Re: st: Leamer's Extreme Bound Analysis and Fixed Effect

["P. Lopes-Monteiro" <pl344@cam.ac.uk>]

Dear Martin

Thank you for your reply. I am interested in knowing how I can obtain the 
max and min coefficient within a 95% confidence interval.

Thanks for all your help 

Patricia

On Jun 8 2011, Maarten Buis wrote:

> On Wed, Jun 8, 2011 at 1:31 PM, pl344  wrote:
> > I know the the built in Leamer's Extreme Bounds Analysis is designed to 
> > run OLS regressions. Does anyone know if there is a way to run fixed 
> > effect panel regression using Leamer's EBA on Stata.
>
> There is no built in Leamer's Extreme Bounds Analysis, there is a user
> written -eba- command available on SSC. I assume you mean that.
>
> All EBA is fitting many models and finding the minimum and maximum
> effect size of variable. If existing programs don't do what you want
> you can easily do it yourself, by looping over all possible
> combinations of a set of control variables, for each such combination
> estimate your model with that combination of control variables and
> store the coefficient of interest. In the example below I only focused
> on the the maximum and minimum coefficient and I used -regress-, you
> can adept that in whatever way you like, e.g.  you can use -xtreg-
> instead of -reg- or you can use -post- (see: help post-) to store all
> coefficients and look at the complete distribution of coefficients
> instead of only the min and the max(For more on examples I sent to the
> Statalist see:
> http://www.maartenbuis.nl/example_faq ). To get the list all possible
> combinations of control variables I used Nick Cox's -tuples- which you
> can get by typing in Stata -ssc install tuples-.
>
> *----------------- begin example ----------------
> sysuse auto, clear
>
> reg price mpg
> tempname maxb maxse minb minse
> scalar `maxb'  = _b[mpg]
> scalar `maxse' = _se[mpg]
> scalar `minb'  = _b[mpg]
> scalar `minse' = _se[mpg]
>
> tuples headroom trunk weight length turn
> forvalues i = 1/`ntuples' {
> 	qui reg price mpg `tuple`i''
> 	if _b[mpg] > `maxb' {
> 		scalar `maxb'  = _b[mpg]
> 		scalar `maxse' = _se[mpg]
> 	}
> 	if _b[mpg] < `minb' {
> 		scalar `minb'  = _b[mpg]
> 		scalar `minse' = _se[mpg]
> 	}
> }
>
> di `maxb' " " `maxse'
> di `minb' " " `minse'
> *---------------- end example -----------------------
> (For more on examples I sent to the Statalist see:
> http://www.maartenbuis.nl/example_faq )
>
> Hope this helps,
> Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
>
> http://www.maartenbuis.nl
> --------------------------
> *
> *   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/

--
Patricia Lopes Monteiro
University of Cambridge
MPhil Economics Candidate 2011
Clare College


*
*   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/