The reason for this is that -dprobit- is a wrapper on -probit-. They
have the same underlying estimation, but just report results
differently, and in the guts of Stata, the probit estimation results
are what really is stored.
the marginal effects are also stored in e(dfdx), and the corresponding
standard errors in e(se_dfdx). my request is to use these instead by
checking e(cmd)
As they say, if this is documented, this is a feature, not a bug :)).
i am glad you agree that it's a documented bug
edwin
On Tue, 21 Sep 2004 17:14:16 +0200 (CEST), Edwin Leuven <[email protected]> wrote:
i find -estimates- a convenient command to combine output from
estimation commands using -estimates table-
one recurring annoyance however is that it doesn't play nice with
-dprobit- :
instead of reporting the marginal effects it reports the probit
coefficients!
i guess this email is directed to statacorp as a request to have
-estimates- report probit coeffients after having run -probit-, and
marginal effects after -dprobit-
if people have opinions on this, please do not hesitate to flame me
edwin
*
* 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/
--
Edwin Leuven
Department of Economics
University of Amsterdam
Roetersstraat 11
1018 WB Amsterdam
the Netherlands
tel: +31 20 525 5241
fax: +31 20 525 4310
http://www.fee.uva.nl/scholar/mdw/leuven
*
* 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/
