Information Technology Reference
In-Depth Information
Table 1.
ANOVA result table based on
df
M
=1. Sorted by
F
.
05
ratios
η
2
dependend Variable
df
R
SS
T
SS
B
SS
M
SS
R
F
.
05
corrections
V
7
15 119.22
42.72
57.78
18.72
46.30
0.76
silence
V
2
15 398.55 129.58 167.92 101.05
24.93
0.62
insights
Q
9
16
18.24
14.9
0.84
2.50
5.36
0.25
reviews
V
6
15
38.01
23.00
3.13
11.88
3.95 0.21
talking
V
1
15 116.56
56.92
10.86
48.79
3.34 0.18
fun
Q
3
16
18.31
15.03
0.55
2.73
3.24 0.17
commitment
Q
5
16
24.68
20.90
0.33
3.45
1.52 0.09
clarity
Q
8
16
32.78
25.78
0.12
6.88
0.27 0.02
motivation
Q
4
16
10.90
9.46
0.05
1.39
0.23 0.04
Table 2.
(one-tailed) t-test comparing groups by method. Ordered like Table 1
dependent
Effect Size
Significance
Confidence Intervals
variable
t.BPM interview
lower boundary upper boundary
corrections
V
7
3.00
0.31
.000
1.85
3.53
silence
V
2
5.54
0.95
.000
2.63
6.54
insights
Q
9
3.75
3.43
.017
0.03
0.60
reviews
V
6
0.81
0.19
.033
-.046
1.30
talking
V
1
4.65
3.49
.044
-0.19
2.52
fun
Q
3
4.16
3.90
.046
-0.05
0.56
commitment
Q
5
3.31
3.51
.118
-0.53
0.14
clarity
Q
8
3.37
3.49
.304
-0.59
0.36
motivation
Q
4
4.45
4.37
.225
-0.14
0.29
between the groups is in that range with 95 percent probability. It should not
include zero to be sure about the effect between the groups.
From both tables we see, that all parameters for
corrections
V
7
,
silence
V
2
and
insights
Q
9
meet scientific standards. For
reviews
V
6
,
talking
V
1
and
fun
Q
3
we
see that they just missed acceptable standards in both tests. E.g. the difference
between the groups is significant in Table 2 but the confidence intervals do not
allow acceptance by rigor scientific standards. Finally,
commitment
Q
5
,
clarity
Q
8
and
motivation
Q
4
did not hold.
5.5 Testing Potentially Influential Factors
We use a two-tailed dependent t-test to compare groups were two different influ-
ences were applied. For example, we had two processes to report on, two exper-
imenters, and two different educational groups. Furthermore, each subject goes
through the experimental task twice. Repetition effects might have influenced
the performance of the subjects.
While the experimenters had no significant influence on the dependent vari-
ables, we found that the second experimental task led to significantly more
clarity
Q
8
about the goal (1st=3.1, 2nd=3.77, p=.001) and more
commitment
Q
5
to the solution (1st=3.2, 2nd=3.63, p=.004).