Statalist The Stata Listserver


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

Re: st: Test for Pooled OLS against Fixed effect model


From   "Keynes M. Smith" <[email protected]>
To   [email protected]
Subject   Re: st: Test for Pooled OLS against Fixed effect model
Date   Thu, 19 Apr 2007 13:17:27 +0100

Hi,

Thanks for your reply.

I've read the econometric book and it state that the null hypothesis
for testing the group

effect is all dummy parameters except one are zero. While the F-test
that state on Stata

result is that all u_i = 0 , so I'm not sure that I can use this
F-test to check for group effect

Best,

A. Sura

On 4/19/07, Johan Hellstrom <[email protected]> wrote:
Before you proceed, I would suggest reading an introductory econometric
textbook on panel data analysis - these issues should be covered there. And
how about using the Stata help command or the manuals? Nonetheless, my quick
(but not complete) answers for some Stata commands are written below.

>Hi,
>
>I've got a question about pooled OLS versus fixed effects. I run a
>fixed effects regression and got an F test that all u_i=0:     F(9,
>137) =     6.92      Prob > F = 0.0000 , does this mean that I should
>run a fixed effect or random effect regression instead (or do I have
>some more serious problem then)?

If the null hypothesis is rejected, you may conclude that the fixed effect
model is better than the pooled OLS model.

For your second question see the help file (and/or manual )for:
hausman

>
>Beside, I've 14 parameters with 10 observations, is this enough to run
>random effect model?
>
>Moreover, how can I then test for autocorrelation and heteroskedastic
>in the pooled OLS model ,fixed effect model and random effect model? I
>try to search for the command but not found.

There is no single Stata command, rather quite a few mainly dependent on if
you have a balanced or unbalanced panel. Except for the classical graphical
regression analysis approaches, see for instance the help files (and/or the
manual) for:
regress postestimation
regress postestimation ts

or
http://www.stata.com/support/faqs/stat/panel.html

>
>Thanks
>
>A. Sura.

Best,
Johan Hellstrom

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index