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: Testing for rho differences after mvprobit

From   <>
To   <>
Subject   st: Testing for rho differences after mvprobit
Date   Fri, 8 Apr 2011 11:59:49 +0100


Date: Fri, 8 Apr 2011 00:47:20 +0200
From: =?iso-8859-1?Q?Jaime_G=F3mez?= <>
Subject: st: Testing for rho differences after mvprobit

Dear Statalisters
I have estimated a model through the "mvprobit" command of Capellari and Jenkins. I would like to test whether the differences in the rhos are statistically significant. For example, I would like to know whether rho21 is different from rho31 or whether rho21>rho42. Is there any way of doing this with Stata?
Thank you very much.
Jaime Gómez.

-mvprobit- is available on SSC (this is a later version than the 2003 Stata Journal version). [Remember the Statalist FAQ reminding you to say where modules are available from.]

You can straightforwardly test equality of correlations using -test- but referring to the ("atrho") transformed versions of the correlations rather than the correlations themselves. Look in e(V) and e(b) using -matrix list- for the names used by the program. Inequality constraints cannot currently be tested directly, as far as I know.

If you have 4 outcome variables, then you might find the Cappellari/Jenkins material in Stata Journal 6(2) 2006 (downloadable from SJ website) of interest -- using code of the kind illustrated there together with the plugin will estimate your model faster. That's relevant if you're having to fit it multiple times.  See also -cmp- on SSC by David Roodman.

Professor Stephen P. Jenkins  <>
Department of Social Policy and STICERD
London School of Economics and Political Science
Houghton Street, London WC2A 2AE, U.K.
Tel. +44 (0)20 7955 6527
Survival Analysis using Stata:
Downloadable papers and software:

Please access the attached hyperlink for an important electronic communications disclaimer:

*   For searches and help try:

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