In addition to my last comment, you should be very careful when you
calculate the marginal effects of interactions in nonlinear models...
Be sure to consult this paper before calculating those effects...
http://www.unc.edu/~enorton/AiNorton.pdf
Justin Smith
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Xiaoheng Zhang
Sent: June 19, 2006 5:32 PM
To: [email protected]
Subject: st: interaction term of logritham of raw data----what form it
should be?
Hello all,
I have two variables that are
agglopharma = number of firms (ranging from 5 to 300)
eatr = effective average tax rate (ranging from 8.5 to 55)
They all enter the conditional logit regression in log form.I would like to
use
interaction term of these two variables(also in log form),which one should I
use?
ln(agglopharma*eatr) = ln(agglopharma)+ln(eatr)
or
ln(agglopharma)*ln(eatr) = ln(agglopharma+eatr)?
Thanks for reading anf TIA!
Kevin
*
* 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/
*
* 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/
