st: re: indirect effects in logit model

[Kit Baum <baum@bc.edu>]

Charlie said

My model is compounded. That is,
Y=X1+X2+X3 (Y= binary dependent variable with very rare case of "1")
but, one independent variable of the equation (X2, continuous
variables) is a function of X1; X2=X1+ X4

As you see in the above, the major independent variable is X1.
X1 has a direct effect on Y and also an indirect effect via X2 on Y.

I want to estimate X1's direct and indirect effects via on Y.


If X2 is a deterministic function of X1, then just substitute its  
equation into the main equation, and estimate
logit y x1 x4 x3

If the original logit is I = b0 + b1 x1 + x2 x2 + b3 x3 + error, then  
you are now estimating
I = b0 + (b1+b2) x1 + b2 x4 + b3 x3 + error

and you can calculate the direct effect of x1 as its estimated  
coefficient minus b2-hat, using lincom.



Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html


*
*   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/