Dear Statalisters,
I would like to manually compute the same delta standard errors as margins after linear regression. This FAQ by Jeff Pitblado explains how to manually calculate the SE's after logit. I could replicate the predicted value but having a problem getting the SE's and CI's.
I would be grateful if you could give me some advice on how to do that.
Thanks!


Code:
sysuse nlsw88, clear
gen lnwage=ln(wage)
reg lnwage i.collgrad c.grade i.union c.hours
margins, over(collgrad)
predict yhat, xb
mean yhat, over(collgrad)