A Fuzzy Implementation for the Socio-Cognitive Approach to Trust - Trust Theory: A Socio-Cognitive and Computational Model - page 326

We have performed three more experiments in order to verify another interesting condition
about learning (250 simulations, 100 tasks). Sometimes it is not possible to learn data in the
same environment where they should be applied. For this reason, we have tested the statistical
trustor by letting them learn without environment and applying data in a normal environment
( EXP2 - positive and negative influences as usual), in an always positive environment ( EXP3 -
only positive influences), and in an always negative environment ( EXP4 - always negative
influences). As easily foreseeable, the mean performance increases in an always positive
environment and decreases in an always negative environment; while this is true for all
strategies, the statistical strategy has more troubles in difficult environments. Figure 11.11

Normal/Normal
No/Normal
No/Neg
No/Pos

44,18
44,18

RANDOM

19,7

67,46

Normal/Normal
No/Normal
No/Neg
No/Pos

72,84

72,84 TRUST

44,42

95,36

Normal/Normal
No/Normal
No/Neg
No/Pos

77,53

77,52 TRUST_L

56,14

96,44

Normal/Normal
No/Normal
No/Neg
No/Pos

62,57

62,57 ABILITY

39,46

79,50

Normal/Normal
No/Normal
No/Neg
No/Pos

61,78

61,78 WILLINGNESS

30,92

88,59

Normal/Normal
No/Normal
No/Neg
No/Pos

64,08

61,20 STAT

34,77

92,20

Figure 11.11 Experiments 1, 2, 3 and 4 compared (measuring the success rates). (Reproduced with
kind permission of Springer Science+Business Media C 2005)