Thanks Rich. This is, in fact, the formula that -esttab- uses. What I
meant is that there should be a way to compute the betas without going
back to the data to get the standard deviations of the variables, but
to somehow derive the betas from e(b) and e(V). This must be possible
since, e.g., -regress- is able to display the betas even after all
data were dropped:
sysuse auto
reg price weight mpg
drop _all
regress , beta
But maybe regress has some secret place where it stores the betas. I don't know.
ben
On 8/23/07, Richard Williams <[email protected]> wrote:
> At 07:06 AM 8/23/2007, Ben Jann wrote:
> >Chris
> >I cannot reproduce this. The beta option works fine on my system.
> >Somehow, the computation of the beta's fails in your case. The only
> >explanation that springs to my mind is that the variables for which
> >the betas need to be computed were deleted from memory after model
> >estimation. The original variables are required to be able to compute
> >the betas (at least as it is implemented in esttab; I believe there is
> >a way to compute the betas from the coefficients vector and the
> >covariance-matrix directly - if someone has the formula, I'd be happy
> >to know).
> >ben
>
> Ben - see the last formula on p. 2 of
>
> http://www.nd.edu/~rwilliam/stats2/l02.pdf
>
>
> -------------------------------------------
> Richard Williams, Notre Dame Dept of Sociology
> OFFICE: (574)631-6668, (574)631-6463
> HOME: (574)289-5227
> EMAIL: [email protected]
> WWW: http://www.nd.edu/~rwilliam
>
> *
> * 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/
>
*
* 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/
