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

st: Interpreting interactions in probit and logit models

From   [email protected]
To   [email protected]
Subject   st: Interpreting interactions in probit and logit models
Date   Sun, 10 Jan 2010 22:00:41 +0100 (CET)

Dear Stata users,
if I have a significant interaction term in probit or logit model, how should I interpet it? 

I have found this sentence: "as long as you interpret the interaction effects in terms of the effect on the latent variable you are ok in simply using the output from -probit- (i.e. the sig. of the coefficient); if you want to interpret the results in terms of the probabilit you should use -inteff-."

What is the difference between these two interpretations? In concrete terms: what can I say if I have a sig. interaction coefficient BUT THE INTERACTION EFFECT IS NOT SIGNIFICANT AFTER HAVING USED INTEFF OR RELATED POSTESTIMATIONS to check the effect on probability?? What are the specific conclusion I get from each of the two interpretations (on the latent variable and on the probability)?

(A final note (if it may help in the answer): I have this problem for two different regressions. In the firs regr, I have an interaction between "one dummy - one continuous" variables; in a second regression I have an interaction between two dummies).
*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index