Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

Re: st: marginal effect Nested Logit regression

From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: marginal effect Nested Logit regression
Date   Tue, 11 Jan 2011 15:49:05 +0000 (GMT)

--- On Tue, 11/1/11, alfredo jimenez palmero wrote:
> i got a comment from a reviewer i do not really understand.
> He says:
> "The author does not indicate whether the coefficients in
> Table 1 are marginal effects or not".
> Table 1 includes the result from a Nested and a Conditional
> Logit Model. What does that "marginal effect" mean? Should
> the coefficient I got from the NLM and the CLM be
> interpreted in a different way than in ordinary
> regressions?

Yes, you are modeling log odds rather than expected values, 
so the correct interpretation takes quite a bit of effort
in these types of models. If you are new to odds ratios,
I would start with trying to understand the results from
a simple -logit- model, move on to a mlogit, and than 
generalize that to your -nlogit- and -clogit- models. 

Hope this helps,

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index