REVISED

Z03_REND6289_10_IM_MOD3.QXD 5/15/08 8:57 PM Page 279

Decision Theory and the Normal Distribution

3
M O D U L E

TEACHING SUGGESTIONS information to the number you obtained in step two. EVPI will al-
ways be equal to EOL.
Teaching Suggestion M3.1: Reviewing the Normal Curve.
Most of the material in this supplement requires the use of the nor- M3-5.
mal curve. A review of the basic principles of the normal curve
found in the probability chapter (Chapter 2) would be helpful be-
VC/U = $16
fore this module is started. P/U = $24
Teaching Suggestion M3.2: Covering Break-Even Analysis First. FC = $160,000
Covering break-even calculations first helps students get into deci-
sion theory and normal curve analysis. This material will also help
students get back into the fundamental principles of normal curve
theory. Once break-even analysis has been mastered by students,
they should be ready for the rest of the material in this module.
Teaching Suggestion M3.3: Spending More Time on EVPI and
the Normal Distribution.
M = 60,000
EVPI and the normal distribution concepts are difficult for many
= 10,000
students. You may need to spend more time on this topic and rein-
force the basic steps involved. Some instructors reduce coverage
or eliminate this topic. FC 160, 000
a. BE = = = 20, 000 books
P / U VC / U 24 16
SOLUTIONS TO QUESTIONS AND PROBLEMS b. EMV (P/U VC/U)(M) FC
M3-1. The purpose of break-even analysis is to help a manager (24 16)(60,000) 160,000
determine at what point overall revenue will equal overall cost. It
$480,000 $160,000 $320,000
can also help the manager to determine at a certain sales volume
what revenues will be generated. This knowledge can assist the M3-6. a. OLF K(BE X) for X BE
manager in making decisions as to whether or not to introduce a OLF 0 for X BE
new product to the market. where
M3-2. The normal distribution can be used in break-even K (P/U VC/U) 8
analysis when sales are symmetrical around the mean expected de-
mand and follow a bell-shaped distribution (when demand is nor- Thus,
mally distributed), and when there is only one random variable. EOL $8 (20,000 X) for X 20,000
Usually, the normal distribution represents the demand for a new $0 for X 20,000
product.
where X is sales in units.
M3-3. The relationship between EMV and the state of nature
b. EOL KN(D)
must be linear when you use the computations presented in Equa-
tion M3-5 in determining EMV from the mean and the standard with
deviation. When this relationship is not linear, the approach used K $8
in computing EMV cannot be used.  10,000
M3-4. When EVPI is to be computed using a state of nature 60, 000 20, 000
that follows a normal distribution, three steps are required. The D= =4
10, 000
first step is to determine the opportunity loss function. The second
and
step is to determine the opportunity loss using the unit normal loss
integral. The third step is to equate the expected value of perfect N(D) 0.000007145.

279
 REVISED
Z03_REND6289_10_IM_MOD3.QXD 5/15/08 8:57 PM Page 280

280 MODULE 3 DECISION THEORY AND THE NORMAL DISTRIBUTION

Thus, d. EVPI EOL

EOL (8)(10,000)(0.000007145) $0.57160 EOL K N(D)
c. EVPI $0.57160, since EOL EVPI K6
20, 000 60, 000
d. Z = = 4  3,571
10, 000
9, 000 6, 000
standard deviations from  D= = 0.8401
3, 571
A Z value for 4 is not found in the table, but we used 0.99997.
N(.84) .1120
Thus,
P(profit) 0.99997 99.99% Thus, EOL (6)(3571)(0.1120) $2,399.71, and Rudy should
be willing to pay up to $2,399.71 for a marketing research study.
P(loss) 0.00003 0.003%
M3-8. FC $24,000
e. The firm should print the book
VC/U $8
M3-7.
P/U $24
FC 24, 000
a. BE = = = 1, 500 sets
P / U VC / U 24 8
b. If D 2,000, True Lens should produce the lenses.
The expected profit would be
Revenue (2,000 $24/set) $48,000
Less expenses
Fixed cost 24,000
20% 20%
Variable cost (2,000 $8/set) 16,000
6,000 = 9,000 12,000 Total expenses (40,000)
Profit $8,000)
(area to the left of M3-9. EMV ($28 $20)(35,000) $16,000
D 12,000 0.80; $264,000
a. Z= from Appendix A,
No effect.
Z value for 0.80 0.84) $10(30 x) for X 30 where X is

Thus,
12, 000 9, 000
M3-10.
a. OLF
0 { otherwise
actual sales

0.84 =
X 45 30
b. D= =
30
0.84 3,000
 3,571 0.5N(D) N(0.5) 0.1978
6, 000 9, 000 3, 000 EOL KN(D) $10 30 0.1978 $59.34
b. Z = = = 0.84
3, 571 3, 571 c. EVPI EOL $59.34
Using Appendix A gives M3-11. EOL K N(D)
Z(0.84) 0.79955 K $8, $10, or $15
1 0.79955 0.20045  30
Thus, 45 30
D= = 0.5
P(loss) 0.20045 20.045% 30
P(profit) 0.79955 79.955%
N(D) 0.1978
c. EMV (P/U VC/U)() FC
Thus,
($10 $4)(9,000) $36,000
EOL if K $8 (8)(30)(0.1978) $47.47
$54,000 $36,000
EOL if K $10 (10)(30)(0.1978) $59.34
$18,000
EOL if K $15 (15)(30)(0.1978) $89.01
Thus, as the loss per lamp increases, the expected opportunity loss
increases.
 REVISED
Z03_REND6289_10_IM_MOD3.QXD 5/15/08 8:57 PM Page 281

MODULE 3 DECISION THEORY AND THE NORMAL DISTRIBUTION 281

M3-12. a. New EMV ($28 $19)(35,000) $32,000 N(D) 0.08332

$283,000. EVPI EOL K N(D)
Go ahead with new process. $15 200 0.08332 $249.96
b. New EMV ($32 $20)(26,000) $16,000 M3-16.  750;  still 200
$296,000.
750 500
Increase selling price. D=
200
Xb 350 200
M3-13. D= = =1 1.25; N(D) decreased to 0.05059
150
EOL $15 0.05059 200 $151.77
N(D) 0.0833
M3-17. Fixed cost $4,000, Profit per job $40. Break even
EVPI EOL KN(D) $80 150 0.0833 $999.60
point 4,000/40 100 jobs.
The most Joe would be willing to pay is $999.60.
M3-18. The EVPI is equal to the minimum EOL. We use the
M3-14. EVPI EOL K N(D) formula
$500 $100 50 N(D) EOL K  N(D); K 80;  15; BEP 100
N(D) 0.1; from OL tables, D 0.9
X b 5, 000 X b BEP 120 100
D= = D= = = 1.33
50 Xb 4,955 15

Break-even point is 4,955 pumps. N(1.33) 0.0427

M3-15.  700 EVPI EOL K N(D) $80 15 0.0427
750 700 51.24

P ( X < 750) = P Z < = 0.60
M3-19. If the selling price increases to $150, the profit per job
would increase to $70 and the break even point would
750 700
= = 200 be 4,000/70 57.14 units.
0.25
X b 700 500
D= = = 1.00
200