Preface

This textbook covers important topics on decision making, presents tools for helping engineers make better decisions, and provides examples to illustrate the concepts and techniques. Students and engineers who study this material and apply these concepts and techniques should become better decision-makers.

Like the products and systems that engineers design, this textbook began as an idea for meeting a need and went through many iterations and revisions over time. In this case, the initial discussions about engineering decision making involved my colleague Linda Schmidt, an expert on design methodologies and design education. She and I discussed how engineers in product development organizations shared information and made decisions, and we decided to begin studying this activity as a system. Then, with our colleague Peter Sandborn, we were awarded a grant from the National Science Foundation to study how firms used information about environmental impacts in product development decision making. After studying multiple firms and publishing our results, the next step was to develop a course in which we could share our insights about decision making with others. We jointly developed a course outline, and in the Spring, 2004, semester I taught the course for the first time. Although a traditional decision analysis textbook was used, the course included topics beyond its scope, so I created course notes and expanded them every time I taught the course.

In the meantime, our research continued, and I developed three perspectives on decision making. This led me to reorganize the course (and the course notes) around these three perspectives, which provide a new way to consider engineering decision making. In addition, I included various topics on risk management, a type of decision-making process. These changes also emphasized the challenges of using a traditional decision analysis textbook that was organized in a completely different way. The organization of this course was not increasing mathematical difficulty but increasing conceptual complexity, and existing texts on decision analysis were inappropriate. The first draft of this textbook was my reorganized set of course notes, which I then divided and rearranged again to form distinct chapters.

This text discusses three perspectives on decision making: (1) the problem-solving perspective, (2) the decision-making process perspective, and (3) the decision-making system perspective. The text introduces these perspectives in Chapter 1 and covers them in sequence as the following paragraphs describe. Techniques for modeling and managing risk are included throughout the text where appropriate within this framework.

Chapters 2-6 consider the components and structure of decisions, which is the problem-solving perspective. Chapter 2 reviews some fundamental topics, including the context of a decision situation, fundamental objectives and means objectives, influence diagrams, rationality, choice strategies, dominance, "framing" a decision situation, risk acceptance criteria, and types of measurement scales. Understanding these important fundamental concepts can help one improve decision making.

After Chapter 2 are two chapters about decisions without uncertainty (Chapters 3 and 4) and then two chapters about decisions with uncertainty (Chapters 5 and 6).

Chapter 3 covers multicriteria decision making, which is a traditional topic in decision analysis and an important skill that is the foundation of decision making. This chapter covers multiple techniques: the Pugh matrix, a version of the analytic hierarchy process (AHP), multiattribute utility theory (MAUT), and conjoint analysis. It also discusses the usefulness of the "Value of a Statistical Life" and the differences between compensating and non-compensating solutions.

Chapter 4 reviews techniques for group decision making. This material follows multicriteria decision making (Chapter 3) because the decisions do not have uncertainty. The chapter covers two primary techniques: ranking (including the Kemeny-Young method) and scoring, including the majority judgment technique. It also discusses the implications of Arrow's Impossibility Theorem.

Chapter 5 introduces decisions with uncertainty (risky decisions) and includes traditional material on decision trees, risk aversion, and expected utility. It discusses different types of uncertainties and subjective probabilities. It also defines different types of robustness measures and presents uncertainty propagation techniques, including sensitivity analysis, the method of moments, and Monte Carlo simulation. (Other approaches for making decisions in the presence of uncertainty are discussed in Chapter 7.)

Chapter 6 then discusses game theory. This chapter is placed after Chapter 5 because the existence of another decision-maker introduces uncertainty, but this uncertainty is quite different from uncertainties that can be represented as random variables (which are discussed in Chapter 5). The chapter discusses two-player simultaneous, zero-sum games (and finding optimal mixed strategies), two-player, simultaneous, mixed-motive games, and two-player Stackelberg games. Game theory is also useful for considering risks due to intelligent adversaries.

Chapters 7-9 discuss the decision-making process perspective: how people make decisions through decision-making and risk management processes. Because different situations require different types of decision-making processes, Chapter 7 begins this part of the text by reviewing many types of useful decision-making processes, the important roles of heuristics and search in decision making, and the composite nature of decisions. It also discusses the secretary problem, a special case in which the decision-making process can be optimized. This chapter also describes product development as a type of decision-making process.

Chapter 8 discusses the value of information, a traditional topic. The decision to gather more information is usually a decision with uncertainty, but it is included in this part of the text because it is a decision about what to do in the decision-making process. This chapter describes how to calculate the expected value of perfect information and the expected value of imperfect information and discusses more generally how to use experimental information to improve decision making.

Chapter 9 explicitly covers the process of risk management, which includes the decision of which risk mitigation activity (or activities) should be performed. The risk mitigation decision is another decision with uncertainty, but this material is included in this part of the text because risk management is a type of decision-making process, and Chapter 9 describes different risk management processes and risk communication, an important part of any decision-making process. Because choosing an appropriate decision-making process reduces the risk of making a poor decision, Chapter 9 discusses poor decisions and how to learn from those that do occur. Finally, because some risk management processes emphasize continuously monitoring an activity and intervening when needed to reduce risk, these processes can be viewed as a control system. This means that the risk management function is also a decision-making system, which is the third perspective.

In the last part of the text, Chapters 10 and 11 describe the decision-making characteristics of organizations and how to improve those decision-making systems, which is relevant to the decision-making system perspective. This perspective considers the flow of information between different decision-makers who have different roles. It views an organization as a dynamic system that makes decisions using decision-making processes, but the quality of the decisions that emerge depends upon not only the decision-making processes but also the culture and the patterns of behavior.

Chapter 10 describes the characteristics and structure of decision-making systems, including different roles and mechanisms of organizational influence. It also describes product development organizations as decision-making systems.

Chapter 11 discusses improving decision-making systems. Because decision- making systems are complex and involve human actors, the usefulness of quantitative techniques is limited. Qualitative approaches can represent more interesting phenomena. The chapter begins with different techniques for modeling decision-making systems, including rich pictures, swimlanes, root definitions, and conceptual models. The chapter also presents an improvement strategy that exploits the insights that these types of models provide.

Thus, the text begins with the aspect of decision making that is, conceptually, the simplest: given a set of alternatives and a decision-maker's preferences, which one should be selected? The text then looks "behind the scenes," so to speak, to describe the processes used to generate and evaluate the alternatives. Finally, the text "steps back" to look at the organization that, to achieve its goals, performs these processes and makes these decisions.

In the first part of the text (Chapters 2-6), the critical skill is choosing the right alternative. In the second part (Chapters 7-9), the critical skill is executing the right decision-making process. In the third part (Chapters 10 and 11), the critical skill is improving the decision-making system.

The content and organization of this text reflect both a pragmatic attitude about improving decision making and the scholarly concern with studying, organizing, and formalizing this activity. This approach has been formed by studying how engineers and...