Example of a contour plot for a binary logistic regression model
main topic     interpreting results     session command     see also 

These data are the results of a survey given to an MBA class at the beginning of the semester. The data contain some characteristics about the students' education and finances and uses them to examine which credit cards students have.

1    Open the worksheet MBASURVEY.MTW.

2    Choose Stat > Regression > Binary Logistic Regression > Fit Binary Logistic Model

3    In Response, enter AmEx.

4    In Continuous predictors enter Cash AIncome.

5    In Categorical predictors enter HDegree.

6    Click Model.

7    In Predictors, highlight Cash.

8    Next to Terms through order, click Add.

9    In Predictors, highlight Cash and AIncome.

10   Next to Interactions through order, click Add.

11   Click OK in each dialog box.

12   Choose Stat > Regression > Binary Logistic Regression > Contour Plot.

13   Click Settings.

14   In Hold categorical variables at, change Setting to Masters.

15   Click OK in each dialog box.

Graph window output

Interpreting the results

You can see that two groups of students have low probabilities of having an American Express card: students with annual incomes below about $50,000 who carry less than about $40 in cash, and students with annual incomes over about $100,000 who carry between $50 and $250 in cash. The contours are steep, so other groups of students are predicted to have over a 90% probability of having an American Express card. The predictions are only for students whose highest level of education is a master's degree. To look at the model for students with other highest levels of education, change the hold value in Settings.

Keep in mind that this plot is based on a model equation. You should be sure that your model is adequate before you interpret the plots.