Academic journal article Journal of Risk and Insurance

Advances in Decision Analysis

Academic journal article Journal of Risk and Insurance

Advances in Decision Analysis

Article excerpt

Advances in Decision Analysis, edited by Nadine Meskens and Marc Roubens, 1999, Norwell, Mass.: Kluwer Academic Publishers

Reviewers: Gordon Johnson, Ph.D., California State University, Northridge, and Joseph J. Launie, Ph.D., CPCU, California State University, Northridge

Advances in Decision Analysis, edited by Nadine Meskens of FUCAM University in Mons, Belgium, and Marc Roubens of the Institute of Liege in Liege, Belgium, contains ten papers presented at the International Conference on Methods and Applications of Multiple Criteria Decision Making (MCDM) held in Mons in May 1997. The focus is on the most recent advances in this field, and the reader should be familiar with MCDM before acquiring this book, although the bibliographies offer a wealth of background reading. For those unfamiliar with MCDM, a recent example (not in the book) might be these two criteria: "Eliminate terrorism" and "Promote good will toward the United States." Bombing villages assists in the first goal but may hurt the second goal. Readers of this journal might also consider these two criteria: "Minimize insurance claims costs" versus "Avoid bad faith litigation."

The first paper, by Herve Raynaud, includes a survey of the conference participants using the Delphi Method. An important result is the opinion that clients need to be more involved with the researcher to increase the likelihood that MCDM methods will be implemented. In particular, researchers need to understand "how people really make decisions in their minds, which is fairly well published if unknown from most decision scientists" (p. 5). Raynaud also shows a proof that the iteration of a strongly monotonic choice function is not a strongly monotonic ranking function. In less mathematical terms, "Methods that have a good outlook on some sets of data may yield weird consequences for more tricky cases" (p. 11). More generally, different people will arrive at different decisions when the criteria are contradictory.

The editors group several papers into a section on the methodological aspects of decision theory. Denis Bouyssou and Marc Pirlot contribute to the theory of MCDM, with emphasis on the failure of human clients to be consistent in actual decision making. With several attributes, people are not always transitive when comparing alternatives.

Jean-Francois Laslier does a survey of game theory methods, using both ordinal and cardinal information. In general, we use decision theory when we face states of nature (earthquakes and hurricanes, for example), while game theory involves another player who anticipates our decisions (for example, chess, competing insurance companies).

Oleg Larichev's paper is user-friendly since it starts with a practical example, the decision regarding where to live. Criteria include cost, size, noise, and commuting distance. Minimizing cost would likely lead to a small house in a noisy location far from work. The author uses utility theory in his paper, along with a decision support system. He describes his approach, Verbal Decision Analysis, and comments on the unrealistic expectations of Von Neumann and Morgenstern in 1947 when mathematicians believed that decision making could be modeled as a physical science. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.