14.9 Simple effects for data in Exercise 14.6:
14.11 (a) SSreading at child 5 50.7; F 5 11.655*
(b) SSitems at adult good 5 4.133; F 5 3.647.14.13 There would be a very decided lack of independence
among items because an increase in one category would
necessitate a decrease in another—that is, the subject
would have less opportunity to draw from all categories.
14.15 This is a graphic.
14.17 (b) The Ffor MEAN is a test on H 0 : m50.
(c) MSwithin cellis the average of the cell variances.14.19 Source column of summary table for a 4-way ANOVA
with repeated measures on A& B:
SourceBetween Ss
C
D
CD
Ss w/in groups
Within Ss
A
AC
AD
ACD
A 3 Ss w in groups
B
BC
BD
BCD
B 3 Ss w in groups
AB
ABC
ABD
ABCD
AB 3 Ss w in groups
Total14.21 Using the mixed models procedure.
The covariance matrix shows a decreasing pattern in co-
variances as you move to the right. Therefore I re-
quested that the solution include a covariance matrix
that was AR1. The results follow:>>>SSprob at noncalc=11.20 F=5.51*SSprob at calc =78.53 F=112.19*SSgroup at mult =52.9 F=28.86*SSgroup at subt =3.6 F=1.96SSgroup at add =0.9 F 61Fixed Effects
Type II Tests of Fixed Effectsa
Numerator Denominator
Source df df F Sig.
Intercept 1 43.256 422.680 .000
Group 2 43.256 3.521 .038
Time 3 81.710 71.356 .000
Group 3 Time 4 81.710 5.578 .001
aDependent Variable: dv.14.23Fixed Effects
Type II Tests of Fixed Effectsa
Numerator Denominator
Source df df F Sig.
Intercept 1 41.724 393.118 .000
Group 2 41.724 2.877 .068
Time 3 70.480 640760 .000
Group 3 Time 4 70.459 5.266 .001
aDependent Variable: dv.
14.25 (a) SPSS printout on gain scores (see next page).
(b) SPSS printout using pretest and posttest (see next
page).
(c) The Fcomparing groups on gain scores is exactly
the same as the Ffor the interaction in the re-
peated measures design.
(e) t 52 0.287. This group did not gain significantly
over the course of the study. This suggests that any
gain we see in the other groups can not be attrib-
uted to normal gains seen as a function of age. f.
Without the control group we could not separate
gains due to therapy from gains due to maturation.
14.27 (a) t 52 0.555. There is no difference in Time 1
scores between those who did, and did not, have a
score at Time 2.
(b) If there had been differences, I would have wor-
ried that people did not drop out at random.
14.29 Differences due to judges play an important role.
14.31 If I was particularly interested in differences between
subjects, and recognized that judges probably didn’t
have a good anchoring point, and if this lack was not
meaningful, I would not be interested in considering
differences among judges.
14.33 Strayer et al. (2006) (see next page).
t1–2 5 0.07; t1–3 5 2.52*; t2–3 5 2.45* Both the Baseline
and Alcohol conditions show poorer performance than
the cell phone condition, but they do not differ from
each other.Answers 749