The data file contains data from a large study to investigate risk-taking behavior. There were three exper-imental conditions: Arnold and Arnold (AA); Coombs (C); and Coombs with no penalty (NC). The experimental conditions differ in terms of what the directions given to students. Each student then responding to two parallel forms of a test given under low and high penalty conditions. There were n1 = 30, n2 = 28, and n3 = 29 students in each experimental condition.
Task 1. Find the mean vector and covariance matrix for each of the three treatments. Does there appear to be differences between the treatments and do you think that the assumption of equal variances is sensible?
Task 2. Perform a test to determine if there is evidence that the covariance matrices are different across groups.
Task 3. Interpret the results of the Wilks test.
Task 4. Compare the within group correlation (0.260) to the correlation between scores on the high and low penalty tests found when ignoring experimental condition. What do you think is the reason for the difference (if there is one).
Task 5. Use the LMATRIX statement to test for significant differences between AA and NC, and C and NC.
Task 6. Interpret the results of the Tukey comparisons for AA versus NC and C versus NC.
The solution provides step by step method for the calculation of MANOVA in SPSS . Formula for the calculation and Interpretations of the results are also included.