Stata: Data Analysis and Statistical Software

Austin -
I recall some very old posting (prior to your empirical testing) where we spoke of 100 clusters needed  for the asymptotics to kick in.
Particularly when clustering at the country level requiring a large number of clusters could become a problem, so I am really interested in your future findings for non-linear models...
Best
Justina


[email protected] schrieb: -----

An: [email protected]
Von: Austin Nichols <[email protected]>
Gesendet von: [email protected]
Datum: 06.01.2011 02:43PM
Thema: Re: Antwort: RE: st: Direction of the effect of the cluster command on the standard error depends on the inclusion of a control variable

Eric-
Those should be the same file, but I guess one got corrupted--thanks
for the heads-up.
Justina-
Of course the CRSE has different properties in a nonlinear model, but
I suspect the same intuition applies with slightly larger minimum
numbers of clusters.  I have not run the relevant simulations, though.

On Thu, Jan 6, 2011 at 7:51 AM, DE SOUZA Eric
<[email protected]> wrote:
> The one at this address opens without problem:
> repec.org/usug2007/crse.pdf
>
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of Neil Shephard
> Sent: 06 January 2011 13:21
> To: [email protected]
> Subject: Re: Antwort: RE: st: Direction of the effect of the cluster command on the standard error depends on the inclusion of a control variable
>
> On Thu, Jan 6, 2011 at 12:12 PM, DE SOUZA Eric <[email protected]> wrote:
>> I did download and save, then open with Acrobat Pro I did this in the
>> past, too In between I have switched from WinXP to Win7 and also have
>> a new version of Acrobat.
>
> I too have problems viewing the PDF after downloading using Adobe Acrobat 9.4.1 under GNU/Linux.
>
> Neil

> *
> *   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/