Module 8 Discussion1. Exercise: (Spend 5 to 10 minutes). Identify the direct value nodes in the Influence diagrams on the following pages in Decision Analysis for the Professional (4th edition). p. 315p. 149p. 1892. Define distinctions for the nodes in your influence diagram, starting with the value node and working backward.
Unformatted Attachment Preview
Decision Analysis
for the Professional
Peter McNamee
John Celona
SmartOrg, Inc.
This work is released under the terms of the
Creative Commons Attribution-Noncommercial-No Derivative works 3.0 License.
Decision Analysis for the Professional
Peter McNamee
John Celona
This work in all its entirety is copyrighted and protected by the Creative Commons
Attribution-Non-Commercial-No Derivative Works 3.0 license (http:// You are free to distribute this
work ONLY under the terms of this license. Distribution in classrooms under the
terms of this license is also to be considered as non-commercial usage (no
commercial profit from the distribution is permitted). If you wish to distribute this
book commercially, or create derivative works, you will need to obtain express
written permission from SmartOrg. Please consider buying this book from Amazon
or SmartOrg to encourage the release of such works to the public under similar
Printed in the United States of America.
10 9 8 7 6 5 4 3 2 1
ISBN 0-9710569-0-0
Copyright © 2001-2008 by SmartOrg, Inc.
Electronic Edition, Oct 2008
Publisher: SmartOrg, Inc.
Editors: Mimi Campbell, Bill Roehl, Mary Story
Typography and Layout: Don Creswell and T/D Associates
Cover Design: Rogodino & Associates
Excel is a registered trademark of Microsoft Corporation. Supertree and Sensitivity are registered
trademarks of SmartOrg, Inc.
We would like to dedicate this book to a number of very special people—
Tereza, Patrick, Christina, and Andrew McNamee, and Karen Schwartz,
who endured computer widowhood and orphanhood during the long weeks,
weekends, and months while this book was first being written. Without
their constant support and indulgence, this project would have never come
to fruition.
In the thirteen years since the first edition of this book, the authors have
seen a dramatic evolution in the practice of decision analysis. The number
of companies using decision analysis as an approach to problem solving has
grown rapidly. Our experience during this period has shown that practical
as well as analytical skills are needed for successful implementation of a
decision analysis program.
As a problem-solving approach, decision analysis involves far more than
the use of decision trees as a calculational tool. It is a process of framing a
problem correctly, of dealing effectively with uncertainty, of involving all the
relevant people, and, above all, of communicating clearly. Accordingly, in
addition to the analytical techniques used in decision analysis, this book
presents material that the authors hope will assist the reader in integrating
these techniques into a practical and effective problem-solving process.
The dialog decision process (DDP) and the language of decision quality
have emerged as a powerful tool in the application of decision analysis in a
world of delegated decision making and cross-functional teams. The team
process combines with the analytical clarity of decision analysis to produce
decisions which can be accepted and implemented by the organization.
This edition splits the material into four major sections. The first
section addresses the tools of decision making and decision analysis. The
second section then shows how these tools can be applied in the complex
corporate environment. The third section is new and presents the process
and language that has been developed for dealing with teams and delegated
decision making. The fourth section deals with more advanced topics which
are of interest to the more advanced practitioner.
The book has been rewritten so that it is independent of software. In
several examples and problems, we use hand calculations to teach readers
what the computer programs do. In principle, this text could be read (and
many of the problems done) without using computer software.
However, as everyone knows, it is impossible to do much in decision
analysis without the aid of supporting software. Descriptions of how to use
some software packages such as Supertree are available from the authors.
We hope this book will lead the student to develop an appreciation of the
power, practicality, and satisfying completeness of decision analysis. More
and more, decision analysis and the dialog decision process are becoming
accepted as the best way to address decision problems. Being a
decision facilitator is an exciting and satisfying occupation. This text
is designed to emphasize this. Furthermore, since texts tend to
remain on students’ shelves, we hope this book will be of assistance
long after the course is done.
The text is intended for a short course in decision analysis in
business schools. It could also be part of an analytical methods course.
The general philosophy of the book, however, is more consonant with
extending the course by having the student apply decision analysis to
more complex cases, perhaps based on real data or problems supplied
by local businesses.
Decision analysis is both young enough that its founders are alive
and active in the field and old enough that the literature on the field
has grown large. We have chosen not to write a book bristling with
footnotes. Rather, we have chosen to list in the bibliography several
books in different areas for readers interested in those topics. We ask
our colleagues not to take offense if their names or works are not
explicitly referred to in this book. We gratefully acknowledge their
contribution of accumulated wisdom and knowledge, which has made
decision analysis a useful and powerful management tool.
We especially thank all the people who contributed their useful
comments and constructive suggestions, including Charles Bonini,
Max Henrion, and Myron Tribus. Dr. Bonini kindly allowed us to use
the IJK Products and Hony Pharmaceutics problems that appear in
Chapters 4 and 7, respectively. We are also indebted to Dr. Udi Meirav
for assistance in the discussion of Options and Real Options. We also
thank Yong Tao, who assisted in constructing problems for the book.
We are particularly indebted to Ronald A. Howard and James E.
Matheson for contributions and insights, which appear throughout
this book.
Peter McNamee
Menlo Park, California
John Celona
San Carlos, California
Chapter 1 – Introduction 1
Origins of Decision Analysis 1
Decision Making 2
A Philosophy 2
A Decision Framework 2
A Decision-Making Process 3
A Methodology 3
Dealing with Complex Problems 3
Discovering the Real Problem 4
Keeping the Analysis Manageable 4
Finding the Insights 5
Dealing with Complex Organizations 5
The Team Approach 5
Structured Dialog 6
Decision Quality 6
Advanced Topics 6
Dealing with Uncertainty 7
Dealing with Complex Informational Relationships 7
Obtaining Reliable Information 7
Focus of This Book 7
Problems and Discussion Topics 8
Part I: Decision Making
Chapter 2 – Uncertainty and Probability 13
Probability: A Language of Uncertainty 13
Why Bother with Probabilities? 17
What Are Probabilities? 18
Using Intuition Effectively 19
The “Divide and Conquer” Approach 19
Passing the Clairvoyance Test 19
Assigning the Numbers 20
Using Trees 20
The Data in the Nodes 21
Drawing the Probability Tree 23
A Value Function 24
Analyzing the Tree 26
How Much Detail Is Enough? 29
Certain Equivalent and Expected Value 30
Encoding Probabilities 32
Obtaining the Data 33
Discretizing the Data 34
Using Range Data 35
Summary 36
Problems and Discussion Topics 37
Chapter 3 – Decisions Under Uncertainty 41
What Is a Good Decision? 41
Recasting the Problem As a Decision Problem 43
Decision and Value Nodes 43
Rules for Constructing Influence Diagrams 45
Constructing the Decision Tree 45
Decision or Uncertainty? 48
Building the Tree 49
Decision Criterion 50
The Value of Nonmonetary, Intangible Goods 52
The Value of Future Money 53
The Trade-off Between Certainty and Uncertainty 54
Analyzing the Tree 55
The Value of Perfect Information 58
The Value of Perfect Control 62
Summary 62
Problems and Discussion Topics 63
Chapter 4 – Probabilistic Dependence 75
Dependence and Independence 75
Dependent Probabilities 76
Obtaining the Data 76
Using the Data 76
Dependent Outcomes 79
Obtaining the Data 79
Using the Data 80
Nature’s Tree 80
Indicators and States of Nature 81
An Example from Medicine 82
Bayes’ Rule 85
A Prototype As an Indicator 86
The Value of Imperfect Information 87
The Value of Imperfect Control 92
Common Interpretive Problems with Reordering the Tree 92
Decision-Dependent Probabilities 92
Decision-Dependent Outcomes 93
Summary 94
Problems and Discussion Topics 94
Chapter 5 – Attitudes Toward Risk Taking 101
The Inadequacy of Expected Values 101
Toward a Consistent Risk Attitude 102
Order Rule 103
Equivalence Rule 104
Substitution Rule 104
Choice Rule 104
Probability Rule 105
What Is a Utility Function? 106
Using a Utility Function 108
Value of Information with an Encoded Utility Function 109
An Exponential Utility Function 109
Value of Corporate Risk Tolerance? 116
An Approximation to the Certain Equivalent 117
Encoding a Utility Function 119
Deriving the Existence of a Utility Function 120
Risk-Free Discount Rates 124
Summary 127
Problems and Discussion Topics 127
Part II: Dealing with Complex Problems
Chapter 6 – The Complexity of Real-World Problems 141
A Cyclical Approach 142
Basis Development
Starting the Process 144
Using Decision Hierarchy
Using Strategy Tables 145
Using Influence Diagrams 147
Deterministic Structuring: Modeling the Problem 150
Modeling Profit 151
Sunk Costs 151
Shutting Down the Business 152
Inflation 153
Terminal Value 154
Deterministic Structuring: Sensitivity Analysis 154
Probabilistic Evaluation: Building and Pruning the Tree 161
Basis Appraisal: Obtaining Information from the Tree 166
Conditional Distributions 166
Value of Perfect Information 170
Value of Perfect Control 171
Sensitivity Analysis to Probabilities, Risk Attitude, and Value Trade-offs 172
Policy Matrix 174
Time to Prepare and Present 179
Summary 179
Problems and Discussion Topics 180
Chapter 7 – Typical Corporate Applications of Decision Analysis 189
New Product Introduction 189
Litigation Decision Analysis 195
Bidding Strategies 198
Investment and Investment Rollover Decisions 200
Real Options
R&D Decisions
R&D Portfolio
Corporate Strategies and Business Portfolios 202
Summary 205
Problems and Discussion Topics 205
Part III: Corporate Decision Making
Chapter 8 – A Decision Making Process 223
Decision Projects 224
Should There be a Project? 224
Choosing the Decision Process 225
Structured Decision Process 226
Dialogue Decision Process 227
Framing Dialog 229
Alternatives Dialog 234
Analysis Dialog 236
Decision Dialog 241
DDP and the Decision Analysis Cycle 243
Project Staffing and Timing 243
Presenting Decision Analysis Results 244
Decision Process Capability Building 245
Summary 246
Problems and Discussion Topics 247
Chapter 9 – Decision Quality
Quality in Decision Making 253
People Quality and Content Quality 254
Elements Used in Measuring Decision Quality 255
Appropriate Frame 256
Creative, Doable Alternatives 257
Meaningful, Reliable Information 258
Clear Values and Tradeoffs 259
Logically Correct Reasoning 260
Commitment to Action 261
Decision Quality and the Smart Organization 261
Summary 263
Problems and Discussion Topics 264
Part IV: Advanced Topics
Chapter 10 – Probability Theory 267
Theory Overview 267
Definition of Events 267
Distinctions 268
Algebra of Events 271
Mutually Exclusive and Collectively Exhaustive Events 271
Mutually Exclusive 272
Collectively Exhaustive 272
Joint Events 273
Tree Representation of Events 274
Probability and States of Information 275
Probability Theory 276
Joint, Marginal, and Conditional Probabilities 277
Bayes’ Rule 280
Probabilistic Independence 281
Multiply or Add Probabilities? 281
Events, Variables, and Values 283
Representations of Probabilities for Discrete Values 284
Tabular Form 284
Tree Form 284
Mass Density Plot 284
Cumulative Probability Graph 284
Histogram 286
Mean, Median, Mode, Variance, and Standard Deviation 286
Moments and Cumulants 288
Representations of Probabilities for Continuous Variables 290
Problems and Discussion Topics 292
Chapter 11 – Influence Diagram Theory 303
Theory Overview 303
Elements of Influence Diagrams 304
Uncertainty 304
Decision 305
Influence 306
Determined Uncertainty 310
Value and Determined Value 310
Rules for Constructing Influence Diagrams 311
Procedures for Manipulating Influence Diagrams 312
Turning an Influence Diagram into a Decision Tree 313
Summary 316
Problems and Discussion Topics 316
Chapter 12 – Encoding a Probability Distribution 321
Level of Detail in Encoding Probability Distributions 321
Problems in Encoding Probability Distributions 322
Motivational Biases 322
Cognitive Biases 323
Availability Bias 324
Representativeness Bias 324
Adjustment and Anchoring Bias 325
Implicit Conditioning Bias 325
Probability Encoding Process 326
Stage 1: Motivating 326
Stage 2: Structuring 327
Stage 3: Conditioning 328
Stage 4: Encoding 329
Stage 5: Verification 331
Experiences and Insights from Practice 331
Summary 332
Problems and Discussion Topics 332
Bibliography 335
Index 337
Authors 341
Origins of Decision Analysis ______________________________________
Decision-making is one of the hard things in life. True decision-making
occurs not when you already know exactly what to do, but when you do not
know what to do. When you have to balance conflicting values, sort through
complex situations, and deal with real uncertainty, you have reached the
point of true decision-making. And to make things more difficult, the most
important decisions in corporate or personal life are often those that put you
in situations where you least know what to do.
Decision science evolved to cope with this problem of what to do. While
its roots go back to the time of Bernoulli in the early 1700s, it remained an
almost purely academic subject until recently, apparently because there was
no satisfactory way to deal with the complexity of real life. However, just after
World War II, the fields of systems analysis and operations research began
to develop. With the help of computers, it became possible to analyze
problems of great complexity.
Out of these two disciplines grew decision analysis: the application of
decision science to real-world problems through the use of systems analysis
and operations research. Decision analysis is a normative discipline, which
means it describes how people should logically make decisions. Specifically,
it corresponds to how (most) people make decisions in simple situations and
shows how this behavior should logically be extended to more complex
This book is divided into four parts, as described in the following pages.
The first part develops the tools of decision making. The second part
describes how these tools can be applied to complex problems. The third part
presents a process for using decision analysis in today’s corporate setting.
Some specialized topics are dealt with in the fourth part.
Decision Making ________________________________________________
Imagine a decision-maker struggling with a difficult decision problem. The
decision analysis approach provides a normative approach that can support
the decision-maker.
Decision analysis functions at four different levels—as a philosophy, as
a decision framework, as a decision-making process, and as a decisionmaking methodology—and each level focuses on different aspects of the
problem of making decisions.
Part I of this book lays the foundation for all four levels.
A Philosophy
As a philosophy, decision analysis describes a rational, consistent way to
make decisions. As such, it provides decision-makers with two basic, invaluable insights.
The first insight is that uncertainty is a consequence of our incomplete
knowledge of the world. In some cases, uncertainty can be partially or
completely resolved before decisions are made and resources committed.
However, in many important cases, complete information is simply not
available or is too expensive (in time, money, or other resources) to obtain.
Although this insight may appear obvious, we are all familiar with
instances in the business world and in personal life in which people seem to
deny the existence of uncertainty—except perhaps as something to be
eliminated before action is taken. For example, decision-makers demand
certainty in proposals brought before them. Twenty-year projections are used
to justify investments without any consideration of uncertainty. Time and
effort are spent to resolve uncertainties irrelevant to the decision at hand. And
this list could, of course, be greatly extended.
The second basic insight is that there is a distinction between good
decisions and good outcomes. Good outcomes are what we desire, whereas
good decisions are what we can do to maximize the likelihood of having good
outcomes. Given the unavoidable uncertainty in the world, a good decision
must sometimes have a bad outcome. It is no more logical to punish the maker
of a good decision for a bad outcome than it is to reward the maker of a bad
decision for a good outcome. (Many types of routine decisions have little
uncertainty about outcomes; thus, in these cases, it is not unreasonable to
associate bad outcomes with bad decisions.)
This insight, too, may seem obvious. Yet how often have we seen corporate
“witch hunts” for someone to blame or punish for unfortunate corporate
A Decision Framework
As a framework for decision-making, decision analysis provides concepts and
language to help the decision-maker. By using decision analysis, the decisionmaker is aware of the adequacy or inadequacy of the decision basis: the set
of knowledge (including uncertainty), alternatives, and values brought to the
decision. There is also a clear distinction between decision factors (factors
completely under the decision-maker’s control) and chance factors (uncertain factors completely outside the decision-maker’s control). Moreover, the
decision-maker is aware of the biases that exist in even the most qualitative
treatments of uncertainty. He or she knows these biases exist because people
are not well trained in dealing with uncertainty and because they are
generally overconfident in describing how well they know things.
A Decision-Making Process
As a decision-making process, decision analysis provides a step-by-step
procedure that has proved practical in tackling even the most complex
problems in an efficient and orderly way. The decision analysis cycle provides
an iterative approach that keeps the focus on the decision and that enables
the decision facilitator* to efficiently compare the decision alternatives.
Modeling, both deterministic and probabilistic, reduces the problem to
manageably sized pieces and allows intuition to function most effectively.
Knowledge of the biases in probability estimation enables the decisionmaker or facilitator to take corrective action.
A Methodology
As a methodology, decision analysis provides a number of specific tools that
are sometimes indispensable in analyzing a decision problem. These tools
include procedures for eliciting and constructing influence diagrams, probability trees, and decision trees; procedures for encoding probability functions and utility curves; and a methodology for evaluating these trees and
obtaining information useful to further refine the analysis.
It is a common mistake to confuse decision analysis with constructing
and manipulating influence diagrams and decision trees. The real contribution and challenge of decision analysis occur at the much more general level
of defining the problem and identi …
Purchase answer to see full

Are you having trouble with the above assignment or one similar?

We offers 100% original papers that are written from scratch.We also have a team of editors who check each paper for plagiarism before it is sent to you.
!-- End of Footer -->