protected]>
To
"[email protected]" <[email protected]>
Subject
st: differences between loneway, xtreg, and xtmixed in calculating ICC
Date
Fri, 13 Sep 2013 01:06:43 +0000
Hi all,
I am using Stata 11.2. I'd like to calculate ICCs for several variables. My dataset has patients nested within doctors.
I've tried three commands to calculate the ICCs - loneway, xtreg, and xtmixed. Here are examples of the code I used to calculate the ICCs for a variable called time:
loneway time doctor (take ICC directly from output where it says Intraclass correlation)
xtreg time , i(doctor) mle (rho in the output is the ICC)
xtmixed time || doctor: , mle (then calculate ICC from output as follows: sd(_cons)^2 / (sd(_cons)^2 + sd(Residual)^2)
The 3 different approaches give me similar results for most of my variables, but in a few cases the ICCs are VERY different. Could someone please shed light on the differences between these approaches, and when each should be used in calculating the ICC? I'm trying to determine which command I should use to report my ICC results.
Many thanks!
Caroline
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/