Regression · Binary outcomes
The Logistic Curve
How can a linear model on the log-odds scale produce valid probabilities?
Page 1 of 6 · Question
Why must a binary-outcome curve bend?
A binary-outcome probability must remain between 0 and 1. Logistic regression lets predictors combine linearly on the unbounded log-odds scale, then transforms the result back to probability.
Use the left and right arrow keys to move between pages.