## Logistic Regression

Nominal logistic regression measures the probability of choosing a specific response level depending on a facor (independent variable). The y-axis of the probability plot shows smooth curves that partition the total probability (=1) among the various response levels. If the lines that separate different response levels are mostly vertical then knowledge about x will give us good predictability of Y. It the lines are mostly horizontal then the probabilities remain the same for a response across the entire factor ran

#### Uses

**characterize the relationship** between a ordinal or nominal dependent and a continuous independent variable to determine the extent, direction, and strength of the association
**seek a quantitative formula** that predicts the probability to obtain a particular response value as a function of the factor

#### Interpretation

- Caveat:
**causality** is the assumption not the conclusion

#### How this is done

**independent** variable: i.e., X, continuous factor
**dependent** variable: i.e., Y, nominal or ordinal response variable.
- use
** -LogLikelihood Measures** for finding the probabilities for obtaining a particular response value for a given factor

last modified: 2/4/08