1. Spreadsheet Modeling & Decision Analysis:
A Practical Introduction to Management Science, 3e
by Cliff Ragsdale

2. Chapter 15 Decision Analysis
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

3. Introduction to Decision Analysis
Models help managers gain insight and understanding, but they can’t make decisions.
Decision making often remains a difficult task due to:
uncertainty regarding the future
conflicting values or objectives
Consider the following example...
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

4. Deciding Between Job Offers
Company A
In a new industry that could boom or bust.
Low starting salary, but could increase rapidly.
Located near friends, family and favorite sports team.
Company B
Established firm with financial strength and commitment to employees.
Higher starting salary but slower advancement opportunity.
Distant location, offering few cultural or sporting activities.
Which job would you take?
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

5. Good Decisions vs. Good Outcomes
A structured approach to decision making can help us make good decisions, but can’t guarantee good outcomes.
Good decisions sometimes result in bad outcomes.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

6. Characteristics of Decision Problems
Alternatives - different courses of action intended to solve a problem.
Work for company A
Work for company B
Reject both offers and keep looking
Criteria - factors that are important to the decision maker and influenced by the alternatives.
Salary
Career potential
Location
States of Nature - future events not under the decision makers control.
Company A grows
Company A goes bust
etc
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

7. An Example: Magnolia Inns
Hartsfield International Airport in Atlanta, Georgia, is one of the busiest airports in the world.
It has expanded numerous times to accommodate increasing air traffic.
Commercial development around the airport prevents it from building additional runways to handle the future air traffic demands.
Plans are being developed to build another airport outside the city limits.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

8. An Example: Magnolia Inns (con’t)
Two possible locations for the new airport have been identified, but a final decision is not expected to be made for another year.
The Magnolia Inns hotel chain intends to build a new facility near the new airport once its site is determined.
Land values around the two possible sites for the new airport are increasing as investors speculate that property values will increase greatly in the vicinity of the new airport.
See data in file Fig15-1.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

9. The Decision Alternatives
1) Buy the parcel of land at location A.
2) Buy the parcel of land at location B.
3) Buy both parcels.
4) Buy nothing.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

10. The Possible States of Nature
1) The new airport is built at location A.
2) The new airport is built at location B.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

11. Constructing a Payoff Matrix
See file Fig15-1.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

12. Decision Rules
If the future state of nature (airport location) were known, it would be easy to make a decision.
Failing this, a variety of nonprobabilistic decision rules can be applied to this problem:
Maximax
Maximin
Minimax regret
No decision rule is always best and each has its own weaknesses.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

13. The Maximax Decision Rule
Identify the maximum payoff for each alternative.
Choose the alternative with the largest maximum payoff.
See file Fig15-1.xls
Weakness
Consider the following payoff matrix
State of Nature
Decision MAX
A <--maximum B
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

14. The Maximin Decision Rule
Identify the minimum payoff for each alternative.
Choose the alternative with the largest minimum payoff.
See file Fig15-1.xls
Weakness
Consider the following payoff matrix
State of Nature
Decision MIN A
B <--maximum
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

15. The Minimax Regret Decision Rule
Compute the possible regret for each alternative under each state of nature.
Identify the maximum possible regret for each alternative.
Choose the alternative with the smallest maximum regret.
See file Fig15-1.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

16. Anomalies with the Minimax Regret Rule
Consider the following payoff matrix
State of Nature
Decision
A 9 2
B 4 6
State of Nature
Decision MAX
A <--minimum B
The regret matrix is:
Note that we prefer A to B.
Now let’s add an alternative...
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

17. Adding an Alternative Consider the following payoff matrix
State of Nature
Decision
A 9 2
B 4 6
C 3 9
The regret matrix is:
State of Nature
Decision MAX A
B <--minimum C
Now we prefer B to A?
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

18. Probabilistic Methods
Sometimes, the states of nature can be assigned probabilities that represent their likelihood of occurrence.
For decision problems that occur more than once, we can often estimate these probabilities from historical data.
Other decision problems (such as the Magnolia Inns problem) represent one-time decisions where historical data for estimating probabilities don’t exist.
In these cases, probabilities are often assigned subjectively based on interviews with one or more domain experts.
Highly structured interviewing techniques exist for soliciting probability estimates that are reasonably accurate and free of the unconscious biases that may impact an expert’s opinions.
We will focus on techniques that can be used once appropriate probability estimates have been obtained.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

19. Expected Monetary Value
Selects alternative with the largest expected monetary value (EMV)
EMVi is the average payoff we’d receive if we faced the same decision problem numerous times and always selected alternative i.
See file Fig15-1.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

20. EMV Caution
The EMV rule should be used with caution in one-time decision problems.
Weakness
Consider the following payoff matrix
State of Nature
Decision EMV
A 15, ,000 5,000 <--maximum
B 5,000 4,000 4,500
Probability
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

21. Expected Regret or Opportunity Loss
Selects alternative with the smallest expected regret or opportunity loss (EOL)
The decision with the largest EMV will also have the smallest EOL.
See file Fig15-1.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

22. The Expected Value of Perfect Information
Suppose we could hire a consultant who could predict the future with 100% accuracy.
With such perfect information, Magnolia Inns’ average payoff would be:
EV with PI = 0.4*$ *$11 = $11.8 (in millions)
Without perfect information, the EMV was $3.4 million.
The expected value of perfect information is therefore,
EV of PI = $ $3.4 = $8.4 (in millions)
In general, EV of PI = EV with PI - maximum EMV
It will always be the case that, EV of PI = minimum EOL
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

23. A Decision Tree for Magnolia Inns1 2 3 4
Buy A
-18
Buy B
-12
Buy A&B
-30
Buy nothing
Land Purchase Decision
Airport Location A B
Payoff 13 -8 11 5 -1 31 6 23 35 29
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

24. Rolling Back A Decision Tree1 2 3 4
Buy A
-18
Buy B
-12
Buy A&B
-30
Buy nothing
Land Purchase Decision
Airport Location A B
Payoff 13 -8 11 5 -1 31 6 23 35 29
0.4
0.6
EMV=-2
EMV=3.4
EMV=1.4
EMV= 0
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

25. Alternate Decision Tree
Land Purchase Decision
Airport Location
Payoff
0.4 A 13 31
Buy A 1
-18
EMV=-2 6 B
-12
0.6
0.4 A -8 4
Buy B 2
-12
EMV=3.4 23 B 11
0.6
0.4 A 5 35
Buy A&B
EMV=3.4 3
-30
EMV=1.4 29 B -1
0.6
Buy nothing
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

26. Using TreePlan TreePlan is an Excel add-in for decision trees.
See file Fig15-14.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

27. About TreePlan
TreePlan is a shareware product developed by Dr. Mike Middleton at the University of San Francisco and distributed with this textbook at no charge to you.
If you like this software package and plan to use for more than 30 days, you are expected to pay a nominal registration fee. Details on registration are available near the end of the TreePlan help file.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

28. Multi-stage Decision Problems
Many problems involve a series of decisions
Example
Should you go out to dinner tonight?
If so,
How much will you spend?
Where will you go?
How will you get there?
Multi-stage decisions can be analyzed using decision trees
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

29. Multi-Stage Decision Example: COM-TECH
Steve Hinton, owner of COM-TECH, is considering whether or not to apply for a $85,000 OSHA research grant for using wireless communications technology to enhance safety in the coal industry.
Steve would spend approximately $5,000 preparing the grant proposal and estimates a chance of actually receiving the grant.
If awarded the grant, Steve would then need to decide whether to use microwave, cellular, or infrared communications technology.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

30. COM-TECH (continued)
Steve would need to acquire some new equipment depending on which technology is used. The cost of the equipment is summarized as:
Technology Equipment Cost
Microwave $4,000
Cellular $5,000
Infrared $4,000
Steve knows he will also spend money in R&D, but he doesn’t know exactly what the R&D costs will be.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

31. COM-TECH (continued)
Cost Prob. Cost Prob.
Microwave $30, $60,
Cellular $40, $70,
Infrared $40, $80,
Best Case Worst Case
Steve estimates the following best case and worst case R&D costs and probabilities, based on his expertise in each area.
Steve needs to synthesize all the factors in this problem to decide whether or not to submit a grant proposal to OSHA.
See file Fig15-26.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

32. Analyzing Risk in a Decision Tree
How sensitive is the decision in the COM-TECH problem to changes in the probability estimates?
We can use Solver to determine the smallest probability of receiving the grant for which Steve should still be willing to submit the proposal.
Let’s go back to file Fig15-26.xls...
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

33. Risk Profiles A risk profile summarizes the make-up of an EMV.
The $13,500 EMV for COM-TECH was created as follows:
Event Probability Payoff
Receive grant, Low R&D costs 0.5*0.9=0.45 $36,000
Receive grant, High R&D costs 0.5*0.1=0.05 -$4,000
Don’t receive grant 0.5 -$5,000
EMV $13,500
This can also be summarized in a decision tree.
See file Fig15-29.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

34. Using Sample Information in Decision Making
We can often obtain information about the possible outcomes of decisions before the decisions are made.
This sample information allows us to refine probability estimates associated with various outcomes.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

35. Example: Colonial Motors
Colonial Motors (CM) needs to determine whether to build a large or small plant for a new car it is developing.
The cost of constructing a large plant is $25 million and the cost of constructing a small plant is $15 million.
CM believes a 70% chance exists that demand for the new car will be high and a 30% chance that it will be low.
The payoffs (in millions of dollars) are summarized below.
Demand
Factory Size High Low
Large $175 $95
Small $125 $105
See decision tree in file Fig15-32.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

36. Including Sample Information
Before making a decision, suppose CM conducts a consumer attitude survey (with zero cost).
The survey can indicate favorable or unfavorable attitudes toward the new car. Assume:
P(favorable response) = 0.67
P(unfavorable response) = 0.33
If the survey response is favorable, this should increase CM’s belief that demand will be high. Assume:
P(high demand | favorable response)=0.9
P(low demand | favorable response)=0.1
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

37. Including Sample Information (con’t)
If the survey response is unfavorable, this should increase CM’s belief that demand will be low. Assume:
P(low demand | unfavorable response)=0.7
P(high demand | unfavorable response)=0.3
See decision tree in file Fig15-33.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

38. The Expected Value of Sample Information
How much should CM be willing to pay to conduct the consumer attitude survey?
Expected Value of Sample Information
Expected Value with Sample Information
Expected Value without Sample Information = -
In the CM example,
E.V. of Sample Info. = $ $126 = $0.82 million
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

39. Computing Conditional Probabilities
Conditional probabilities (like those in the CM example) are often computed from joint probability tables.
High Demand Low Demand Total
Favorable Response
Unfavorable Response
Total
The joint probabilities indicate:
The marginal probabilities indicate:
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

40. Computing Conditional Probabilities (cont’d)
High Low
Demand Demand Total
Favorable Response
Unfavorable Response
Total
In general,
So we have,
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

41. Bayes’s Theorem
Bayes’s Theorem provides another definition of conditional probability that is sometimes helpful.
For example,
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

42. Utility Theory
Sometimes the decision with the highest EMV is not the most desired or most preferred alternative.
Consider the following payoff table,
State of Nature
Decision EMV
A 150, , ,000 <--maximum
B 70, ,000 55,000
Probability
Decision makers have different attitudes toward risk:
Some might prefer decision alternative A,
Others would prefer decision alternative B.
Utility Theory incorporates risk preferences in the decision making process.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

43. Common Utility Functions
risk averse
1.00
risk neutral
0.75
risk seeking
0.50
0.25
0.00
Payoff
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

44. Constructing Utility Functions
Assign utility values of 0 to the worst payoff and 1 to the best.
For the previous example,
U(-$30,000)=0 and U($150,000)=1
To find the utility associated with a $70,000 payoff identify the value p at which the decision maker is indifferent between:
Alternative 1: Receive $70,000 with certainty.
Alternative 2: Receive $150,000 with probability p and lose $30,000 with probability (1-p).
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

45. Constructing Utility Functions (cont’d)
If decision maker is indifferent when p=0.8:
U($70,000)=U($150,000)*0.8+U(-30,000)*0.2 =1*0.8+0*0.2=0.8
When p=0.8, the expected value of Alternative 2 is:
$150,000*0.8 + $30,000*0.2 = $114,000
The decision maker is risk averse. (Willing to accept $70,000 with certainty versus a risky situation with an expected value of $114,000.)
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

46. Constructing Utility Functions (cont’d)
If we repeat this process with different values in Alternative 1, the decision maker’s utility function emerges (e.g., if U($40,000)=0.65):
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

47. Comments
Certainty Equivalent -- the amount that is equivalent in the decision maker’s mind to a situation involving risk.
(e.g., $70,000 was equivalent to Alternative 2 with p = 0.8)
Risk Premium -- the EMV the decision maker is willing to give up to avoid a risky decision.
(e.g., Risk premium = $114,000-$70,000 = $44,000)
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

48. Using Utilities to Make Decisions
Replace monetary values in payoff tables with utilities.
Consider the following utility table from the earlier example,
State of Nature Expected
Decision Utility A
B <--maximum
Probability
Decision B provides the greatest utility even though it the payoff table indicated it had a smaller EMV.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

49. The Exponential Utility Function
The exponential utility function is often used to model classic risk averse behavior:
U(x)
1.00
0.80
R=100
R=200
0.60
0.40
R=300
0.20
0.00
-0.20
-0.40
-0.60
-0.80
-50
-25 25 50 75
100
125
150
175
200
225
250
275
300
325
350 x
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

50. Incorporating Utilities in TreePlan
TreePlan will automatically convert monetary values to utilities using the exponential utility function.
First determine a value for the risk tolerance parameter R.
R is equivalent to the maximum value of Y for which the decision maker is willing to accept the following gamble:
Win $Y with probability 0.5,
Lose $Y/2 with probability 0.5.
Note that R must be expressed in the same units as the payoffs!
In Excel, insert R in a cell named RT.
(Note: RT must be outside the rectangular range containing the decision tree!)
On TreePlan’s ‘Options’ dialog box select, ‘Use Exponential Utility Function’
See file Fig15-38.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

51. Multicriteria Decision Making
Decision problem often involve two or more conflicting criterion or objectives:
Investing: risk vs. return
Choosing Among Job Offers: salary, location, career potential, etc.
Selecting a Camcorder: price, warranty, zoom, weight, lighting, etc.
Choosing Among Job Applicants: education, experience, personality, etc.
We’ll consider two techniques for these types of problems:
The Multicriteria Scoring Model
The Analytic Hierarchy Process (AHP)
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

52. The Multicriteria Scoring Model
Score (or rate) each alternative on each criterion.
Assign weights the criterion reflecting their relative importance.
For each alternative j, compute a weighted average score as:
wi = weight for criterion i
sij = score for alternative i on criterion j
See file Fig15-41.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

53. The Analytic Hierarchy Process (AHP)
Provides a structured approach for determining the scores and weights in a multicriteria scoring model.
We’ll illustrate AHP using the following example:
A company wants to purchase a new payroll and personnel records information system.
Three systems are being considered (X, Y and Z).
Three criteria are relevant:
Price
User support
Ease of use
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

54. Pairwise Comparisons
The first step in AHP is to create a pairwise comparison matrix for each alternative on each criterion using the following values:
Value Preference
1 Equally Preferred
2 Equally to Moderately Preferred
3 Moderately Preferred
4 Moderately to Strongly Preferred
5 Strongly Preferred
6 Strongly to Very Strongly Preferred
7 Very Strongly Preferred
8 Very Strongly to Extremely eferred
9 Extremely Preferred
Pij = extent to which we prefer alternative i to j on a given criterion.
We assume Pji = 1/Pij
See price comparisons in file Fig15-43.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

55. Normalization & Scoring
To normalize a pairwise comparison matrix,
1) Compute the sum of each column,
2) Divide each entry in the matrix by its column sum.
The score (sj) for each alternative is given by the average of each row in the normalized comparison matrix.
See file Fig15-43.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

56. Consistency
We can check to make sure the decision maker was consistent in making the comparisons.
The consistency measure for alternative i is:
where
Pij = pairwise comparison of alternative i to j
sj = score for alternative j
If the decision maker was perfectly consistent, each Ci should equal to the number of alternatives in the problem.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

57. Consistency (cont’d) Typically, some inconsistency exists.
The inconsistency is not deemed a problem provided the Consistency Ratio (CR) is no more than 10%
where,
RI =
for n =
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

58. Obtaining Remaining Scores & Weights
This process is repeated to obtain scores for the other criterion as well as the criterion weights.
The scores and weights are then used as inputs to a multicriteria scoring model in the usual way.
See file Fig15-43.xls
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.

59. End of Chapter 15
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning.