[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: bivariate probit

From   Maarten buis <>
Subject   Re: st: bivariate probit
Date   Sat, 14 Jun 2008 19:55:07 +0100 (BST)

--- Andrea Rispoli <> wrote:
> I am using a bivariate probit (biprobit) estimation of two equations.
> I am interested in looking at the correlation between the two
> regressions. I noticed that Stata output does not provide with the
> p-value for rho (as it happens in the case of trivariate probit).
> Should I look at the P>|z| provided for the athrho?
> What does the athrho refers to?

athrho is the Fisher's Z transformation of the correlation, also known
as the arc-hyperbolic tangent. To get your estimates in the correlation
metric you can apply the inverse of this transformation, the hyperbolic
tangent, which is built into Stata as the -tanh()- function:

*--------------- begin example ----------------
webuse school
biprobit private vote logptax loginc years
nlcom tanh([athrho]_b[_cons])
testnl tanh([athrho]_b[_cons])=0
*--------------- end example ------------------
(For more on how to use examples I sent to the Statalist, see )

The thing you probably don't want to do is to perform tests on the
edges of the parameter space, in this case testing whether the
correlation is 1 or -1. Things usually get a weird in these areas. 

-- Maarten

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Sent from Yahoo! Mail.
A Smarter Email
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index