Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: definition of pseudo R^2 for dprobit or probit

From   "Renzo Comolli" <[email protected]>
To   <[email protected]>
Subject   Re: st: definition of pseudo R^2 for dprobit or probit
Date   Mon, 27 Oct 2003 18:43:07 -0500

Hi Santiago,

Fancy seeing you on Statalist. I write here what is of interest to all,
write you privately for more

I find curious that the answer to your question is not in the manual, but
from my (5 seconds) read of the manuals it is not there. (I'd love to be
contraddicted by others who read the manuals more carfully).

I think I have the answer nonetheless, but I invite you to try because I
used a "nonsafe" method
I installed fistat by typing 
. ssc install fitstat

If you are using a computer where the ado folder is not writable (i.e. at
the Department), you have to do
. net set ado "c:\temp"
. adopath + "c:\temp"
. ssc install fitstat

After that, I did
. probit employed age education
. fitstat

And I eyeballed to which definition of pseudo R^2 generated by fitstat, the
pseudo R^2 generated by probit corresponds.
It looks like it corresponds to McFadden's R2

With this knowledge, I went to read Regression Models for Categorical
Dependent Variables Using Stata
For McFadden's R^2 it gives 1-[log likelihood of the model you are
estimating/ log likelihood of an alternative model which contains only the


>From   Santiago Bazdresch <[email protected]> 
>Subject   st: definition of pseudo R^2 for dprobit or probit 

>I cant seem to find the exact definition of the statistic "pseudo R^2" that
stata 7 reports with probit or dprobit 
>anybody know it?
>how should I find it?

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index