Example of a Generalized Linear Mixed Model
This example illustrates using the Generalized Linear Mixed Model personality of the Fit Model platform to evaluate two teaching programs. In an educational study, classes of students from multiple schools were assigned to one of two programs. At the end of the program, students took an evaluation test for which they receive a grade of Pass or Fail.
1. Select Help > Sample Data Folder and open Student Testing.jmp.
2. Select Analyze > Fit Model.
3. Select Grade and click Y.
When you add this column as Y, the fitting Personality becomes Nominal Logistic.
4. Select Generalized Linear Mixed Model from the Personality list. Alternatively, you can select the Generalized Linear Mixed Model personality first, and then click Y to add Grade.
5. Select Program and click Add on the Fixed Effects tab.
Figure 9.2 Completed Fit Model Launch Window Showing Fixed Effects
6. Select the Random Effects tab.
7. Select School and Class and click Add.
8. Select School from the Select Columns list, select Class from the Random Effects tab, and then click Nest.
Figure 9.3 Completed Fit Model Launch Window Showing Random Effects Tab
9. Click Run.
The Generalized Linear Mixed Model report is shown in Figure 9.4. The effect of the teaching program is not statistically significant. However, the class effect nested within school is statistically significant.
Figure 9.4 Generalized Linear Mixed Model Report Window
10. Click the red triangle next to Binomial and select Conditional Model Inference > Conditional Profiler.
You can use the Conditional Model Profiler to explore the differences in pass rates for various classes and schools.
Figure 9.5 Conditional Model Profiler
