Thanks for the link to the nice searchable version of the SJ paper.
I'm still perplexed as to why my computations (see code in first message of this thread) are not closer to reproducing the standard errors of the marginal effects from -mfx- and -margeff- (I know the 1st is at the mean and the 2nd is the average by default). I'm close for the marginal effects, but I'm not very close for the S.E. of the marginal effect for either.
Thanks in advance,
Jason Franken
*
* 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/
