From the course: SAS Essential Training: 2 Regression Analysis for Healthcare Research

Unlock the full course today

Join today to access over 24,100 courses taught by industry experts.

Categorizing continuous outcomes

Categorizing continuous outcomes - SAS Tutorial

From the course: SAS Essential Training: 2 Regression Analysis for Healthcare Research

Start my 1-month free trial Buy for my team

Categorizing continuous outcomes

“

- [Instructor] I love logistic regression. Let's say I wanted to run a logistic regression model to predict sleep duration. I'd have to change sleep duration, a continuous variable, into a categorical outcome. Logistic regression is easiest to interpret if the outcome is binary, one or zero, so let's categorize sleep duration into a binary variable that could be used as a dependent variable in logistic regression. You'll see I'm using the exercise file with this movie named 710_Categorizing Continuous Outcomes. See these procs, proc univariate and proc freq? I have them there because I want to remind you to first look at the distribution of your continuous variable before you choose a cut point. After all, you need to be able to defend why you chose that cut point. But since by now, we are very familiar with the distribution of the SLEPTIM1 variable, I thought I'd skip running this code up here and go down to where I made…

Contents