Information Technology Reference
In-Depth Information
the within-subjects or error component of S/ABC . Additional sources of
variation include the four interaction effects: A
×
B , A
×
C , B
×
C , and
C .
We agree with the sentiments offered by Keppel and Wickens (2004)
when they note that “you will almost certainly use a computer to perform
the brute numerical work when you analyze a three-factor design. . . .
Nevertheless, it is instructive to look at the computational procedure”
(p. 476). Because of the tedium involved in hand calculations with three
or more factors and the questionable pedagogical payoff, we will forego
our usual elaboration of the numerical computations of this and all future
types of three-factor designs in this text.
However, we will provide you with the computational formulas
for conducting a three-way between-subjects design by hand for those
intrepid students and instructors who desire this additional level of detail.
These formulas can be found in Table 9.2. Further discussion of this topic
can be found in Keppel (1991) and Keppel and Wickens (2004).
A
×
B
×
9.6 PERFORMING THE OMNIBUS ANALYSIS IN SPSS
9.6.1 STRUCTURING THE DATA FILE
A portion of the SPSS data file containing the variables for the analysis
of the numerical example we have been using throughout this chapter is
presented in Figure 9.4. That data file is structured as follows. The first
Figure 9.4
The data file for the three-way numerical example.
Search WWH ::




Custom Search