Assessing the value of information to design site investigation and construction quality assurance programs - Risk and Reliability in Geotechnical Engineering - page 486

-15

Perfect test for B excluding Δ C test
Pilot test for B excluding Δ C test

-20

-25

Preferred alternative
in prior decision

Alternative A

-30

Alternative B

-35
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Prior probability that alternative B works

Figure 13.8 Sensitivity of expected monetary values to prior probability that in situ bioremediation will
work for contaminated lagoon.

A sensitivity analysis for the value of information is shown in Figures 13.8 and 13.9 . The
expected monetary value for Alternatives A and B is shown in Figure 13.8 versus the prior
probability that Alternative B will work, P ( W B ) in Figure 13.3 . The preferred alternative is
Alternative A at P ( W B ) < 0.55 and Alternative B at P ( W B ) > 0.55 ( Figure 13.8 ); the alterna-
tives are equally preferable at P ( W B ) = 0.55. The expected monetary values for the pilot test
and the perfect test without including the cost of the test are also shown in Figure 13.8 ; the
value of information is the distance between the expected monetary value for a test without
including the cost of the test and the expected monetary value for the preferred alternative
in the prior decision. The value of information versus the prior probability that Alternative B
will work is shown in Figure 13.9 . The value of information is a maximum when there is the
greatest potential to change the decision, which is at the indifference point of P ( W B ) = 0.55
( Figure 13.9 ). The value of information for the pilot test is zero when the prior probability
that Alternative B will work is small [ P ( W B ) less than about 0.15] or large [ P ( W B ) greater
than about 0.95]; at these extremes, there is no possible outcome from the pilot test that will
change the decision about whether or not to implement Alternative B.

6

Perfect test for
alternative B

5

4

3

2

Pilot test for
alternative B

1

0 0.0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Prior probability that alternative B works

Figure 13.9 Sensitivity of value of information to prior probability that in situ bioremediation will work for
contaminated lagoon.