510 Chapter 14 Repeated-Measures Designs
14.18The SPSS printout in Exhibit 14.2 was obtained by treating the data in Exercise 14.10 as
though all variables were between-subjects variables (i.e., as though the data represented a
standard three-way factorial). Show that the error terms for the correct analysis represent a
partition of the error term for the factorial analysis.MARGINAL 5.90000 5.55000 4.85000 5.43333
COUNT 10 10 10 30
STANDARD DEVIATIONS FOR 1-ST DEPENDENT VARIABLE
GROUP = * 1.0000 * 2.000 * 3.0000
TIME
PRE 1 1.68655 1.76698 1.52388
POST 2 2.16025 2.45855 1.89737
SUM OF DEGREES OF MEAN TAIL
SOURCE SQUARES FREEDOM SQUARE F PROBABILITY
MEAN 1771.26667 1 1771.26667 322.48 0.0000
GROUP 11.43333 2 5.71667 1.04 0.3669
1 ERROR 148.30000 27 5.49259
TIME 19.26667 1 19.26667 9.44 0.0048
TG 20.63333 2 10.31667 5.06 0.0137
2 ERROR 55.10000 272.04074Exhibit 14.1 (continued)Type III
Sum of
Squares
170.800a
1058.400
29.400
68.267
60.400
3.267
.000
.933
8.533
82.800
1312.000
253.600Source
Corrected Model
Intercept
AGE
READER
PA R T
AGE * READER
AGE * PART
READER * PART
AGE * READER * PART
Error
Total
Corrected Totaldf
11
1 1 1 2 1 2 2 248
60
59Mean
Square
15.527
1058.400
29.400
68.267
30.200
3.267
.000
.467
4.267
1.725F
9.001
613.565
17.043
39.575
17.507
1.894
.000
.271
2.473Sig.
.000
.000
.000
.000
.000
.175
1.000
.764
.095Tests of Between-Subjects Effects
Dependent Variable: DVaR Squared 5 .674 (Adjusted R Squared 5 .599)Exhibit 14.214.19 Outline the summary table for an A 3 B 3 C 3 Ddesign with repeated measures on Aand B
and independent measures on Cand D.
14.20 Foa, Rothbaum, Riggs, and Murdock (1991) ran a study comparing different treatments for
posttraumatic stress disorder (PTSD). They used three groups (plus a waiting list control)
One group received Stress Inoculation Therapy (SIT), another received a Prolonged Expo-
sure (PE) treatment, and a third received standard Supportive Counseling (SC). All clients