Hello. I am trying to calculate standard errors for marginal effects following a logit estimation by hand (i.e. not -mfx-), since -mfx- is taking a long time and all I need is an approximation. I'm using the delta method approximation for standard errors (Greene, 2000, p. 824), which is straightforward for continuous variables.
For dummy variables, however, my "marginal effect" is the change in predicted probability associated with changing the dummy from 0 to 1. I can use the delta method to approximate standard errors for the *individual* predicted probabilities. But what is the appropriate standard error for the *difference* between the two predicted probabilities? How does Stata handle standard errors on dummy variables for -mfx-?
A formula or reference would be greatly appreciated. Thanks.
*
* 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/
