Re: st: testing differences between coefficients

[Maarten Buis <maartenlbuis@gmail.com>]

On Thu, Dec 20, 2012 at 10:38 AM, Ebru Ozturk wrote:
> Maarten, what you said is really complicated for me. Simply, I want to test this hypothesis: The positive effect of direct ties is lower for the X technology relative to Y technology. And I use tobit/probit regression.
>
> And I would like to test it by looking at the difference for the coefficients of direct ties variable for these two models (the first model regressing on X technology and the second model regressing on Y technology).

This is in essence a repetition of the first question. Unfortunately
but not very surprisingly, my answer remains the same. You seem to
think that this is a mere technical question, and there is a technical
aspect to that question. The answer to that I have given in my first
answer: -suest-. You followed up with some more detailed question on
how to implement it. The problem there is that that depend on what you
mean with "effect", as what an "effect" is can be defined in various
ways in your models. Because of that there are multiple answers
possible to your answer. When something like this happens than it is
almost always a sign that your research question and your hypotheses
are not precise enough. In this case you need to be more precise what
you mean with "effect". I have given you the steps you need to go
through in order to define your effect.

I realize this is not the answer you are looking for, but sometimes
there is no easy answer (or easy answers would be misleading).

-- Maarten

---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

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