Statalist


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

st: Testing endogeneity with xtivreg2


From   "Ivan Etzo" <ivanetzo@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Testing endogeneity with xtivreg2
Date   Fri, 28 Sep 2007 16:34:21 +0200 (ora legale Europa occidentale)

hello to everyone,
 
I'm using xtivreg2 for a panel with N=380 nd T=6. I tested for endogeneity using the option -endog(x)- and I'd like to know whether I followed the proper procedure.
The main problem is that I have four covariates and all of them might be endogenous.
What I did so far is to use only one covariate per time, treating it as endogenous, and testing its endogeneity, that is:
 
-xtivreg2 y (x1=z1), fe cluster(i) endog(x1)-
for each x.
 
In this way I found that only two covariates are endogenous, should I estimate the model with this two covariates contemporaneously treated as endogenous? Is that right?
 
Thank u for ur help
 
Ivan
 
 
 
 



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index