Information Technology Reference
In-Depth Information
The variable color appears as
the second variable. It will
be the variable for the most
global break and must increment
most slowly in the Estimated
Marginal Means tables.We
therefore need to place color at
the start of the list of variables
on this subcommand.
Figure 15.17
The syntax underlying the analysis.
The simple effects analyses will be keyed to the plots we show in
Figure 15.3. Each plot shows the results for a particular toy color, and
thus represents the most global of divisions of the factors. This most
global breakdown needs to correspond to the order that the factors are
specified on the
/EMMEANS
TABLES
subcommand containing the
three-way interaction. The most global break would be the variable that
increments most slowly on the subcommand, that is, the variable that
is furthest toward the left in the parentheses. Therefore, in this instance,
we need to have
color
asthefirstvariableintheparenthesesonthe
/EMMEANS
=
TABLES
subcommand. We note in Figure 15.17 that
color
is the second rather than the first named variable; this can be changed by
simply dragging the variables into (or deleting and typing) their necessary
order. We show this modification in the ordering in Figure 15.18.
We have added our simple effects syntax as shown in Figure 15.19
by duplicating the
/EMMEANS
=
TABLES
subcommand containing the
revised ordering of our three-way interaction. We have added our
compare
syntax focusing on
toytype
and
gender
to the first and second of our
duplicatelines,respectively.Nowweselect
Run
➜
All
from the main
menu.
=
15.7 OUTPUT FOR THE POST-ANOVA ANALYSIS IN SPSS
Thesimpleeffectsoutputisbestviewedwhenlookingattheplotofthe
results in Figure 15.3. In this way, you can relate the statistical outcomes
