Information Technology Reference
In-Depth Information
Figure 12.8
All of the variables are now defined.
Click the
Contrasts
pushbutton (we do not show this window). SPSS
uses the
Polynomial
contrast as its default. Since the repeated measures
do not represent equally spaced intervals, we should not perform such
an analysis. Highlight each variable in turn, select
None
from the drop-
down menu, and click
Change
. The contrasts in parentheses next to each
variable name will now say
None
.Click
Continue
to return to the main
dialog window.
Click the
Options
pushbutton to reach the dialog window shown
in Figure 12.9. The top portion of the window is devoted to
Estimated
Marginal Means
. Bring over all of the effects to the panel labeled
Display
Means for
; this will produce a line of
/EMMEANS
TABLES
syntax and
tables of estimated marginal means for these effects. In the bottom portion
of the window labeled
Display
,check
Descriptive statistics
to output the
observed means and standard deviations. Click
Continue
to return to the
main dialog window.
Click the
Plots
pushbutton to set up a plot for the three-way interaction
should it be found to be significant. We intend to ask SPSS to generate
a couple of different plots, not knowing which if we will settle on once
the data analysis is completed. In the first plot, we placed
boxcolor
on
the horizontal axis, used separate lines for the
granules
variable, and
used separate plots for
powcolor
. The separate plots for powder color
correspond to the structure of our data file in which this variable is the
=