Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Looping with regression controls


From   David Hoaglin <dchoaglin@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Looping with regression controls
Date   Wed, 29 Feb 2012 21:29:30 -0500

Dear Ras,

The key to interpreting your model is in Steve's reply.  As in
multiple regression, the coefficient of each variable tells you about
the effect of that variable, adjusting for the contributions of the
other variables (since you're fitting a logit model, the adjustments
are on the log-odds scale).

It is a common error to interpret the coefficient of a variable as the
effect of a change (of 1 unit) in that variable when the other
variables are held constant.  That is not how ordinary multiple
regression works, and the same interpretation (adjusting for the
contributions of the other variables) applies to regression-like
models in general.  In most data, the variables are not controlled;
they are simply observed.  Thus, the model does not "control for" the
other variables.

It may be appropriate to make predictions by assigning values to the
other variables and changing only one variable, as long as you stay
within (or close to) the region covered by your data.

David Hoaglin
*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index