i m working on a log log model.
first run a regression of log wage on education, log experience, female, married, large-firm.
use the corrected prediction to calculate the marginal effect of an additional year of experience for an unmarried male at a small firm at the average education and experience.

do i do this right?
gen lnhrearn=ln(hrearn)
gen lnexper=ln(exper)
scalar avexper=r(mean)
scalar lnavexper=ln(avexper)
scalar aveducrecode=r(mean)
scalar sig2b=e(rss)/e(df_r)
di "corrected factor="exp(sig2b/2)
scalar predlnhrearn1=_b[_cons]+_b[educrecode]*aveducrecode+_b[married]*0+_b[female]*0+_b[largfirm]*0+_b[lnexper]*lnavexper
scalar sig2b=e(rss)/e(df_r)
scalar yhatd=exp(predlnhrearn1)*exp(sig2b/2)
disp _b[lnexper]*(yhatd/avexper)

i ask because i got different stata outcome with hand calculation.