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

st: oprobit marginal effects with interactions

From   "Zohal Hessami" <>
Subject   st: oprobit marginal effects with interactions
Date   Sun, 08 Feb 2009 15:27:55 +0100

Hello guys,
currently, I am dealing with ordered probit estimations where I want to investigate the effect of government size on people's life satisfaction. To do so I have interacted government size with a few institutional dummies such as the extent of corruption.
I read the article by Ai and Norton (2003) which clearly points out that the calculation of marginal effects in the framework of nonlinear estimation techniques such as ordered probit is quite difficult and often incorrectly done. I also found out about the inteff module in STATA, which appears to calculate just what I need.
My do file reads as follows:
oprobit lifesat govtexp corruption corrgovtexp x, r 
inteff lifesat govtexp corruption corrgovtexp x, savedata(C:\Model3govtexp, replace) savegraph1(C:\graphs\fig1, replace) savegraph2(C:\fig2, replace)

However, when I tried to do this I got the following error message:
last estimates not found

Is it possible that inteff can only be used for binary probit/logit and not for ordered probit/logit? If not, what else could be the reason why I get this error message?
Thx in advance,
Zohal Hessami
University of Konstanz
Department of Economics
Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kann`s mit allen:
*   For searches and help try:

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