Information Technology Reference
In-Depth Information
home and that consider themselves to be liberal. Those who voted in
the; last election (
LSMEAN Number 1
) had a mean satisfaction with
the schools of 36; those who did not vote in the last election (
LSMEAN
Number 4
) had a mean satisfaction with the schools of 35. These two
groups do not differ from each other in the degree of satisfaction they
reported (
p
=
1
.
000).
9.13.3 MAPPING SIMPLE EFFECTS ANALYSES TO THE GRAPH
OF THE INTERACTION
Mapping the simple effects to the plot of the interaction (see Figure 9.3)
was discussed in the context of SPSS in Section 9.9.3, and the reader is
referred to that discussion if needed to determine which data points in the
plot significantly differ from which others.
9.14 COMMUNICATING THE RESULTS
A written summary of the results is as follows.
×
×
2 between-subjects design investigated citizen attitudes toward the
public school system. The independent variables were whether or not there
were children in the home (yes or no), political preference (liberal, moderate,
or conservative), and whether or not participants had voted in the last election
(yes or no). Overall, there were significant main effects for whether there were
children living in the home,
F
(1, 48)
A2
3
2
=
9
.
29,
p
<.
05,
η
=
.
05, and political
2
preference,
F
(2, 48)
=
9
.
29,
p
<.
05,
η
=
.
24. For this latter effect, a Tukey
test indicated that liberals (
M
26) were more satisfied with
thepublicschoolsystemthaneithermoderates(
M
=
33
.
00,
SD
=
6
.
=
24
.
75,
SD
=
10
.
63) or
conservatives (
M
44).
These main effects were obtained in the context of a significant three-
way interaction,
F
(2, 48)
=
20
.
25,
SD
=
11
.
2
20, the plot for which is
presented in Figure 1. Bonferroni adjusted simple effects tests holding the
alpha level at .05 revealed the following.
For those who voted in the last election, liberals were more satisfied than
either of the other two political groups. For those who did not vote conser-
vatives were less satisfied than either of the other two groups. Furthermore,
moderates who voted were less satisfied with the public schools than those
who did not vote; for the other two political groups, whether people voted or
not did not matter with respect to their level of satisfaction.
For those who voted in the last election, conservatives were less satisfied
with the public schools than the other political groups. For those who did not
vote, moderates were less satisfied than the other two groups. Furthermore,
moderates who voted in the last election were more satisfied than those who
did not vote; for the other two political groups, whether people voted or not
did not matter with respect to their level of satisfaction.
=
19
.
56,
p
<.
05,
η
=
.
