Statalist


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

AW: st: AW: LSDV vs. FE


From   "Martin Weiss" <martin.weiss1@gmx.de>
To   <statalist@hsphsun2.harvard.edu>
Subject   AW: st: AW: LSDV vs. FE
Date   Sat, 28 Feb 2009 17:18:34 +0100

<> 

Look at the example adapted from Baum (2006), chapter 9.1
http://www.stata-press.com/books/imeus.html


*************
u http://www.stata-press.com/data/imeus/traffic, clear

qui{
	xtreg fatal beertax spircons unrate perincK, fe
	estimates store fe
	xi: reg fatal i.state beertax spircons unrate perincK
	estimates store dummy
}

* see the same test as in xtreg, fe
testparm _Istate_*

* table of results
 estimates table fe dummy, style(oneline)
*************

The parameter estimates are the same, and the -testparm- line replicates the
F-test after -xtreg,fe-


HTH
Martin

-----Ursprüngliche Nachricht-----
Von: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Asgar
Khademvatani
Gesendet: Samstag, 28. Februar 2009 09:51
An: statalist@hsphsun2.harvard.edu
Betreff: Re: st: AW: LSDV vs. FE

Hello Martin,

Thanks for the comments. My understanding from Baum (2006) is that in 
running LSDV, which is equivalent to one-way FE, we should run with 
constant term as,( for instance if we have 4 Units):
regress y d1 d2 d3 x         where d2 d3 d3 are dummies.

Am I right?

Thank you.
Asgar

Martin Weiss wrote:
> <> 
>
> You want to consult Baum (2006), chapter 9.1.1, where this is explained
> quite clearly. See 
> http://www.stata-press.com/books/imeus.html
>
>
> HTH
> Martin
>
> -----Ursprüngliche Nachricht-----
> Von: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Asgar
> Khademvatani
> Gesendet: Freitag, 27. Februar 2009 19:48
> An: statalist@hsphsun2.harvard.edu
> Betreff: st: LSDV vs. FE
>
> Hello All,
>
> In general, I understand that LSDV estimation method  is an alternative
> way of modeling fixed-effectsin in panel data approach , or it is
> equivalent to FE method. But, in STATA, we can specify a LSDV model in two
> ways:
>
> (1)with common constant as:
> LS y d1 d2 d3 x1 x2 x3   or
>
> (2)we can run  it without constant or with non-constant option as:
> (for example)LS d1 d2 d3 x1 x2 x3, nonconstant
>
> I got a couple of concerns and questions as follows:
>
> First,  which syntax of LSDV given above is appropriate or reliable for
> modeling a fixed-effects model? I mean it should be run with constant or
> without constant as shown above?
>
> Second, when we run FE model in Stata as:
>
> xtreg y x1 x2 x3, fe
>
> My question is that is this equivalent to LSDV with constant or without
> constant?
>
> Thank you in advance.
>
> Regards,
> Asgar Khademvatani
> Ph.D. Candidate-Department of Economics
> The Faculty of Social Sciences
> University of Calgary,
> 2500 University Dr., NW., Calgary,
> Alberta, Canada, T2N 1N4
> Email: akhademv@ucalgary.ca
> Tel: +1(403)210-2574
>
> *
> *   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/
>
>
> *
> *   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/
>
>
>   


-- 
Sincerely,
Asgar Khademvatani, PhD Candidate
Department of Economics
University of Calgary




*
*   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   |   What's new   |   Site index