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

Re: st: Fixed effects using binary numbers + F test for Fixed effects

From   "Austin Nichols" <>
Subject   Re: st: Fixed effects using binary numbers + F test for Fixed effects
Date   Fri, 2 Nov 2007 13:59:26 -0400

What does "fixed effects using binary" mean?  What are d04 and d05? By
MCO do you mean "Monte Carlo Optimization" to estimate numerical
integration via sampling?

The test that all FE are 0, or var(u_i)=0, is reported by -xtreg, fe-
as a matter of course, I believe, and available as e(F_f) after

Can you restate your problem in terms of available data, e.g. start with
   webuse grunfeld, clear
or the like?

On 11/1/07, joelle farajallah <> wrote:
> Hi everyone
> I have quite a few questions about the fixed effects that might sound very
> beginner level things but i'm getting really stuck with them
> We have a certain model
> lcrmrteit = 0 + 1 lprbarrit + 2 lprbconvit +d04 + d05 + uit.
> The first question is to develop a model with fixed effects using binary
> that can be estimated by MCO and estimate this model using Stata
> The second question is to use an F test to validate the usage of fixed
> effects (restricted F test)
> Third question is to estimate the model using the fixed effects and use an
> F statistic to validate the presence or absence of fixed effects ( using
> integrated commands ?? )
> And they finally ask us what the difference is between the 3 F statistics
> obtained in these 3 questions , which for me kind of all seem to be the
> same.
> Thanks in advance for your help.
> *
> *   For searches and help try:
> *
> *
> *

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index