To test for equal variances for the whole- plot portion of the model in Question 15, calculate the aver-age for each whole- plot unit. In our study, the whole- plot unit is Subject. Thus, calculate the average score for each of the 12 subjects (the average of the four tests for each subject). Then create an individual value plot (or boxplot) of subject average score by major. Each major will have three values, since there are three subjects for each major. While there is some evidence that the four majors have unequal variances, the small sample sizes and many group comparisons make it difficult to interpret. If the split- plot level variances are equivalent, then we can use this plot to test for equal variances in the whole- plot factors. However, if the split- plot level variances are clearly unequal, it is difficult to interpret whether the unequal variances are caused by subject variability or by submeasurements made within the split-plot.*