Medium
Logistic Regression
SPSS
Free
A researcher uses logistic regression to predict whether a person votes (1 = yes, 0 = no). Predictor PoliticalInterest is coded 1 = low, 2 = medium, 3 = high. SPSS output shows for PoliticalInterest: B = 0.80, Exp(B) = 2.23, p = .01. Which is the best interpretation of Exp(B)?
A
Moving from low to high interest doubles the probability of voting (by 2.23).
B
For each one-step increase in interest (e.g., low→medium), the odds of voting are multiplied by 2.23.
C
People with high interest have odds 2.23 times higher than people with low interest, regardless of how the variable is coded.
D
A one-unit increase in interest increases the odds of voting by 0.80.