Information Technology Reference
In-Depth Information
A. Political Preference
Multiple Comparisons
Dependent Variable: satis public schools
Tukey HSD
Mean
Difference
(I-J)
95% Confidence Interval
(I) political pref
(J) political pref
Std. Error
Sig.
Upper Bound
Lower Bound
e
liberal
moderate
conservative
liberal
conservative
liberal
moderate
8.2500*
1.87528
.000
3.7147
12.7853
12.7500*
1.87528
.000
8.2147
17.2853
moderate
- 8.2500*
1.87528
.000
- 12.7853
- 3.7147
4.5000
1.87528
.052
- .0353
9.0353
conservative
-12.750*
1.87528
.000
- 17.2853
- 8.2147
- 4.5000
1.87528
.052
-9.0353
.0353
Based on observed means.
The mean difference is significant at the .05 level.
*.
B. Homogeneous Subsets
The mean for liberal is in its own column
indicating that it is significantly different from
both the conservative and moderate political
preference groups. However, we see a
nonsignificant
p
value of 1.000 because the
mean is compared to itself. Conservative and
moderate are not significantly different from
each other (Sig. = .052).
satis public schools
Tukey HSD
a,b
Subset
political pref
N
1
2
conservative
moderate
liberal
Sig.
20
20.2500
20
24.7500
20
33.0000
.052
1.000
Means for groups in homogeneous subsets are displayed
Based on Type III Sum of Squares
The error term is Mean Square(Error) = 35.167.
Uses Harmonic Mean Sample Size = 20.000.
a.
b.
Alpha = .05.
Figure 9.16
Post hoc results on the political preference main effect.
9.9.4 MAIN EFFECTS POST HOC ANALYSES: RESIDE
Figure 9.16 presents the results of the post hoc Tukey HSD test on the
main effect of
politics
. As can be seen, liberals were more satisfied with
the public school system than were the other two political groups. Since
the eta squared value associated with this main effect was relatively high
compared to the other values, it is probably worthwhile to mention this
analysis in the results.
9.10 PERFORMING THE OMNIBUS ANALYSIS IN
SAS ENTERPRISE GUIDE
9.10.1 STRUCTURING THE DATA FILE
A portion of the Excel file containing the data for the numerical example
we have been using throughout this chapter is shown in Figure 9.17. That
data file is structured as follows. The first variable is a subject identification
