60. A nuclear power company is deciding whether to build a nuclear power plant at Diablo Canyon or at Roy Rogers City.

however, and an earthquake occurs at Diablo during the next five years, construction will be terminated and the compan
company believes there is a 20% chance that an earthquake will occur at Diablo during the next five years. For $1 million
that an earthquake will not occur. The geologists past record indicates that she will predict an earthquake on 95% of the
Should the power company hire the geologist? Also, calculate and interpret EVSI and EVPI.

Diablo $10,000,000
Roy Rogers $20,000,000
Geologist $1,000,000

P(Earthquake) 0.20
P(No earthquake) 0.80

P(Negative | No earthquake) 0.90

P(Negative | Earthquake) 0.05
P(Positive | No earthquake) 0.10
P(Positive | Earthquake) 0.95

No earthquake Earthquake Total

Negative 0.72 0.01 0.73
Positive 0.08 0.19 0.27
Total 0.80 0.20

P(Earthquake | Positive) = P(Positive | Earthquake) * P(

P(Positive|Earthquake) * P(Earthquake) + P(Positive

P(Earthquake | Negative) = P(Negative | Earthquake) * P

P(Negative|Earthquake) * P(Earthquake) + P(Negativ

0.27
Positive

To hire

Negative
0.73

To not hire
Diablo
EVSI:

EVSI = Minimum expected cost with sample information - Minimum exp

= (-13,900,000 + 1,000,000) - (-14,000,000)
= $1,100,000

Yes, the company should hire the geologist as it reduces the expected cost by $100,000. The cost o
hence it is worth it to hire the geologist.
Depending on the prediction, the company can then decide on where to build the nuclear power pl
Diablo Canyon. If she predicts an earthquake, then the plant should be build in Roy Rogers City ins

EVPI:
Since perfect prediction information does not exist, it will be impossible to allocate a cost to it.

EVPI = Minimum expected cost with sample information - Minimum exp

= (-12,000,000) - (-14,000,000)
= $2,000,000

Even though an exact cost of a perfect survey cannot be allocated, the company should not pay m

0.20
Positive
EMV:
$12,000,000

To hire

EMV: Negative
$12,000,000 0.80

To not hire EMV: $14,000,000

Diablo
at Diablo Canyon or at Roy Rogers City. The cost of building the power plant is $10 million at Diablo and $20 million at Roy Rogers Cit
ion will be terminated and the company will lose $10 million (and will still have to build a power plant at Roy Rogers City). Without fu
uring the next five years. For $1 million, a geologist can be hired to analyze the fault structure at Diablo Canyon. She will predict eith
ll predict an earthquake on 95% of the occasions for which an earthquake will occur and no earthquake on 90% of the occasions for
nd EVPI.

P(Positive | Earthquake) * P(Earthquake) = 0.19

ke) * P(Earthquake) + P(Positive|No Earthquake) * P(No Earthquake) 0.27

P(Negative | Earthquake) * P(Earthquake) = 0.01

ke) * P(Earthquake) + P(Negative|No Earthquake) * P(No Earthquake) 0.73

0.7037037037 EMV
Got earthquake
$31,000,000
No earthquake 0.296296296
Diablo
$11,000,000
Roy Rogers $21,000,000

0.0136986301
Got earthquake $31,000,000

Diablo No earthquake $11,000,000

0.9863013699
Roy Rogers $21,000,000

0.20
Got earthquake $30,000,000
 No earthquake 0.80 $10,000,000

Roy Rogers
$20,000,000

mple information - Minimum expected cost without sample information

ed cost by $100,000. The cost of geologist of $1,000,000 is $100,000 lesser than the EVSI of $1,100,000,

re to build the nuclear power plant. If she predicts no earthquake, then the company should build the plant a
be build in Roy Rogers City instead.

sible to allocate a cost to it.

mple information - Minimum expected cost without sample information

the company should not pay more than $2,000,000 for this information despite it is perfect.

Diablo $30,000,000

Roy Rogers $20,000,000

$10,000,000
Diablo

Roy Rogers
$20,000,000

0.20
Got earthquake $30,000,000

No earthquake 0.80 $10,000,000

Roy Rogers
$20,000,000
 0 million at Roy Rogers City. If the company builds at Diablo,
oy Rogers City). Without further expert information, the
anyon. She will predict either that an earthquake will occur or
n 90% of the occasions for which an earthquake will not occur.

= 0.7037037037

= 0.0136986301

Weighted
Probabilities EMV Minimum
average

0.7037 $21,814,815
$25,074,074 \\
$21,000,000
0.2963 $3,259,259 $21,000,000
1.00 $21,000,000

0.0137 $424,658
$11,273,973
0.9863 $10,849,315 $11,273,973
$21,000,000 \\
1.00 $21,000,000

0.20 $6,000,000
$14,000,000
 0.80 $8,000,000 $14,000,000

$20,000,000 \\
1.00 $20,000,000

VSI of $1,100,000,

hould build the plant at

Probabilit Weighted
EMV Minimum EMV
y average

0.27 $5,670,000

$13,900,000

0.73 $8,230,000 $13,900,000

$14,000,000 \\
1.00 $14,000,000
76. Sharp Outfits is trying to decide whether to ship some customer orders now via UPS or wait until after the threat of a
the company will incur $60,000 in delay and shipping costs. If Sharp Outfits decides to ship the customer orders via UPS
customer orders via UPS, the company will incur $10,000 in delay costs regardless of whether UPS goes on strike. Let p r
a. For which values of p, if any, does Sharp Outfits minimize its expected total cost by choosing to postpone shipping its
b. Suppose now that, at a cost of $1000, Sharp Outfits can purchase information regarding the likelihood of a UPS strike i
information will predict a strike with probability 0.75, and if there will not be a strike, the information will predict no strike
cost?

a. Use the tree from part b to find the EVSI when p = 0.15. Then use a data table to find EVSI for p from 0.05 to 0.30 in in

b. Continuing part b, compute and interpret the EVPI when p = 0.15.

a. Expected amount Probability

Ship during UPS Strike $60,000 0.1785714286
No strike $4,000 0.8214285714
Ship after threat is over $14,000 1

To minimize its expected total cost, Sharp Outfits should postpone shipping its custome

b. Shipping cost $4,000

Delay & shipping cost if got strike $60,000
Postpone shipping delay cost $10,000
Information cost 1000

P(Strike) 0.15
P(No strike) 0.85

P(Negative | No strike) 0.85

P(Negative | Strike) 0.25
P(Positive | No strike) 0.15
P(Positive | Strike) 0.75

P(Strike | Positive) = P(Positive | Strike) *

P(Positive|Strike) * P(Strike) + P(Positiv

P(Strike | Negative) = P(Negative | Strike) *

P(Negative | Strike) * P(Strike) + P(Negat

0.24
Positive
 Purchase

Negative
0.76

Don't purchase
Ship

To minimize its expected total cost, Sharp Outfits should first purchase the information r
strike, then Sharp Outfits should postpone shipping its customer orders via UPS. If it ass
shipping customer orders via UPS.

EVSI:

EVSI = Minimum expected cost with sample information - Minimum expec

= (-8,540 + 1,000) - (-12,400)
= $2,460

Alternative EMV with info EMV without info EVSI

$8,540 $10,000
0.05 $1,020
0.1 #VALUE!
0.15 #VALUE!
0.2 #VALUE!
0.25 #VALUE!
0.3 #VALUE!

EVPI:
Since perfect prediction information does not exist, it will be impossible to allocate a co

EVPI = Minimum expected cost with sample information - Minimum expec

= (-5,500) - (-12,400)
= $7,500

Even though an exact cost of a perfect survey cannot be allocated, the company should
 0.15
Positive
EMV:
$4,900

Purchase

EMV: Negative
$4,900 0.85

Don't purchase EMV: $12,400

Ship
 until after the threat of another UPS strike is over. If Sharp Outfits decides to ship the requested merchandise now and the UPS strik
customer orders via UPS and no strike occurs, the company will incur $4000 in shipping costs. If Sharp Outfits decides to postpone s
UPS goes on strike. Let p represent the probability that UPS will go on strike and impact Sharp Outfitss shipments.
to postpone shipping its customer orders via UPS?
likelihood of a UPS strike in the near future. Based on similar strike threats in the past, the company assesses that if there will be a s
ation will predict no strike with probability 0.85. Provided that p = 0.15, what strategy should Sharp Outfits pursue to minimize its ex

or p from 0.05 to 0.30 in increments of 0.05, and chart EVSI versus p.

Weighted average EMV

$10,714 To find p, goal seek to make the EM
$3,286 14000 UPS equals to the EMV of shipping a
$14,000 14000 is over.

shipping its customer orders via UPS when p is more than 0.1786.

No strike Strike Total

Negative 0.72 0.0375 0.76
Positive 0.13 0.1125 0.24
Total 0.85 0.15

P(Positive | Strike) * P(Strike) = 0.1125

* P(Strike) + P(Positive|No Strike) * P(No Strike) 0.24

P(Negative | Strike) * P(Strike) = 0.0375

* P(Strike) + P(Negative | No Strike) * P(No Strike) 0.76

EMV
Strike
$61,000
No strike
Ship
$5,000
Postpone $11,000

Strike $61,000
 Ship No strike $5,000

Postpone $11,000

Strike $60,000

No strike $4,000

Postpone
$10,000

hase the information regarding the likelihood of a UPS strike. If it assess that there will be a
rders via UPS. If it assess that there will not be a strike, then Sharp Outfits should continue

tion - Minimum expected cost without sample information

EVSI Graph of EVSI vs p

$1,200

$1,000

$800

$600

$400

$200

$0
0.05 0.1 0.15 0.2 0.25 0.3
p

sible to allocate a cost to it.

tion - Minimum expected cost without sample information

, the company should not pay more than $6,900 for this information despite it is perfect.
Ship $60,000

Postpone $10,000

$4,000
Ship

Postpone
$10,000

Strike 0.15 $60,000

No strike 0.85 $4,000

Postpone
$10,000
ndise now and the UPS strike takes place,
Outfits decides to postpone shipping its
hipments.

esses that if there will be a strike, the

fits pursue to minimize its expected total

al seek to make the EMV of shipping during

o the EMV of shipping after threat of UPS strike

= 0.46875

= 0.0493421

Weighted
Probabilities EMV Minimum Probability
average

0.4688 $28,594
$31,250 \\
$11,000 0.24
0.5313 $2,656 $11,000
1.00 $11,000

0.0493 $3,010
 $7,763
0.9507 $4,753 $7,763 0.76
$11,000 \\
1.00 $11,000

0.15 $9,000
$12,400
0.85 $3,400 $10,000 1.00

$10,000 \\
1.00 $10,000

0.3
p
Weighted Minimum
EMV
average EMV

$2,640

$8,540
 $5,900 $8,540

$10,000 \\

$10,000