Information Technology Reference
In-Depth Information
Trustfulness
.02
External Factors
Internal Factors
.5
.5
.01
.14
.5
Opportunity
.6
Danger
0.2
.8
Ability
.04
.05
1.0
.61
.5
.5
.5
Unharmfulness
.5
.5
.5
.5
.0
.0
.5
.05
.5
.3
.5
.5
.5
.0
Availability
DE
DE
.0
DE
.0
.0
.5
C
.7
.0
.0
.5
O/R
.12
.5
O/R
C
.5
.0
.2
C
R
R
O/R
.0
.2
.2
O/R
R
DE
.5
.0
.5
.5
C
R
.5
.0
.3
.0
DE
O/R
Positive Impact
.0
C
R
Negative Impact
Figure 11.7
Emergency Visit FCMs for the Doctor. (Reproduced with kind permission of Springer
Science+Business Media
C
2003)
The results also change drastically: trustfulness for the doctor is
only slightly positive
(
0.29)
(see Figure 11.7 and 11.8).
The FCMs are very stable; altering some settings for the doctor (Ability - Direct Experience:
very good
and Danger - Categorization:
only little danger
) results in a change in the trustfulness
value that becomes
almost good
but does not overcome the machine's one. We obtain the same
results if we suppose that Doctor's Ability - Direct Experience:
perfect
and Ability's Causal
Power:
very strong
.
On the contrary, if we introduce a big danger (
+
0.02)
and for the machine it is
quite good (
+
1) either internal (
harmfulness
) or external
(
danger
) in each FCM the trustfulness values fall to
negative
in both cases (respectively
+
−
0.59
and
−
0.74 for the doctor; and
−
0.52 and
−
0.67 for the machine).
11.12.3 Trustfulness and Decision
We consider three steps: evaluation (i.e. how much trust do I have); decision (to assign or not
assign a task); delegation (make the decision operative). Obtaining the trustfulness values is
only the first step. In order to make the final choice (e.g. between a doctor and a machine in
our scenarios) we have to take into account other factors, mainly costs and possible saturation
thresholds for the various features.
Search WWH ::
Custom Search