Monday, February 20, 2006
st: glm - predicted values for different x values
How can I compute predicted values for different x values after estimating a model with glm, family(bin) and link(logit)?
A little background: My endogenous variable is a percentage variable that is by definition limited between zero and one hundred, and about half of the observations have a value zero. Therefore, I estimate the empirical model using the fractional logit estimator introduced by Papke and Wooldridge in 1996 by using the command
glm y x1 x2 x3 x4 x5, fam(bin) link(logit) cluster(z).
To understand how important the statistically significant effects of x1 - x5 are from an economic point of view, I want to compute the predicted values for different sets of values for these x variables.
I know how to do this for other non-linear estimators like probit (there is a FAQ on this, by the way), but I have no idea how to proceed with glm. Any help is highly appreciated!
* * 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/
Links to this post: