Examining the Selected Cases

As a first step in examining the cases in node 9, you might want to look at the variables not used in the model. In this example, all variables in the data file were included in the analysis, but two of them were not included in the final model: education and car loans. Since there's probably a good reason why the procedure omitted them from the final model, they probably won't tell us much, but let's take a look anyway.

From the menus choose:
Analyze > Descriptive Statistics > Crosstabs...

Figure 2. Crosstabs dialog box
Select Credit rating for the row variable.
Select Education and Car loans for the column variables.
Click Cells.
Figure 3. Crosstabs Cell Display dialog box
In the Percentages group, select (check) Row.
Then click Continue, and in the main Crosstabs dialog box, click OK to run the procedure.

Examining the crosstabulations, you can see that for the two variables not included in the model, there isn't a great deal of difference between cases in the good and bad credit rating categories.

Figure 4. Crosstabulations for cases in selected node

For education, slightly more than half of the cases with a bad credit rating have only a high school education, while slightly more than half with a good credit rating have a college education—but this difference is not statistically significant.
For car loans, the percentage of good credit cases with only one or no car loans is higher than the corresponding percentage for bad credit cases, but the vast majority of cases in both groups has two or more car loans.

So, although you now get some idea of why these variables were not included in the final model, you unfortunately haven't gained any insight into how to get better prediction for node 9. If there were other variables not specified for the analysis, you might want to examine some of them before proceeding.