Statalist


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

st: Equality of regression coefficients across different models (with different weights)


From   Michela Bia <[email protected]>
To   [email protected]
Subject   st: Equality of regression coefficients across different models (with different weights)
Date   Sat, 16 Jan 2010 00:13:03 +0100 (CET)

Dear statalisters,
happy new year! :)

I have to test equality of regression coefficients across different models. 
The problem is that the models I implement are regressions with different weights (same covariates) and this is not allowed by the test command.

I am using Stata/SE 11 for Windows.

Here a simple example:

clear
set obs 2000

ge xa = 10 + invnorm(uniform())
ge xb = 20 + 3*invnorm(uniform())

ge y = xa + xb + 2*invnorm(uniform())

ge w = invnorm(uniform())
replace w = 0.1 if (w < 0) | (w > 1)

ge v = 2*invnorm(uniform())
replace v = 0.2 if (v < 0) | (v > 1)

reg y xa xb
estimates store alfa

reg y xa xb [aw = w]
estimates store beta

reg y xa xb [aw = v]
estimates store gamma

estimates table alfa beta gamma, eq(1)
return li

suest beta gamma
test [beta_mean = gamma_mean]

Do you have any suggestions?

Thanks in advance!
Michela Bia
*
*   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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index