--- On Tue, 20/10/09, [email protected] wrote:
> due to the log transformation i have the following problem:
> exp(E(ln(Consumption))) is not equal to E(Consumption).
> I thought more precise estimates would help-but as it
> turned out now- they don't.
Appart from the two solutions I already gave you can use the
fact this way you get the conditional geometric means rather
than the conditional arithmatic means. All you have to do is
find some reason why in your case geometric means are
preferable over arithmatic means ;-) In the example below I
illustrate that -regress- with log transformed dependent
variables can be used to get the geometric means, and -glm-
can be used to ge the arithmatic means.
*------------- begin example -----------
sysuse nlsw88, clear
gen ln_w = ln(wage)
xi: regress ln_w i.race
est store ols
xi: glm wage i.race, link(log)
est store glm
bys race: ameans wage
est restore ols
adjust, by(race) exp
est restore glm
adjust, by(race) exp
*------------ end example ---------------
(For more on examples I sent to the Statalist see:
http://www.maartenbuis.nl/example_faq )
More seriously, this is interesting is a nerdish
sorta way, but practically I would use -glm-.
Hope this helps,
Maarten
--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany
http://www.maartenbuis.nl
--------------------------
*
* 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/
