Exercises for Chapter 9

This exercise utilizes the data set jobsatisfaction_b.sav, which can be downloaded from this website:

Previously we conducted a discriminant analysis on job satisfaction, with had four categories. Both unsatisfied groups were relatively small. Conduct a stepwise discriminant analysis with satjob3 that combines the unsatisfied employees and therefore only has three groups: 1 = very satisfied, 2 = moderately satisfied, 3 = not satisfied. Use the same IVs.

IVsage, Rincome20, educ, and hrs1

1. Develop a research question.

2. Which predictor variables show significant group differences?

3. Can homogeneity of covariance be assumed?

4. Which variables were entered into the model?

5. How many functions were generated? Why?

6. Is (are) the function(s) significant? Explain.

7. Calculate the effect size for each function.

8. Evaluating the function coefficients and correlation coefficients, what would you label the function(s)?

9. What is the discriminant function equation?

10. How accurate is (are) the function(s) in predicting income level for each group and the total sample?

12. What is your conclusion regarding these results?