Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Intepretation of interaction terms


From   Maarten Buis <maartenlbuis@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Intepretation of interaction terms
Date   Tue, 24 May 2011 09:41:15 +0200

On Tue, May 24, 2011 at 9:15 AM, lreine ycenna <lreine.ycenna@gmail.com> wrote:
> In fact,
> another similar question:
>
> y = a + b + a x b
>
> I want to test this equation for group1 (dummy G1 ==1), while a , b, y
> are numeric.
>
> In Stata, it would be:
>
> regress y a b axb G1 axG1 bxG1 axbxG1
>
> we must include G1 individually in the equation.

That is not necessarily true, it just depends on what you want.

> Also, we need to add up the coef of G1 to get the true effect for this group.

That is also not necessarily true.

> My question is, since G1 is a dummy, how should we interprete G1? i.e.
> what does the coefficient for G1 mean to Y (GDP)?

That depends on how you set up your model and how you choose to
interpret your results. Typically with interaction effects there are
multiple interpretations possible.

You seem to be looking for a "cook book recipe" for doing interaction
effects, and no such thing exists. Typically there multiple ways of
correctly estimating the same model  (and many more ways to do it
incorrectly). The only way of doing this right is to understand the
logic and derive the model yourself. Deriving once own model sounds
hard, but it is not. The only hard part is not making mistakes (it is
easy to forget a minus sign or misplace a bracket), but that is just a
matter of concentration.

Even though I say it is easy, it is probably best learned by doing
this a couple of times with someone who can give you face to face feed
back while you are doing this. I suggest you find someone in your
institution who can spare you an hour or so, and lets you go through
successively more difficult examples till you get to your actual model
of interest.

Hope this helps,
Maarten

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


http://www.maartenbuis.nl
--------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index