It’s generally easier to think and speak in terms of the odds as opposed to the log odds (discussed previously). We can convert the simple logistic regression model from one that predicts the log odds to one that predicts the odds by exponentiating both sides:

Odds = eβ0+β1*X = (eβ0) * (eβ1*X)

This provides an equivalent, but slightly different interpretation for the coefficients. If, for example, (eβ1) = 2, then we could say, “For a one unit increase in X, the odds increase by a multiplicative factor of 2.” This is an important distinction as effects caused by increases in X multiply the odds but add to the log odds.

Note that in Prism’s results output, odds ratios are reported for β0 and β1. These values are mathematically equivalent to eβ0 and eβ1.