st: RE: Number of observations: xtreg and xtivreg2

["Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>]

Nadia,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu 
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of 
> Abou Nabout, Nadia
> Sent: 29 September 2010 10:05
> To: statalist@hsphsun2.harvard.edu
> Subject: st: Number of observations: xtreg and xtivreg2
> 
> Dear Statalist readers,
> 
> I am running a fixed effects model on a dataset with about 
> 97,000 observations.
> 
> While xtreg seems to use all these observations (xtreg ln_cpc 
> position t qs qs_sq, i(keyword_id) fe), xtivreg2 (2SLS model 
> with fixed effects: xtivreg2 ln_cpc t qs qs_sq 
> (position=aktgebot), fe) only uses about 96,000 of these 
> observations. I checked all the variables and instruments for 
> missing values but there are none. I also do not have any 
> lagged dependent variables in there, so I am wondering why 
> the number of observations is lower for xtivreg2 compared to xtreg.

Are any of these observations singletons?  These are panel units with
only one observation.  The FE within transformation wipes them out and
they are not used in the estimation.  xtivreg2 notes this and adjusts
the number of obs accordingly.  If xtivreg2,fe and xtreg,fe give you
otherwise identical results, this is probably the explanation.

--Mark

> Any help would be highly appreciated!
> 
> Best,
> Nadia
> 
> 
> 
> *
> *   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/
> 


-- 
Heriot-Watt University is a Scottish charity
registered under charity number SC000278.


*
*   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/