Stanford News

4/1/97

CONTACT: Stanford University News Service (415) 723-2558


Symposium on decision-making to focus on the legacy of Amos Tversky

Rarely can an academic conference promise to draw people from as many backgrounds as that planned to honor the late Amos Tversky on Saturday, April 12. Tversky, whose death at age 59 last year brought tributes from around the world, was a professor of psychology at Stanford who some considered a candidate for the Nobel Memorial Prize in economic sciences.

Along with Princeton's Daniel Kahneman, Tversky showed that the classical picture of people as rational beings who act in their best interest didn't correspond to everyday behavior. They showed that even the most experienced and educated adults repeatedly make errors in judgment, not randomly but in ways that the two psychologists could predict and categorize.

Using Tversky's carefully crafted examples, researchers in many fields went on to show that systematic errors were made by juries and judges in courtrooms, by patients and their doctors in hospitals, by business people hiring employees or investing capital, by arms control negotiators and divorcing couples trying to negotiate agreements. The two have caused philosophers to rethink what it means to be happy, statisticians to rethink probability, economists and engineers to rethink how they analyze the costs and benefits, for example, of siting a hazardous waste facility or new factory.

Hoping to capture some of the remarkable flexibility, practicality and depth of Tversky's work, planners of the symposium have selected prominent non-Stanford academics in several fields as speakers. The public event begins with coffee at 9 a.m. in Room 290 of the Law School and a welcome from Stanford President Gerhard Casper at 9:45 a.m.

Casper will be followed at 10 a.m. by Richard Nisbett, a social psychologist from the University of Michigan and at 11 a.m. by Cass Sunstein from the University of Chicago School of Law. After a lunch break, Persi Diaconis of Harvard's Department of Mathematics will speak at 2 p.m. Two Harvard economists, David Laibson of the Economics Department and Richard Zeckhauser of the John F. Kennedy School of Government, will discuss Tversky's impact on economics at 3 p.m. followed by economist Thomas Schelling of the University of Maryland at 4:15 p.m.

Symposium planners say they can't predict what will be said but point to some interesting connections between the visiting speakers and Tversky. Diaconis, a statistician (who is also "one of the world's best card magicians," according to Stanford statistician Bradley Efron), frequently stayed at Tversky's house on his West Coast visits where the two discussed the problem of "thinking too much."

"Amos seemed to believe ­ and he was usually right ­ that one way to make bad decisions is to spend too much time thinking about them," Efron said. "Persi has done some probabilistic research on that topic."

Nisbett is likely to talk about Tversky's role as the "standard bearer" for psychology, said Stanford Professor Lee Ross, who is a co-author with Nisbett. "To anyone who criticizes psychology for not saying enough about everyday human behavior that is true, new and useful," Ross said, "psychologists always can respond, 'What about Amos Tversky?' " Nisbett will trace Tversky's influence.

Ellen Markman, chair of the psychology department, said she believed that Tversky had a "profound, serious impact on a huge range of disciplines" partly because of his elegant theories and partly because he was able to explain them with compelling examples. "Most of us immediately recognized his examples [of inferential folly] as something we would do and yet also recognized that they violated our concept of standard decision-making."

Ross points to one of Kahneman's and Tversky's famous illustrations of the "representative heuristic": A woman named Linda is described to people in such a way that her interests, background and views correspond to their image of a feminist and poorly to their image of a bank teller. When Tversky asked people to assess the likelihood that Linda was a bank teller, most said the possibility was less than the odds that she was both a bank teller and a feminist, defying laws of probability. In their book on human inference, Nisbett and Ross drew heavily on Kahneman's and Tversky's work, claiming that it showed how people function as intuitive psychologists in everyday life and how their inferential shortcomings and triumphs are cut from the same cloth.

Economists at the symposium are likely to touch on ways Tversky's and Kahneman's work on "prospect theory" forced their field to focus more on actual, rather than ideal, human choice. Critiquing the role of rational choice in economic "utility" theory, Tversky and Kahneman showed, among other things, that "the threat of a loss has a greater impact on a decision than the possibility of an equivalent gain." Their concept of "framing" revealed that the way a choice is stated influences people's preferences. For example, if the statistical risk of undergoing a certain medical treatment is stated in terms of percentage of lives lost, patients and their doctors are less likely to choose the treatment than if the same statistical risk is stated in terms of lives saved.

For the legal profession, the implications are equally if not more numerous than for economics and medicine. Tversky's and Kahneman's work showing the different shape of the gain and loss curves in people's mental accounting can explain, for example, why plaintiffs frequently settle out of court when they could win and why defendants continue to litigate when "rational" outsiders conclude they should settle.

Philosophers of the mind are as interested in the work as philosophers who focus on how to live happy, satisfying lives, Ross said. Schelling, for instance, has written about the problem of making wise choices, given the varying needs of the individual delineated in Tversky's and Kahneman's work.

The symposium was organized by Professors Markman, Ross, Efron and Kenneth Arrow. It is sponsored by the Psychology Department in cooperation with the Dean's Office of the School of Humanities and Sciences. For more information, contact Joni Bonham at 725-2449.

-30-

By Kathleen O'Toole