# Logistic Regression: Dummy Variables

Some concepts get overlooked because it requires an understanding. There are a lot of procedures that a user can run and execute to get the odds ratio. In SAS software, you can construct a logistics regression and it will yield particular odds ratio(s).  However, you should make sure your independent variables can be interpreted in terms of the odds. If your independent variable is age, a continuous variable, it is impossible to summarize each and every odds ratio for a particular age. Therefore , most statisticians will convert the age into discrete groups for odds ratio comparisons. Think of the logistic regression as a series of on and off switches. Each switch represents an independent variable (1=on 0=off). Your main concern is what happens when that switch turns ‘on’ versus ‘off’. You are less concerned about the various gradations of that independent variable. You are simply commenting on (in your interpretation) – this versus that….odd of being in one group versus the other.  So it’s strongly suggested that your independent variables for dichotomous. Keep it simple.

If you enjoyed this short blog there are more to follow-Amy

This site uses Akismet to reduce spam. Learn how your comment data is processed.