Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

st: Constraining covariance (HLM)

From   "Saul G. Alamilla" <[email protected]>
To   Statalist <[email protected]>
Subject   st: Constraining covariance (HLM)
Date   Thu, 13 Jan 2011 12:34:19 -0800 (PST)

Dear Statalist Members:

I have a question about constraining covariance(s) in HLM:

Say I have a random intercepts and slopes model w/nested data. I believe STATA's default is to make the random intercept and slope uncorrelated. You have to specify cov(un) to do otherwise. HLM's default appears to be: cov(un), and there doesn't appear to be a way to UNDO this, so that the two effects are assumed uncorrelated. 

Is there a way to have the two uncorrelated in HLM?


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index