The Probabilistic Approach to Preferences Measurement - Probabilistic Composition of Preferences, Theory and Applications - page 28

Table 3.1 Cars probabilities
Beauty Comfort Consumption Power Price Reliability Safety

Car1

0.0705 0.0756 0.1054 0.0034 0.0960 0.0585 0.0018

Car2

0.0705 0.0756 0.1054 0.0034 0.0960 0.0016 0.0621

Car3

0.0705 0.0756 0.1054 0.0034 0.0040 0.0585 0.0621

Car4

0.0705 0.0756 0.0047 0.0881 0.0040 0.0585 0.0621

Car5

0.0705 0.0756 0.0047 0.0881 0.0040 0.0585 0.0018

Car6

0.0705 0.0756 0.0047 0.0881 0.0040 0.0016 0.0621

Car7

0.0705 0.0756 0.0047 0.0034 0.0960 0.0585 0.0621

Car8

0.0705 0.0756 0.0047 0.0034 0.0960 0.0585 0.0018

Car9

0.0705 0.0756 0.0047 0.0034 0.0040 0.0585 0.0621

Car10

0.0705 0.0025 0.1054 0.0881 0.0040 0.0585 0.0621

Car11

0.0705 0.0025 0.1054 0.0034 0.0960 0.0016 0.0621

Car12

0.0705 0.0025 0.0047 0.0881 0.0960 0.0585 0.0621

Car13

0.0705 0.0025 0.0047 0.0881 0.0040 0.0585 0.0621

Car14

0.0705 0.0025 0.0047 0.0034 0.0960 0.0585 0.0621

Car15

0.0022 0.0756 0.1054 0.0881 0.0960 0.0585 0.0018

Car16

0.0022 0.0756 0.1054 0.0881 0.0040 0.0585 0.0621

Car17

0.0022 0.0756 0.1054 0.0034 0.0960 0.0585 0.0621

Car18

0.0022 0.0756 0.0047 0.0881 0.0040 0.0585 0.0621

Car19

0.0022 0.0025 0.1054 0.0881 0.0040 0.0585 0.0621

Car20

0.0022 0.0025 0.0047 0.0881 0.0960 0.0585 0.0621

Table 3.1 presents the results of the seven probabilistic transformations to the
columns of Table 2.4 .
The weighted average scores generated multiplying by the weight of the criterion
the value assigned to each model by each criterion and adding the products of the
multiplications, along with those generated combining by the Choquet integral with
respect to the capacity employed in Chap. 2 , are given in Table 3.2 .
The scores are followed in this table by the ranks derived from them. There is a
considerable agreement between the present ranks and those of Chap. 2 , buta
discrepancy could be noticed in the rank of Car2, which passes from 15 to 9. In the
middle of the ranking, the differences between probabilistic scores are small, what
generates large differences in the ranks. If the probabilistic scores were approxi-
mated to the second decimal, then the probabilistic rank of Car2 would be 11,
re

ecting a tie between eleven models, from the 6th to the 16th.