Statalist


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

st: RE: RE: XTLSDVC and R-squared


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: RE: XTLSDVC and R-squared
Date   Mon, 14 Apr 2008 18:01:51 +0100

A small refinement here: use -summarize, meanonly- in each instance
here. 
In this example the saving will be trivial, but it's a good habit to get
into. 
There is more at

SJ-7-3  st0135  . . . . . . . . . . . Stata tip 50: Efficient use of
summarize
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N.
J. Cox
        Q3/07   SJ 7(3):438--439                                 (no
commands)
        tip on using the meanonly option of summarize to more
        quickly determine the mean and other statistics of a
        large dataset

Rodrigo Alfaro A.

I think that you could report within R2 from the LDV+FE model. 
It seems to me that you panel is large in T, is it unbalanced? 

Bruno (2005) -the author of -xtlsdvc- reports simulations on 
the effect of unbalancedness. You need the arithmetic (A), and 
harmonic (H) averages. You could check his results on the paper
and use the following code to compute the statistics using
your data. 

qui {
	* Arellano-Bond dataset
	webuse abdata
	sum year
	gen time = year - r(min) + 1
	* Getting the max time for firm
	bysort id: egen tmax=max(time)
	* Leaving only the result in last period
	replace tmax=. if time!=tmax
	gen double inv_tmax = 1/tmax
	sum tmax
	local A = r(mean)
	sum inv_tmax
	local H = 1/r(mean)
	local R = `A'/`H'
}
di in ye "Arithmetic =" `A' ", Harmonic=" `H' ", H/A=" `R'

Bruno, G. (2005) "Approximating the bias of LSDV estimator for 
dynamic unbalanced panel data models" Economic Letters, 87: 361-6


-----Original Message-----
Rodrigo,

The coefficients in the LSD+FE are very close to the ones from
-xtlsdvc-, a little different but very close. I have around 11,000
observations, so it takes close to 2 hours to run one model in xtlsdvc,
but I can get a good quick sense of what the results will be with just
and the LSD+FE.

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   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