SPost9 is a collection of Stata ado files for the post-estimation interpretation of regression models for categorical outcomes. These commands works with Stata 9 and Stata 10; most commands will also work with Stata 8. SPost13 is a set of much more powerful commands based on Stata's margins command. If you are using Stata 11 or later, I highly recommend the new commands which are discussed in the Third Edition of Regression Models for Categorical Dependent Variables Using Stata.
If you use the commands in your work, please cite: J. Scott Long and Jeremy Freese, 2005, Regression Models for Categorical Outcomes Using Stata. Second Edition. College Station, TX: Stata Press.
If you like SPost, you might find The Workflow of Data Analysis Using Stata to be useful.
Warning! The update to Stata 11 for Windows (at least the 64 bit version) on 2010-02-17 has changed the way the Stata interface works. When you alt-tab to Stata, control is in the output window instead of the command window. So, page up scrolls the output rather than scrolling through commands from the Review Windows. I much prefer the earlier behavior and StataCorp plans to fix this inadvertant change. If you install the new release and don't like it, you can rename the new file (say StataMP-64.exe to StataMP-64_argh.exe) and then rename the backup file (say StataMP-64_old.exe to StataMP-64.exe),
Stata 11 news: As of November 17, 2009 we believe that all of the SPost routines work correctly with Stata 10 and Stata 11. They do NOT, however, support factor varaibles. If you have problems, let us know. 17Nov2009
08Oct09: listcoef has been updated to work with mlogit under Stata 11.
01Oct09: The 18Aug09 update to the Stata 10 executable disabled the windowing used by -mlogview-. This has been fixed in the 01Oct09 update. -mlogview- still does not work in Stata 11.
10Sep09: The package spost9_ado does not work with factor variables. We have updated prchange (v 1.8.1) and countfit (0.8.1) to work with Stata 11. We are still working with StataCorp to make SPost work with mlogit under Stata 11. In the short run, the easiest thing is run mlogit under version control. So instead of: "mlogit y a b c" use "version 10.1: mlogit y a b c" Then the SPost commands should work fine. If you encounter other problems using Stata 11, let us know.
SPost and estout: Ben Jann and Scott Long have an article forthcoming in the Stata Journal on using SPost with estout/esttab. Here is a preprint. 08Oct09
About the 2nd Edition
The Second Edition of Regression Models for Categorical Outcomes Using Stata by J. Scott Long and Jeremy Freese began shipping in November 2006. For the full table of contents and purchasing information, click here.
This edition is nearly 50% longer than the previous edition. A few errors in the second edition are corrected here.This edition has a new chapter on estimating and interpreting models for nominal outcomes with alternative-specific data, including the multinomial-probit model, the rank-ordered logit model, and conditional logistic regression. New sections on the estimation and interpretation of the stereotype-logistic model and zero-truncated count models have also been added. Many of the interpretation techniques have been updated to include both point and interval estimates.
Technical details on the models
Many of the procedures implemented in SPost are more fully documented in:
J. Scott Long, 1997, Regression Models for Categorical and Limited Dependent Variables. Advanced Quantitative Techniques in the Social Sciences, Volume 7. Sage Publications. ISBN 0-8039-7374-8.
Our commands are free, but Stata is not. For information on obtaining Stata, contact the folks at Stata. At Indiana University, you can purchase the program from the IU Stat/Math Center. Details on installation are here.
If you don't use Stata
If you don't use Stata, take a look at these Excel spreadsheets that allow you to compute post-estimation predictions.