UVA Research Data Services: Binary Logistic Regression in R
A binary event either happens or not. For example, a subject’s condition in a clinical trial either improves or does not. Is the probability of improvement related to a novel treatment, or to patient characteristics such as sex, weight, or age? This is what binary logistic regression can do for us. It allows us to model the probability of a binary outcome given various predictors. In this workshop we’ll cover the basics of implementing, interpreting, and evaluating binary logistic regression models using R.
Previous experience with R and linear modeling will be helpful but not required.