Dear all,
I have a 3-level categorical variable and depending on whether it is
specified as -i.levels- or -i.level2 i.level3- margins gives slightly
different values. My manual calculation and mfx agree with the second
specification. Could anyone point out what exactly is the difference
in the marginal effect calculation?
Thank you
Regards
Manos
** EXAMPLE
sysuse nlsw88.dta, clear
drop if occupation>3 | occupation==. | union==.
tab occupation,g(oc)
*
probit union i.occupation
margins, dydx(*) atmeans
*
probit union i.oc2 i.oc3
margins, dydx(*) atmeans
*
probit union oc2 oc3
mfx
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/
