Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

st: Parametric survival model - Plotting interaction effects

From   "Christ, Florian" <[email protected]>
To   "[email protected]" <[email protected]>
Subject   st: Parametric survival model - Plotting interaction effects
Date   Sun, 17 Oct 2010 13:35:16 +0200

Dear statalisters,

I would greatly appreciate your help with regard to my event-history problem. 

I am currently using a parametric survival model (Generalized gamma distribution; following the AIC and BIC criterion) for my research project [streg-command, Stata]. 

In order to illustrate my postulated interaction effects (following Aiken/West 1991) I would like to constrain all variables to the mean except the two variables (both continuous), which are part of the respective interaction term. Similarly, I would like to extract specific hazard estimates for certain values of the two variables, which are part of the interaction term (one standard deviation below and one above the mean) to plot the effects. Unfortunately, I am not sure on how to best do this with the event-history technique. 

Also I am not really sure on how to best incorporate time, when plotting the interaction effects in an event-history setting. Can I set time to the median/mean for simplicity reasons (not having analysis time as additional axis) or should I use the stcurve command and build 4 different graphs (for the combinations of my postulated interaction effects)? Probably you can give me any advice on the suitable commands to best plot those graphs for the event-history methodology. 

I am sorry for the "rookie" questions, but I am still pretty new to this methodology. 

Thanks so much in advance for your help and support.

Best regards,


Florian Christ

WHU – Otto Beisheim School of Management
Doctoral Research Fellow
*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index