ANOVA Assumptions - Analysis of Variance Designs - page 64

Figure 5.11 Main Explore window.

The left panel of the dialog window contains a list of all the variables in
the analysis. We click over the continuous dependent variable (GAFscore)
to the Dependent List window. The Factor List window allows you to
click over categorical independent variables that will “break” or partition
univariate outliers by each group or level of the independent variable. We
click over our remaining categorical independent variable ( therapy )tothe
Factor List window to obtain these profiles. Below the panel of variables
(on the left side) is the Display panel, which allows you to toggle between
arequestfor Statistics (basic descriptive statistics) or Plots (window plots
or stem-and-leaf plots) for describing each variable. The SPSS default is to
displayboth.Ontherightsideofthedialogwindowarethreepushbuttons:
Statistics , Plots , and Options . These three pushbuttons allow you to
request additional descriptive statistics, plots, and ways of overriding the
default (listwise) missing values option.
Click the Statistics pushbutton and the Explore: Statistics dialog win-
dow is produced (see Figure 5.12). This dialog window is composed of four
check windows, two of which concern us here, Descriptives (the default)
displays basic descriptive statistics, and Outliers displays cases with the
five largest and smallest ExtremeValues for each dependent variable. Each
of these checkboxes has been clicked. Clicking Continue brings you back
to the main dialog window.
A click of the Plots pushbutton produces the Explore: Plots dialog
window (see Figure 5.13). The Boxplots panel produces boxplots for
each group or level of the independent variable. Such plots allow the
investigator to see the spread or variability of the scores within each
level of the independent variable. The Plots panel produces stem-and-leaf
and histogram plots of continuous dependent variables partitioned by