Overview for Fit Binary Logistic Model and Binary Logistic Regression

Fit Binary Logistic Model and Binary Logistic Regression perform the same analysis from different menus. Use these analyses to describe the relationship between a set of predictors and a binary response. A binary response has two outcomes, such as pass or fail. You can include interaction and polynomial terms, perform stepwise regression, fit different link functions, and validate the model with a test sample or with cross-validation.

For example, marketers at a cereal company investigate the effectiveness of an ad campaign for a new cereal. The marketers can use binary logistic regression to determine whether people who saw the ad are more likely to buy the cereal.

After you perform the analysis, Minitab stores the model so that you can do the following:

Predict the probability of an event for new or existing observations.
Plot the relationships among the variables.
Find values that optimize multiple responses.

For more information, go to Stored Model Overview.

Where to find this analysis

To fit a binary logistic regression model, choose Stat > Regression > Binary Logistic Regression > Fit Binary Logistic Model.

This analysis has the same capabilities as Predictive Analytics Module > Binary Logistic Regression. The version of the analysis from the Predictive Analytics Module has the following differences.

You access analyses that use the fitted model from the output pane instead of from the menu. Analyses of the fitted model are available for any model that has output in the navigator, not the most recent model only.
The fitted model is available no matter what worksheet is active, so you can predict for columns of data that are in a different worksheet from the response variable.
Minitab Statistical Software saves the model in a project file (*.MPX).

When to use an alternate analysis

If your response variable contains three or more categories that have a natural order, such as strongly disagree, disagree, neutral, agree, and strongly agree, use Ordinal Logistic Regression.
If your response variable contains three or more categories that do not have a natural order, such as scratch, dent, and tear, use Nominal Logistic Regression or CART^® Classification.
If your response variable counts occurrences, such as the number of defects, use Fit Poisson Model.
If your data have a missing value pattern that interferes with the construction of the model or if the binary logistic model does not fit well, consider CART^® Classification.

When to use a predictive analytics model

For some applications, you consider different approaches to model construction. For more information on different types of models, go to Types of predictive analytics models in Minitab Statistical Software. Minitab offers CART^® Regression, TreeNet^® Regression, Random Forests^® Regression, and MARS^® Regression analyses with the Predictive Analytics Module. The Discover Best Model (Continuous Response) analysis compares the performance of different model types in 1 analysis. Click here for more information about how to activate the module.