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 at the end of May, and its replacement, statalist.org is already up and running.


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

st: Testing equality of 2 coefficients after FE regression: How does Stata compute the Pooled SE and test statistic?


From   andreas nordset <andreas.nordset@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Testing equality of 2 coefficients after FE regression: How does Stata compute the Pooled SE and test statistic?
Date   Wed, 8 Dec 2010 11:15:18 +0100

Dear all,

after estimating a FE model, I am testing the equality of two coefficients.

The difference between them is about 3000 and the SE in the regression
output are respectively 1700 and 1600. I would thus have thought that
the test statistic should be
t = 3000/sqrt(1700^2+1600^) =1.29, insufficient to reject the Null
that the two coefficients are identical.

Yet --test coeff1=coef2-- gives me F(1,1386)=6.01 and hence Prob
>F=0.0143, thus rejecting the Null at the 5% confidence level.

So I'm wondering why this is: Does Stata use smaller within-person
standard errors for the test, given that the regression is a FE
regression? If so, should I also use such smaller SEs (rather than
those displayed next to the coefficients in the standard regression
output) for plotting the coefficients?

Thanks a lot,
Andreas
*
*   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