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

Re: st: Differences in regression slopes

From   "Austin Nichols" <>
Subject   Re: st: Differences in regression slopes
Date   Wed, 20 Feb 2008 13:34:56 -0500

E. Paul Wileyto <>:
Yours is the appropriate response to a question about comparing slopes
in a linear regression.  In a logit model, the issue is not so cut and
dried.  Allison (1999), drawing on Amemiya and others, famously
criticized the use of interaction terms to compare slopes across
groups; similar criticisms apply to most other methods.  Williams
(2006 and much subsequent work) usefully summarizes and critiques
possible solutions.  A quote from "Comparing Logit & Probit
Coefficients Across Groups"
( is particularly
helpful: "[Allison's proposed solution] is part of a larger class of
models that is variously known as location-scale models (McCullaph &
Nelder 1989) and heterogeneous choice models (Alvarez & Brehm, 1995;
Keele & Park, 2006)."  Some of this work is peer-reviewed and some
not, but in every case, caveat lector.

Allison, Paul. 1999. "Comparing Logit and Probit Coefficients Across
Groups." Sociological Methods and Research 28(2): 186-208.

Alvarez, R. Michael and John Brehm. 1005. "American Ambivalence
Towards Abortion Policy: Development of a Heteroskedastic Probit Model
of Competing Values." American Journal of Political Science, 39(4):

Keele, Luke and David K. Park. 2006. "Difficult Choices: An Evaluation
of Heterogenous Choice Models"

McCullaph & Nelder. Generalized Linear Models, 2nd edition, 1989

Williams, Richard. 2006. "Generalized Ordered Logit/ Partial
Proportional Odds Models for Ordinal Dependent Variables." The Stata
Journal 6(1):58-82. A pre-publication version is available at

On Feb 20, 2008 12:14 PM, E. Paul Wileyto <> wrote:
> Responses so far have sent you this way and that.  Just look up -test-
> in STATA help.
> To get to the point of using -test- for your purpose, you would need to
> specify a model that has group-specific slopes, or combine two
> regressions, one for each group, using -suest-.
> Paul
> Barth Riley wrote:
> > Hello
> >
> > Is there a procedure in stata to determine if a regression slope
> > (coefficient) for a predictor variable is significantly different between
> > groups? I would like to apply this procedure for coefficients in logistic
> > regression nalysis.
> >
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index