Suchergebnisse

This paper presents the current state of evolution of a language for teaching and practicing decision analysis that may avoid confusion of students, clients, and ourselves. Many of the terms currently used are inaccurate, arcane, or unnecessary. Restricting decision language to terms that are accurate, familiar, and fundamental contributes to clarity of thought and understanding. To illustrate the type of changes advocated, I propose replacing dependence with relevance, expectation with e-value, utility with u-value, and eschewing subjective probability, confidence, uncertainty about probability, any distinction between risk and uncertainty, and states of nature. I show how to incorporate the new terminology in teaching and practice.

These remarks respond to the comments offered by four colleagues on my paper "Speaking of Decisions: Precise Decision Language" (Howard 2004). The response gives me the opportunity to clarify and expand on the paper and to note different views. In particular, I further discuss random variable, distinction, expected value, option, utility as a measure of happiness or satisfaction, subjective probability, the necessity of a value measure for computing the value of clairvoyance, decision analysis as a normative discipline, and the inability of the clairvoyant to provide probabilities. I conclude with a concept map for the language and a table showing where the terms are most useful in conversations with students and clients.

The issue of how to think about "uncertainty about probability" is framed and analyzed from the viewpoint of a decision analyst. The failure of nuclear power plants is used as an example. The key idea is to think of probability as describing a state of information on an uncertain event, and to pose the issue of uncertainty in this quantity as uncertainty about a number that would be definitive: it has the property that you would assign it as the probability if you knew it. Logical consistency requires that the probability to assign to a single occurrence in the absence of further information be the mean of the distribution of this definitive number, not the median as is sometimes suggested. Any decision that must be made without the benefit of further information must also be made using the mean of the definitive number's distribution. With this formulation, we find further that the probability of r occurrences in n exchangeable trials will depend on the first n moments of the definitive number's distribution. In making decisions, the expected value of clairvoyance on the occurrence of the event must be at least as great as that on the definitive number. If one of the events in question occurs, then the increase in probability of another such event is readily computed. This means, in terms of coin tossing, that unless one is absolutely sure of the fairness of a coin, seeing a head must increase the probability of heads, in distinction to usual thought. A numerical example for nuclear power shows that the failure of one plant of a group with a low probability of failure can significantly increase the probability that must be assigned to failure of a second plant in the group.

We present an analogy between joint cumulative probability distributions and a class of multiattribute utility functions, which we call attribute dominance utility functions. Attribute dominance utility functions permit assessing multiattribute utility functions using common techniques of joint probability assessment such as marginal-conditional assessments and the method of copulas. By itself, this class of utility functions appears in many cases of decision analysis practice. Furthermore, we show that many functional forms of multiattribute utility function can be decomposed into attribute dominance utility functions that are easier to elicit. We introduce the notion of utility inference analogous to Bayes' rule for probability inference and provide a graphic representation of attribute dominance utility functions, which we call utility diagrams.

We often make small ethical compromises for "good" reasons: We lie to a customer because our boss asked us to. We exaggerate our accomplishments on our résumé to get an interview. Temptation blindsides us. And we make snap decisions we regret. Minor ethical lapses can seem harmless, but they instill in us a hard-to-break habit of distorted thinking. Rationalizations drown out our inner voice, and we make up the rules as we go. We lose control of our decisions, fall victim to the temptations and pressures of our situations, taint our characters, and sour business and personal relationships. In Ethics for the Real World, Ronald Howard and Clinton Korver explain how to master the art of ethical decision making by: Identifying potential compromises in your own life Applying distinctions to clarify your ethical thinking Committing in advance to ethical principles Generating creative alternatives to resolve dilemmas Packed with real-life examples, this book gives you practical advice to respond skillfully to life's inevitable ethical challenges. Not only can you make right decisions, you can acquire new habits that will realize the best in yourself and transform your relationships

Finding the best national strategy to prevent or delay a country from acquiring nuclear weapons continues to be a critical issue for U.S. policy makers. In this paper, we build on previous work to develop a model that addresses this question. This model identifies the strategy that minimizes the disutility of the overall cost of the strategy and the cost of the consequences resulting from the strategy. We illustrate the insights that the model provides with a case study of Iran's nuclear weapons program.

Influence diagrams were first used in 1973 as a way to model political conflicts in the Persian Gulf and measure the value of information collected by the Defense Intelligence Agency. The number of scenarios for events in the region was too large to be represented as a conventional decision tree model. Influence diagrams were initially conceived as a way to create smaller, coalesced decision trees that required fewer probability assessments. However we found that they also facilitated communication between analysts, experts, and policy makers. Influence diagrams later became the basis for new ways of solving decision models.

Coastal ecosystems are under pressure from a vast array of anthropogenic stressors, including development and climate change, resulting in significant habitat losses globally Conservation policies are often implemented with the intent of reducing habitat loss. However, losses already incurred will require restoration if ecosystem functions and services are to be recovered. The United States has a long history of wetland loss and recognizes that averting loss requires a multi-pronged approach including mitigation for regulated activities and non-mitigation (voluntary herein) restoration. The 1989 "No Net Loss" (NNL) policy stated the Federal government's intent that losses of wetlands would be offset by at least as many gains of wetlands. However, coastal wetlands losses result from both regulated and non-regulated activities. We examined the effectiveness of Federally funded, voluntary restoration efforts in helping avert losses of coastal wetlands by assessing: (1) What are the current and past trends in coastal wetland change in the U.S.?; and (2) How much and where are voluntary restoration efforts occurring? First, we calculated palustrine and estuarine wetland change in U.S. coastal shoreline counties using data from NOAA's Coastal Change Analysis Program, which integrates both types of potential losses and gains. We then synthesized available data on Federally funded, voluntary restoration of coastal wetlands. We found that from 1996 to 2010, the U.S. lost 139,552 acres (~565 km(2)) of estuarine wetlands (2.5% of 1996 area) and 336,922 acres (~1,363 km(2)) of palustrine wetlands (1.4%). From 2006 to 2015, restoration of 145,442 acres (~589 km(2)) of estuarine wetlands and 154,772 acres (~626 km(2)) of palustrine wetlands occurred. Further, wetland losses and restoration were not always geographically aligned, resulting in local and regional "winners" and "losers." While these restoration efforts have been considerable, restoration and mitigation collectively have not been able to keep pace with wetland ...