Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Predicted probabilities for different x values after xtlogit, fe

From   joachim Wagner <>
Subject   st: Predicted probabilities for different x values after xtlogit, fe
Date   Mon, 18 Aug 2003 08:54:35 +0200

Dear listers,

I am currently reading a paper reporting results from a fixed-effects logit model (estimated with Stata). Some of the estimated regression coefficients are highly statistically significant, and the author concludes from the prob-values reported that these variables are highly important. Unfortunately, this interpretation is based on the statistical significance alone. What I would like to do is the following: I would like to compute the predicted probability for a hypothetical unit (here:plant) with average values for all the exogenous variables, and then I would like to change - one by one - the values of the statistically significant variables and to compute the predicted probabilities for these hypothetical units to see how much these predicted probabilities change. In this way I hope to learn whether the statistically significant variables are really important in the sense that a reasonable change in their value have a large impact on the predicted probability. I usually do this kind of "simulation" when I work with simple probit models.

Question: Using Stata 8.1, how can I do what I want to do based on the output from xtlogit, fe (and a table of descriptive statistics for the variables included in the model) alone - without having access to the data used to estimate the model? Is this possible at all?

If it should not be possible, how would I do all this when I would have access to the data? (Maybe, I can contact the author and convince him to perform this exercise if I could send him a do-file ....)

Many thanks in advance!

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index