CHAPTER 8

DECISION ANALYSIS

SOLUTIONS TO DISCUSSION QUESTIONS

8-1. The purpose of this question is to make students use a personal experience to distinguish between
good and bad decisions. A good decision is based on logic and all of the available information. A bad
decision is one that is not based on logic and the available information. It is possible for an unfortunate or
undesirable outcome to occur after a good decision has been made. It is also possible to have a favorable
or desirable outcome occur after a bad decision.

8-2. The decision-making process includes the following steps: (1) define the problem, (2) list the
alternatives, (3) identify the possible outcomes, (4) evaluate the consequences, and (5) select an
evaluation criterion and make the appropriate decision. The first four steps or procedures are common for
all decision-making problems. Step 5, however, depends on the decision-making model used.

8-3. An alternative is a course of action over which we have complete control. A state of nature is an
event or occurrence in which we have no control. An example of an alternative is deciding whether or not
to take an umbrella to school or work on a particular day. An example of a state of nature is whether or
not it will rain on a particular day.

8-4. The basic differences between decision-making models under certainty, risk, and uncertainty depend
on the amount of chance or risk that is involved in the decision. A decision-making model under certainty
assumes that we know with complete confidence the future outcomes. Decision-making-under-risk
models assume that we do not know the outcomes for a particular decision but that we do know the
probability of occurrence of those outcomes. With decision making under uncertainty, it is assumed that
we do not know the outcomes that will occur, and furthermore, we do not know the probabilities that
these outcomes will occur.

8-5. EMV is the expected monetary value. This is the expected return that we would realize if the
decision were repeated an infinite number of times. EVwPI is the expected value with perfect
information. This is the return or value of making the same decision an infinite number of times when we
have perfect or complete information. EVPI is the expected value of perfect information. This is simply
the difference between EMV and EVwPI. It is the amount that we would be willing to pay for perfect
information.

8-6. A decision tree is preferred to a decision table when a number of sequential decisions are to be made.
A sequential decision situation is one in which the outcome of one decision becomes an important factor
in making future decisions. For example, if a decision maker is considering the possibility of acquiring
additional information and a decision of whether or not to build a new plant, the decision to acquire the
new information is made first. Then, based on the results of the new information (if it is gathered), the
decision to build the plant is made. Therefore, these decisions are sequential. One is made before the
other.

8-7. A prior probability is one that exists before additional information is gathered. A posterior
probability is one that can be computed based on prior probabilities and additional information.
8-8. The purpose of Bayesian analysis is to determine posterior probabilities based on prior probabilities
and new information. Bayesian analysis can be used in the decision-making process whenever additional
information is gathered. This information can then be combined with prior probabilities in arriving at
posterior probabilities. Once these posterior probabilities are computed, they can be used in the decision-
making process as any other probability value.

8-9. The overall purpose of utility theory is to incorporate a decision maker’s preference for risk in the
decision-making process.

8-10. A utility function can be assessed in a number of different ways. A common way is to use a
standard gamble. With a standard gamble, the best outcome is assigned a utility of 1, and the worst
outcome is assigned a utility of 0. Then, intermediate outcomes are selected and the decision maker is
given a choice between having the intermediate outcome for sure and a gamble involving the best and
worst outcomes. The probability that makes the decision maker indifferent between having the
intermediate outcome for sure and a gamble involving the best and worst outcomes is determined. This
probability then becomes the utility of the intermediate value. This process is continued until utility
values for all economic consequences are determined. These utility values are then placed on a utility
curve.

8-11. When a utility curve is to be used in the decision-making process, utility values from the utility
curve replace all monetary values at the terminal branches in a decision tree or in the body of a decision
table. Then, expected utilities are determined in the same way as expected monetary values. The
alternative with the highest expected utility is selected as the best decision.

8-12. A risk seeker is a decision maker who enjoys and seeks out risk. A risk avoider is a decision maker
who avoids risk even if the economic payoff is higher. The utility curve for a risk seeker increases at an
increasing rate. The utility curve for a risk avoider increases at a decreasing rate.
SOLUTIONS TO PROBLEMS

8-13. See file P8-13.XLS.

Outcomes (a) Maximax (b) Maximin (c) EMV
Alternatives Fav mkt Unfav mkt Maximum Choice Minimum Choice EMV Choice
Sub 100 $300,000 -$200,000 $300,000 Best -$200,000 $150,000 Best
Oiler J $250,000 -$100,000 $250,000 -$100,000 $145,000
Texan $75,000 -$18,000 $75,000 -$18,000 Best $47,100

8-14. See file P8-14.XLS.

(a) Maximax (b) Maximin (c) Equally Likely (d) Hurwicz (0.6)
Alternatives Maximum Choice Minimum Choice Average Choice Realism Choice
Expand $56,000 -$29,000 $16,000 $22,000
Move $70,000 Best -$45,000 $20,000 Best $24,000 Best
No change $30,000 $5,000 Best $15,000 $20,000

Regret Outcomes (e) Minimax

Alternatives Good Average Poor Maximum Choice
$14,00 $34,00
Expand 0 $14,000 0 $34,000 Best
$50,00
Move $0 $0 0 $50,000
$40,00
No change 0 $25,000 $0 $40,000

8-15. See file P8-15.XLS.

Alternatives EMV Choice
Expand $14,750
Move $19,750 Best
No change $13,500

Expected Value WITH Perfect Information (EVwPI) = $34,750

Best Expected Monetary Value (EMV) = $19,750
Expected Value OF Perfect Information (EVPI) = $15,000

Alternatives EOL Choice

Expand $20,000
Move $15,000 Best
No change $21,250

8-16. See file P8-16.XLS.

(c) Equally
(a) Maximax (b) Maximin Likely (d) Hurwicz (0.5) (e) Minimax
Alternatives Maximum Choice Minimum Choice Average Choice Realism Choice Maximum Choice
Stocks 12% Best -10% 3% 1% 12%
 Bonds 7% 1% 4% Best 4% Best 5% Best
Money Mkt 4% 2% Best 3% 3% 8%

8-17. See file P8-17.XLS.

Alternatives EMV Choice
Stocks 0.00%
Bonds 3.25% Best
Money Market 2.75%

Expected Value WITH Perfect Information (EVwPI) = 5.40%

Best Expected Monetary Value (EMV) = 3.25%
Expected Value OF Perfect Information (EVPI) = 2.15%

Alternatives EOL Choice

Stocks 5.40%
Bonds 2.15% Best
Money Market 2.65%

8-18. See file P8-18.XLS.

(a) Outcomes (b) Equally Likely (c) Hurwicz (0.75)
Alternatives Grow Same Average Choice Realism Choice
Large wing $150,000 -$85,000 $32,500 Best $91,250 Best
Small wing $60,000 -$45,000 $7,500 $33,750
No wing $0 $0 $0 $0

8-19. See file P8-19.XLS.

(c) Equally
(a) Maximax (b) Maximin Likely (d) Hurwicz (0.4) (e) Minimax
Alternatives Maximum Choice Minimum Choice Average Choice Realism Choice Maximum Choice
100 $75 $75 Best $75 $75 Best $150
200 $150 -$50 $83 Best $30 $125 Best
300 $225 Best -$175 $25 -$15 $250

8-20. See file P8-20.XLS.

Alternatives EMV Choice
100 $75 Best
200 $70
300 -$5

Expected Value WITH Perfect Information (EVwPI)

= $139
Best Expected Monetary Value (EMV) = $75
Expected Value OF Perfect Information (EVPI) = $64
 Alternatives EOL Choice
100 $64 Best
200 $69
300 $144

8-21. See file P8-21.XLS.

(c) Equally
(a) Maximax (b) Maximin Likely (d) Hurwicz (0.7) (e) Minimax
Alternatives Maximum Choice Minimum Choice Average Choice Realism Choice Maximum Choice
50 $500 $500 Best $500 $500 $1,500
100 $1,000 -$300 $675 Best $610 $1,000 Best
150 $1,500 -$1,100 $525 $720 $1,600
200 $2,000 Best -$1,900 $50 $830 Best $2,400

8-22. See file P8-22.XLS.

Alternatives EMV Choice

50 $500
100 $740 Best
150 $525
200 -$15

Expected Value WITH Perfect Information (EVwPI)

= $1,225
Best Expected Monetary Value (EMV) = $740
Expected Value OF Perfect Information (EVPI) = $485

Alternatives EOL Choice

50 $725
100 $485 Best
150 $700
$1,24
200 0

8-23. See file P8-23.XLS.

Outcomes
Alternatives 6 7 8 9 EMV Choice
6 $300 $300 $300 $300 $300.00
7 $255 $350 $350 $350 $340.50
8 $210 $305 $400 $400 $352.50 Best
9 $165 $260 $355 $450 $317.00
Probability 0.1 0.3 0.5 0.1
8-24. See file P8-24.XLS.
(c) Equally
(a) Maximax (b) Maximin Likely (d) Hurwicz (0.8) (e) Minimax
Alternatives Maximum Choice Minimum Choice Average Choice Realism Choice Maximum Choice
Small $50,000 -$10,000 Best $20,000 $38,000 $250,000
Medium $80,000 -$20,000 $30,000 $60,000 $220,000
Large $100,000 -$40,000 $30,000 $72,000 $200,000
Very large $300,000 Best -$160,000 $55,000 Best $208,000 Best $150,000 Best

8-25. See file P8-25.XLS. Note: In this problem, lower times are the preferred outcome. Therefore we
select the alternative with the lowest EMV. Calculate regret as (time taken - best time taken), then select
alternative with lowest EOL. EVPI = Best EMV – EVwPI. Alternatively, we can multiply all payoffs by
-1, and solve this problem like any other problem.

PAYOFFS Outcomes
Alternatives None Mild Severe EMV Choice
Tennessee 15 30 45 25.00
Back roads 20 25 35 24.17 Best
Expressway 30 30 30 30.00

Best Expected Time (EMV) = 24.17

Expected Value WITH Perfect Information (EVwPI) = 20.83
Expected Value OF Perfect Information (EVPI) = 3.33

REGRET Outcomes
Alternatives None Mild Severe EOL Choice
Tennessee 0 5 15 4.17
Back roads 5 0 5 3.33 Best
Expressway 15 5 0 9.17

8-26.
(a) EMV if we construct the clinic = 0.5 * $100,000 + 0.5 * (-$40,000) = $30,000. EMV if we do nothing
= $0. Therefore, construct clinic.

(b) See file P8-26.XLS for the TreePlan solution. Conduct study. Construct clinic if result is positive. Do
not construct clinic if result is negative. EMV = $36,140

(c) EVSI = $11,140. Thus, the physicians would pay up to $11,140 more for the survey. Note: Since
EVSI should be calculated assuming no cost to gather the sample information, $5,000 had to be added
back to $36,140.

(d) EVwPI = $50,000. Best EMV = $30,000. EVPI = $20,000. Efficiency = 55.70%.

8-27.
(a) See file P8-27.XLS, sheet (a) for the utility curve. They are risk avoiders.
(b) See file P8-27.XLS, sheet (a) for the Tree Plan solution. Expected utility if we conduct the survey =
0.761. Expected utility if we do not conduct the survey = 0.90. Therefore, the medical professionals
should not conduct the survey, and should not construct the clinic.

8-28. See file P8-28.XLS for Tree Plan solution.

(a) Conduct survey. If results are favorable, build large shop. If the results are unfavorable, don't build
any shop.

(b) EVSI = $11,000. EVwPI = $35,400. Best EMV = $19,000. EVPI = $16,400. Efficiency = 67.07%.

8-29. See file P8-29.XLS for the TreePlan solution. Note: All costs have been expressed as negative
amounts in the decision tree. Hence, we select highest expected values at all nodes.
(a) Use supplier A. Expected cost is $90 with supplier A versus $113 with supplier B. Note: In the tree,
we pick the EMV of -$90 over the EMV of -$113.

(b) $37 + ($113 - $90) = $60 less than supplier A.

8-30. See file P8-30.XLS for the TreePlan solution.

Market the new golf balls. If competitor enters market, set price Medium. If competitor does not enter
market, set the price High. EMV = $325,000.

8-31. See file P8-31.XLS for the TreePlan solution.

Hire Samantha Adams. If she says cold, Ajay should buy a new blower. If she says not cold, Ajay should
repair the old blower. Expected profit = $176.88. EVSI = $101.88. EVwPI = $340.00. Maximum EMV =
$125.00. EVPI = $215.00. Efficiency = 47.38%.

8-32. See file P8-32.XLS for the TreePlan solution.

(a) Hire Susan. If she says good chance of being sunny, take lemonade. If she says bad chance of being
sunny, take cocoa. Expected profit = $50.36.

(b) EVSI = $10.36

8-33. See file P8-33.XLS for the TreePlan solution.

Build the pilot plant. If the pilot plant succeeds, build facility. If the pilot plant fails, don't build facility.
Expected profit = $209,500.

8-34.
(a)
Prior Probabilities
P(Facility works) = 0.60
P(Facility fails) = 0.40
Conditional probabilities
P(Pilot works | Facility works) = 0.80
 P(Pilot fails | Facility works) = 0.20
P(Pilot works | Facility fails) = 0.15
P(Pilot fails | Facility fails) = 0.85

Posterior probabilities GIVEN pilot works

Outcome P(Pilot works | Outcome) Prior prob Jt prob Post. prob
Facility works 0.80 0.60 0.48 0.89
Facility fails 0.15 0.40 0.06 0.11
P(Pilot works) = 0.54
Posterior probabilities GIVEN pilot fails
Outcome P(Pilot fails | Outcome) Prior prob Jt prob Post. prob
Facility works 0.20 0.60 0.12 0.26
Facility fails 0.85 0.40 0.34 0.74
P(Pilot fails) = 0.46

(b) See file P8-34.XLS for the revised TreePlan solution. Build the pilot plant. If the pilot plant succeeds,
build facility. If the pilot plant fails, don't build facility. Expected profit = $244,300.

8-35. See file P8-35.XLS for the TreePlan solution. Accept the wager. Use a hard first serve. If the first
serve is out of play, use a hard second serve. Expected payoff = $17.60.

8-36.
(a) See file P8-36.XLS, sheet (a) for the TreePlan solution. Accept the wager. Use a hard first serve. If the
first serve is out of play, use a hard second serve. Expected utility = 0.588.

(b) See file P8-36.XLS, sheet (b) for the utility curve. You are a risk avoider.

8-37. See file P8-37.XLS for the TreePlan solution.

Test the land. If the test result is positive, drill for oil. If the test result is negative, sell the land. Expected
profit = $565,000.

8-38.
(a)
Prior Probabilities
P(Oil well) = 0.20
P(Dry well) = 0.80
Conditional probabilities
P(Positive test | Oil well) = 0.85
P(Negative test | Oil well) = 0.15
P(Positive test | Dry well) = 0.25
P(Negative test | Dry well) = 0.75

Posterior probabilities GIVEN positive test

Outcome P(Positive test | Outcome) Prior prob Jt prob Post. prob
Oil well 0.85 0.20 0.17 0.46
Dry well 0.25 0.80 0.20 0.54
P(Positive test) = 0.37
 Posterior probabilities GIVEN negative test
Outcome P(Negative test | Outcome) Prior prob Jt prob Post. prob
Oil well 0.15 0.20 0.03 0.05
Dry well 0.75 0.80 0.60 0.95
P(Negative test) = 0.63

(b) See file P8-38.XLS for the TreePlan solution.

Test the land. If the test result is positive, drill for oil. If the test result is negative, sell the land. Expected
profit = $427,300.

8-39.
(a) See file P8-39.XLS, sheet (a) for the TreePlan solution. Test the land. If the test result is positive, drill
for oil. If the test result is negative, sell the land. Expected utility = 0.246.

(b) See file P8-39.XLS, sheet (b) for the utility curve. Shamrock Oil is a risk seeker.

8-40.
(a) See file P8-40.XLS for the TreePlan solution. Conduct the survey questionnaire. If the response is
positive, produce razor. If the response is negative, do not produce razor. Expected return = $24,160.

(b) EVPI = $30,000. EVSI = $9,160. Efficiency = 30.53%.

(c) EVPI = $30,000. EVSI = $17,080. Efficiency = 56.93%.

8-41.
(a) See file P8-41.XLS, sheet (a) for the TreePlan solution. Conduct the survey questionnaire. If the
survey response is positive, produce razor. If the survey response is negative, do not produce razor.
Expected utility = 0.823.

(b) See file P8-41.XLS, sheet (b) for the utility curve. Jim is a risk avoider.

8-42. See file P8-42.XLS for the TreePlan solution.

Don't lock-in now. If the rate increases after one month, lock-in then. If the rate is unchanged or decreases
after one month, don't lock-in and accept current rate at settlement. Expected cost = $270.

8-43.
(a) See file P8-43.XLS, sheet (a) for the utility curve. Jason is a risk seeker.

(b) See file P8-43.XLS, sheet (b) for the TreePlan solution. The use of utilities does not affect Jason’s
decision. The recommended course is still to not lock-in now. If the rate increases after one month, lock-
in then. If the rate is unchanged or decreases after one month, don't lock-in and accept current rate at
settlement. Expected utility = 0.261.

8-44. See file P8-44.XLS for the TreePlan solution.

(a) Get the information. If it is favorable, build the store. If it is unfavorable, don't build the store.
Expected return = $29,200.

(b) EVPI = $32,000. EVSI = $21,200. Efficiency = 66.25%.

8-45.
(a) See file P8-45.XLS, sheet (a) for the TreePlan solution. Get the information. If it is favorable, build
the store. If it is unfavorable, don't build the store. Expected utility = 0.62.
(b) See file P8-45.XLS, sheet (b) for the utility curve. Sue is a risk seeker.

8-46. See file P8-46.XLS for the TreePlan solution.

If the survey is not conducted, the medium facility, with an EMV of $670,000, is selected. If the survey is
conducted, we first compute the revised probabilities using Bayes’ theorem, as follows.

Prior Probabilities
P(Low demand) = 0.15
P(Fair demand) = 0.40
P(High demand) = 0.45
Conditional probabilities
P(Low survey | Low demand) = 0.70
P(Fair survey | Low demand) = 0.20
P(High survey | Low demand) = 0.10
P(Low survey | Fair demand) = 0.40
P(Fair survey | Fair demand) = 0.50
P(High survey | Fair demand) = 0.10
P(Low survey | High demand) = 0.10
P(Fair survey | High demand) = 0.30
P(High survey | High demand) = 0.60

Posterior probabilities GIVEN low survey

Outcome P(Low survey | Outcome) Prior prob Jt prob Post. prob
Low demand 0.70 0.15 0.105 0.339
Fair demand 0.40 0.40 0.160 0.516
High demand 0.10 0.45 0.045 0.145
P(Low survey) = 0.310
Posterior probabilities GIVEN fair survey
Outcome P(Fair survey | Outcome) Prior prob Jt prob Post. prob
Low demand 0.20 0.15 0.030 0.082
Fair demand 0.50 0.40 0.200 0.548
High demand 0.30 0.45 0.135 0.370
P(Fair survey) = 0.365
Posterior probabilities GIVEN high survey
Outcome P(High survey | Outcome) Prior prob Jt prob Post. prob
Low demand 0.10 0.15 0.015 0.046
Fair demand 0.10 0.40 0.040 0.123
High demand 0.60 0.45 0.270 0.831
P(High survey) = 0.325
If the survey results are low, the best decision is to build the medium facility with an EMV of $495,000.
If the survey results are medium, the best decision is also to build the medium plant with an EMV of
$646,000. Finally, if the survey results are high, the best decision is to build the large facility with an
EMV of $821,000. Now, we can take these results and multiply them times the probability of having low
survey results, medium survey results, and high survey results to get the expected value of taking the
survey and selecting the appropriate decision. The resulting EMV is approximately $656,000. Because
the EMV for not conducting the survey is greater ($670,000), the best decision is to not conduct the
survey and to build the medium-sized facility.

8-47. See file P8-47.XLS.

(a) Posterior probabilities GIVEN favorable survey
Outcome P(Favorable | Outcome) Prior prob Jt prob Post. prob
Success 0.80 0.50 0.400 0.727
Failure 0.30 0.50 0.150 0.273
P(Favorable survey) = 0.550
(b) Posterior probabilities GIVEN unfavorable survey
Outcome P(Unfavorable | Outcome) Prior prob Jt prob Post. prob
Success 0.20 0.50 0.100 0.222
Failure 0.70 0.50 0.350 0.778
P(Unfavorable survey) = 0.450

8-48. See file P8-48.XLS.

Posterior probabilities GIVEN favorable survey
Outcome P(Favorable | Outcome) Prior prob Jt prob Post. prob
Fav market 0.875 0.550 0.481 0.855
Unfav market 0.182 0.450 0.082 0.145
P(Favorable survey) = 0.563
Posterior probabilities GIVEN unfavorable survey
Outcome P(Favorable | Outcome) Prior prob Jt prob Post. Prob
Fav market 0.125 0.550 0.069 0.157
Unfav market 0.818 0.450 0.368 0.843
P(Unfavorable survey) = 0.437

8-49. See file P8-49.XLS.

Posterior probabilities GIVEN growth prediction
Outcome P(Growth predn | Outcome) Prior prob Jt prob Post. prob
Actual growth 0.75 0.30 0.225 0.696
Actual steady 0.18 0.45 0.081 0.250
Actual decline 0.07 0.25 0.018 0.054
P(Growth prediction) = 0.324
Posterior probabilities GIVEN steady prediction
Outcome P(Steady predn | Outcome) Prior prob Jt prob Post. prob
Actual growth 0.08 0.30 0.024 0.058
Actual steady 0.80 0.45 0.360 0.870
Actual decline 0.12 0.25 0.030 0.072
P(Steady prediction) = 0.414
Posterior probabilities GIVEN decline prediction
Outcome P(Decline predn | Outcome) Prior prob Jt prob Post. prob
Actual growth 0.05 0.30 0.015 0.054
Actual steady 0.12 0.45 0.054 0.195
Actual decline 0.83 0.25 0.208 0.750
P(Decline prediction) = 0.277
8-50.
(a) See file P8-50.XLS, sheet (a). Travel time on expressway = 30 minutes. Expected travel time on
Broad Street = (40)(0.5) + (15)(0.5) = 27.5 minutes. Use Broad Street.
(b) See file P8-50.XLS, sheet (b). Utility of travel time on expressway = 0.7. Expected utility on
Broad Street = (0.2)(0.5) + (0.9)(0.5) = 0.55. The expressway maximizes utility.
(c) See file P8-50.XLS, sheet (c) for the utility curve. Lynn is a risk avoider.

Case: Ski Right

(1) See file P8-Ski.XLS. The best decision is to have Leadville Barts make the helmets and have
Progressive Products make the rest. Expected value = $2,600. The option of not using Progressive was,
however, very close with an expected value of $2,500.

PAYOFFS Outcomes
Alternatives Poor Average Good Excellent EMV Choice
PP -$5,000 -$2,000 $2,000 $5,000 $700
LB and PP -$10,000 -$4,000 $6,000 $12,000 $2,600 Best
TR and PP -$15,000 -$10,000 $7,000 $13,000 $900
CC and PP -$30,000 -$20,000 $10,000 $30,000 $1,000
LB, CC, and TR -$60,000 -$35,000 $20,000 $55,000 $2,500
Probability 0.1 0.3 0.4 0.2

(2 and 3) The opportunity loss and EVPI computations are as follows.

Expected Value WITH Perfect Information (EVwPI) = $17,900

Best Expected Monetary Value (EMV) = $2,600
Expected Value OF Perfect Information (EVPI) = $15,300

REGRET Outcomes
Alternatives Poor Average Good Excellent EOL Choice
PP $0 $0 $18,000 $50,000 $17,200
LB and PP $5,000 $2,000 $14,000 $43,000 $15,300 Best
TR and PP $10,000 $8,000 $13,000 $42,000 $17,000
CC and PP $25,000 $18,000 $10,000 $25,000 $16,900
LB, CC, and TR $55,000 $33,000 $0 $0 $15,400
Probability 0.1 0.3 0.4 0.2

(4) Clearly, there are a number of options that Bob did not consider.

Case: Blake Electronics

(1) MAI’s proposal directly gives Steve the conditional probabilities he needs (e.g., probability of a
successful venture given a favorable survey). Although the information from Iverstine and Kinard (I&K)
is different, we can easily use Bayes’ theorem to on I&K information to compute the revised probabilities
(see file P8-Blake.XLS, sheet Posterior). As such, does not need any additional information from I&K.
(2) Steve’s problem involves three decisions. First, should he contract the services of an outside research
agency? Second, if a survey is warranted, should he employ MAI or I&K? Third, in any case, should the
new product line be introduced?

The TreePlan solution for Steve’s problem is shown in file P8-Blake.XLS.

If Steve decides not to conduct a survey, the decision is to introduce the product with an EMV of
$700,000 [= (0.6)($1,500,000) + (0.4)(-$500,000)].

If Steve decides to conduct the survey, he has to choose between MAI and I&K. If he chooses MAI for
the survey, the best choice is to introduce the product irrespective of whether the survey results are
favorable or unfavorable. The EMV is $800,000 if the survey results are favorable, while the EMV is
only $200,000 if the survey results are unfavorable. The overall EMV of hiring MAI is $500,000 [= (0.5)
($800,000) + (0.5)($200,000)].

If Steve chooses I&K for the survey, the best choice is to introduce the product if survey results are
favorable, for an EMV of $940,000. On the other hand, if the survey results are unfavorable, the best
decision is to not introduce the product for an EMV of -$300,000 (the cost of the survey). The overall
EMV of hiring MAI is $468,800 [= (0.62)($940,000) + (0.38)(-$300,000)].

Comparing these alternatives, Steve should not hire either firm to do the survey. He should simply choose
to introduce the produce right away for an EMV of $700,000.