Statalist The Stata Listserver

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

Re: st: frustrated: MFX after regoprob

From   Richard Williams <>
Subject   Re: st: frustrated: MFX after regoprob
Date   Thu, 22 Jun 2006 12:17:53 -0400

At 10:26 AM 6/22/2006, Carol Bao wrote:
Hi, everyone,

I am estimating an random effects generalized ordered
probit model, and though I get the coefficients fine,
I cannot get marginal effects by using the mfx command
since I am getting the "Warning: ... rescaling"
First off, I don't think you are doing anything wrong. regoprob is a hybrid of gologit2 and reoprob, using code adapted from both. Early versions of gologit2 had the exact same problem with mfx not working. After spending some time on a wild goose chase of my own, I was finally able to fix that problem in gologit2, but I believe regoprob/goprobit still use the earlier problematic code. (Make sure you have the most current versions of regoprob & goprobit though.)

If you post a simple replicable example (e.g. something using auto.dta or whatever) I may be able to suggest a workaround. I'm thinking it may just be a matter of getting regoprob to use gologit2's predict command instead, but I'm not sure.

Incidentally, you may want to install -mfx2- as well. It is like a slow version of margeff! But it works with more commands, including gologit2 and goprobit. If, say, your dv has 4 outcomes, mfx2 automates the process of calling mfx 4 times, and also stores the results in a form that can be used by routines like outreg2 and estout.

Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX: (574)288-4373
HOME: (574)289-5227
EMAIL: Richard.A.Williams.5@ND.Edu
WWW (personal):
WWW (department):
* For searches and help try:

© Copyright 1996–2020 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index