Information Technology Reference
In-Depth Information
Yellow Colored Toys
Blue Colored Toys
Gender
Subid
Hands-On
Pretend
Hands-On
Pretend
1
5
6
0
2
2
9
20
0
7
Girls
3
7
12
3
4
4
4
10
5
5
5
6
15
2
3
6
18
4
9
7
7
12
8
6
5
Boys
8
8
7
8
9
9
10
3
7
8
10
14
6
8
11
Figure 15.1
Data for the numerical example.
Looking at the yellow toys, we see that the boys spend more time than
girls with the hands-on toys but the girls spend more time than boys
with the pretend toys. Given that all three factors seem to differentially
affect performance depending on their combination, it is likely that we
will obtain a significant three-way interaction.
15.2 EFFECTS OF INTEREST
The effects of interest in this complex mixed design are the same as in
any three-way design, including the three-way complex mixed design
described in Chapter 14: three main effects (gender, color, and toy type),
three two-way interactions (Gender
×
Color, Gender
×
Toy Type, and
Color
×
Toy Type). What makes this design especially interesting is that it contains
two repeated measures. With more than a single within-subjects factor
in the study, we have multiple within-subjects error terms; each of these
error terms is keyed to a within-subjects effect just as they are for fully
within-subjects designs as discussed in Chapters 11 and 12. This can be
seen in the summary table presented in Table 15.1 that shows the results
of the analysis.
With only one between-subjects factor (Factor A :gender),the
between-subjects portion of the variance is partitioned into the effect
×
Toy Type), and one three-way interaction (Gender
×
Color
Yellow Colored Toys Blue Colored Toys
Hands-On Pretend Hands-On Pretend
Gender
Girls
6.20
12.60
2.00
4.20
Boys
12.40
5.60
7.60
8.00
Figure 15.2
Cell means for the numerical example.
Search WWH ::




Custom Search