Simple Mixed Design - Analysis of Variance Designs - page 379

Tukey's post-hoc comparisons
indicating democratic scored
greater than authoritarian and
lasses faire.

A. Multiple Comparisons

Measure: MEASURE_1
Tukey HSD

Mean
Difference
(I-J)

95% Confidence Interval

(I)leadership style
authoritarian

(J)leadership style
democratic
lassez faire
authoritarian
lassez faire
authoritarian
democratic

Std. Error Sig.

Lower Bound Upper Bound

-3.10 * .500 .000 -4.43 -1.77

.30 .500 .823 -1.03 1.63

democratic

3.10 * .500 .000 1.77 4.43

3.40 *

.500 .000 2.07 4.73

lassez faire

-.30 .500 .823 -1.63 1.03

-3.40 * .500 .000 -4.73 -2.07

Based on observed means.
*. The mean difference is significant at the .05 level.

Tukey's post-hoc comparisons indicating
democratic scored greater than
authoritarian and lasses faire.

B. MEASURE_1

Tukey HSD a,b,c

Subset

leadership style
lassez faire
authoritarian
democratic

N 1 2

5 3.20

5 3.50

5 6.60

Sig. .823 1.000

Means for groups in homogeneous subsets are displayed
Based on Type III Sum of Squares
The error term is Mean Square(Error) = .625.
a. Uses Harmonic Mean Sample Size = 5.000.

b. The group sizes are unequal. The harmonic

of the group sizes is used. Type I error l
not guaranteed.

c. Alpha = .05.

Figure 13.20 The Tukey results for leadership style.

sufficiently compelling that to address style of leadership without consid-
ering the nature of the projects on which the teams are working would
oversimplify the results. Thus, we will not address this main effect at all
in our narrative of the results.

13.9 PERFORMING THE OMNIBUS ANALYSIS IN SAS
13.9.1 STRUCTURING THE DATA FILE
The data file in an Excel spreadsheet ready to import into a SAS Enterprise
Guide project is displayed in Figure 13.21. It is structured in univariate or
stacked mode as explained in Section 10.15.2, and is similar to those we