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: suest vs interaction term

From   Ankit Sakhuja <>
Subject   st: suest vs interaction term
Date   Sun, 2 Feb 2014 12:46:12 -0600

I am trying to assess if the odds ratio of mortality for those with
condition X in comparison to those without condition X are different
between males and females. Which is a better method to understand that
in a multivariable logistic model:

1. Adding an interaction term between condition X and variable for sex
in the multivariable logistic model
2. Running 2 separate multivariable logistic regression models, each
for a subpopulation of males or females and then comparing the OR for
mortality for condition X in the two models by -suest-.

I understand that by this method an interaction term is implicitly
added between variable for sex and all explanatory variables but then
which one would give a more meaningful and accurate answer?

*   For searches and help try:

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