Information Technology Reference
In-Depth Information
Descriptive Statistics
Dependent Variable: satis public schools
Std.
Deviation
kids in home vote last elect political pref
yes
Mean
N
f
This is the mean satisfaction score
for those voting in the last election
that have children in the home.
yes
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
liberal
moderate
conservative
Total
36.0000
3.53553
5
14.0000
6.51920
5
10.0000
6.67083
5
20.0000
12.97800
15
no
35.0000
4.52769
5
32.0000
6.51920
5
15.0000
6.51920
5
27.3333
10.64134
15
This is the mean
satisfaction score for
everyone who has
children in the home.
Total
35.5000
3.86580
10
23.0000
11.30388
10
12.5000
6.75360
10
23.6667
12.24276
30
no
yes
33.0000
7.00000
5
35.0000
4.52769
5
20.0000
6.51920
5
This is the mean
satisfaction score for
moderates who did not
vote in the last election
and who have no
children in the home.
29.3333
8.90960
15
no
28.0000
7.51665
5
18.0000
5.70088
5
36.0000
4.00000
5
27.3333
9.38591
15
Total
30.5000
7.33712
10
26.5000
10.18986
10
28.0000
9.85450
10
28.3333
9.04904
30
Total
yes
34.5000
5.46199
10
This is the mean
satisfaction score for
those voting in the last
election.
24.5000
12.26784
10
15.0000
8.15135
10
24.6667
11.92313
30
no
31.5000
6.91616
10
25.0000
9.36898
10
25.5000
12.18606
10
27.3333
9.85877
30
Total
33.0000
6.25763
20
24.7500
10.62705
20
20.2500
11.43805
20
26.0000
10.92967
60
Figure 9.8
The descriptive statistics output.
data from those who have children in the home. Thus, on that
Total
row,
we see the mean of
liberal
to be 35.50. This represents liberal respondents
with children in the home. As another example, the bottom third of the
table is given the label
Total
under the
kids in home
column. Thus, the
means in this portion of the table combine the data from those with and
without children in the home. On the row labeled as
no
under the
vote
last elect
column we see the mean for
conservative
to be 25.50. This
represents conservative participants with no children in the home. It may
take a bit of work, but it is possible to discern all of the means for all of
the effects from this table.
The
GLM Univariate
procedure uses the Levene test to determine if
the assumption of homogeneity of variance has been met. In Figure 9.9,
we see that the
F
ratio for that test is not significant; we may thus conclude
that the data meet the homogeneity of variance assumption.
