--- "John L. Worrall" <[email protected]> wrote:
> So if I don't provide the varlist before the "by" option it is not
> the same as holding the others constant, as in a typical regression
> scenario?
Correct. Lets simplify the problem by assuming you have one other
variable, bmi. typing
adjust , by(custcode) pr
gives you for each category of custcode the predicted probability for
an individual with an average bmi within that category, while typing
adjust bmi, by(custcode) pr
will give you for each category of custcode the predicted probabilities
for an individual with a bmi equal to the overall mean. In the first
case part of the differences in probability is due to differences in
the average bmi across custcode categories, while in the latter case
the bmi is held constant at the overall mean.
Hope this helps,
Maarten
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* 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/
