Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: panel data xtmixed vs xtreg
From
Joerg Luedicke <[email protected]>
To
[email protected]
Subject
Re: st: panel data xtmixed vs xtreg
Date
Fri, 3 Jun 2011 10:06:39 -0400
A statistical model itself does not care about causality. You may or
may not be able to set up a model in a certain way to give your
estimates a causal interpretation. This depends on many things like
theory, measurement, and the model set-up. Given the jargon you use, I
assume that your work is related to the field of economics. You should
check back with the econ literature about common solutions in that
field. However, I do not think there is a one-size-fits-all-solution
for the problem of reverse causality, or more general, complex causal
structures. And certainly not on the level of a statistical model
(like if you would say it does not matter how my data is gathered etc.
I can just run model xyz and get rid of _"the"_ endogeneity problem).
But econ people could have a different opinion here, so you should see
what the habits are in that field.
J.
On Thu, Jun 2, 2011 at 12:59 PM, yuying shi <[email protected]> wrote:
> Thanks very much! Another question is : is there any way to control for the endogeneity problem when using xtmixed? I checked the xtmixed manual, but could not find it. I can certainly use a two stage least square by regressing the endogenous variable on other exogenous variables first and then substitute the predicted value back, but would xtmixed implicitly assume all variables are exogenous?
>
> Best,
> Sunny
>
> --- On Thu, 6/2/11, Joerg Luedicke <[email protected]> wrote:
>
>> From: Joerg Luedicke <[email protected]>
>> Subject: Re: st: panel data xtmixed vs xtreg
>> To: [email protected]
>> Date: Thursday, June 2, 2011, 7:55 AM
>> On Thu, Jun 2, 2011 at 12:34 AM,
>> yuying shi <[email protected]>
>> wrote:
>> > Dear Statalist,
>> > Currently I am using the xtmixed to analyze
>> an unbalance panel data.=
>> > Could any tell me what the difference between xtreg
>> and xtmixed? I got dif=
>> > ferent result by the two commands.=20
>> > Thanks!
>>
>> If you run a 2-level varying (aka random) intercept model
>> (i.e.
>> without varying slopes) using -xtmixed-, you should get the
>> same
>> results as with -xtreg-'s default re option. You need to
>> post your
>> code and output (or relevant parts of it) in order to see
>> what went
>> wrong in your case.
>>
>> J.
>>
*
* 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/