CONTACT: Stanford University News Service (650) 723-2558
Amos Tversky, leading decision researcher, dies at 59
STANFORD -- Stanford psychologist Amos Tversky, one of the world's leading experts in judgment and human decision making, died Sunday, June 2, of metastatic melanoma at his Stanford home. He was 59.
Tversky, a cognitive psychologist who was a dominant figure in decision research and a leading psychological theorist, seriously challenged economic theory by showing that people frequently do not behave rationally to maximize their welfare. His work on the limits of human rationality also had a major impact on philosophy, statistics, political science, law and medicine.
His work had a great impact on economics, said Kenneth Arrow, professor emeritus of economics, because he tested hypotheses of rationality that are central to predicting how economies behave. "The hypothesis of rational behavior has been central to economics, though always held with some discomfort," Arrow said from All Souls College, where he is on sabbatical. "Previous criticism of economic postulates by psychologists had always been brushed off by economists, who argued, with some justice, that the psychologists did not understand the hypotheses they criticized. No such defense was possible against Amos' work," he said.
Tversky's early work on judgment began in Israel with another Israeli-born psychologist, Daniel Kahneman, now of Princeton. They detailed 11 "cognitive illusions," or biasing characteristics of human judgment, and proposed systematic explanations for them. Tversky's later work on decision making, some of it with Kahneman, showed how people make choices under conditions of uncertainty. His experiments and theories are so precise and broadly applicable, said Robyn Dawes, a professor of social and decision sciences at Carnegie Mellon University, that teachers can make up their own experiments, try them in any class and be reasonably certain the students will demonstrate the judgment pattern that Tversky has specified.
Tversky once explained his work by saying that "people use mental approximations to understand an uncertain world. As a result, we make certain types of errors in judgment." He admitted that even after studying the nature of these judgmental errors for years, he sometimes caught himself making them.
When he won a five-year MacArthur Foundation fellowship in 1984, Tversky said with typical modesty that much of what he had studied was already known to "advertisers and used car salesmen." His theoretical modeling and carefully designed experiments, however, elucidated the basis for such phenomena as consumers getting upset if a store charged a "surcharge" for using a credit card but being pleased if a store offered a "discount" for paying with cash.
"He certainly changed my life, applying just the concept that people's reasoning is imperfect, susceptible to error and amenable to corrective procedures," said Donald Redelmeier, a physician at the University of Toronto who did research with Tversky. Tversky's and Kahneman's work on "framing" the idea that small differences in how data are presented to people have a substantial effect on their decisions has influenced the way doctors view informed consent from patients for medical procedures, according to Redelmeier and Barbara McNeil, the Ridley Watts Professor of Health Care Policy at Harvard. McNeil worked with Tversky on what she refers to as the 1 percent99 percent problem. One and 99 add up to 100, so that if someone is told he or she has a 1 percent chance of dying during a given medical procedure, that person also has a 99 percent chance of living. Their studies showed that people will be more optimistic or pessimistic about the procedure, depending upon which way the information was stated to them. "At a time when medical technology has advanced and patients are being asked to make more decisions about medical options, this is even more important than it was in 1980" when the research was done, McNeil said.
Scientists who worked with Tversky said he was a perfectionist who held high standards for himself and others, but was also friendly, modest and helpful. "You were happy being in his presence," said Persi Diaconis, a professor of mathematics at Harvard. "There was a light shining out of him."
Stanford psychology Professor Lee Ross said that Tversky "was modest, unassuming and never a prima donna. That had the good effect of making it impossible for everyone else" to demand special treatment or favors.
While they were personal friends, Ross added that his sense of loss is "greater than can be explained by personal relationships. Everyone, not just people at Stanford, feel the field has been diminished, and that intellectual life has been diminished" as a result of Tversky's death.
"People gave him respect bordering on awe. For many people, it was just plain awe," Kahneman said. Tversky continued to work at home until the last days of his life with good humor,and a calmness that was inspiring, collaborators said.
Ellen Markman, chair of the psychology department, said the department will have to learn to make its decisions in new ways. "I don't think there has been a serious issue in the department since he's been here where people haven't sought his advice and counsel on how to proceed. There was a sense of calm, fairness and thoughtfulness about him that made him a tremendously valued colleague, independent of his work. People would always say, 'Have you talked to Amos yet?' " They also could always count on his "almost boyish sense of humor," Markman said.
President Gerhard Casper said that Tversky, the inaugural Davis-Brack Professor of Behavioral Sciences at Stanford, "came as close to being the ideal of a university faculty member as any colleague I have known in my almost four decades in higher education. He was an outstanding, original, challenging scholar and teacher. He maintained the highest standards of professional ethics. He always spoke his mind and responded to disagreements in the same rational way in which he had argued to begin with. His dedication to Stanford and its institutions of faculty governance was exemplary until the very end. His death is a terrible loss, not only for Barbara and the children, but also for the university."
Tversky's 1974 Science article with Kahneman on cognitive illusions triggered a "cascade of related research," Science News wrote in a 1994 article tracing the recent history of research on reasoning. Decision theorists in economics, business, philosophy and medicine as well as psychologists cited their work.
Tversky was the more mathematically minded of the two, colleagues say, but he also was an astute observer of how people made decisions and was good at explaining his ideas to others.
"One of his most beautiful pieces in recent years started out from observations he made at a Stanford faculty meeting," Kahneman said. Tversky observed that the faculty wanted to make appointment offers to two people for two vacancies but had decided to hold up the offer to one until after they had made the offer to the first, whom they wanted more. When Tversky pointed out that there was no reason to hold off on the offer to the second, the committee made both offers and Tversky went on to demonstrate similar behavior in a number of laboratory experiments.
Asked in 1985 about his ability to provide practical examples of all his theories and mathematical models, Tversky said that "growing up in a country that's fighting for survival, you're perhaps more likely to think simultaneously about applied and theoretical problems."
Tversky was born in Haifa, Israel, on March 16, 1937, to parents who had emigrated from Poland and Russia. His father, Yosef, was a veterinarian and his mother, Genia, was a member of the Knesset from its establishment in 1948 until her death in 1964. Tversky was an officer in the paratroopers, an elite unit, eventually rising to captain and serving in three wars. He became a "legend" within the unit, Kahneman said, after he was awarded Israel's highest honor for personal bravery during a 1956 border skirmish.
"On maneuvers, a soldier went forward with an explosive charge to blow up some barbed wire and froze, literally lying down on top of the explosive after lighting the fuse," Kahneman said. "Amos was a few meters behind him. He got up and in a few seconds managed to get to that man, raise him and throw him to safety while his commanding officer was yelling at him not to do it. He was wounded and the other person emerged without a scratch. He was not seriously injured but he had metal in him all his life."
Tversky earned a bachelor's degree from Hebrew University in 1961 and his doctorate in 1965 from the University of Michigan, where he met and married his wife, Barbara, a fellow student in cognitive psychology. Barbara Tversky is also a professor of cognitive psychology at Stanford.
After holding several teaching positions at Michigan, Harvard and Hebrew University, Amos Tversky came to Stanford in 1970 as a fellow at the Center for Advanced Study in the Behavioral Sciences. He joined the Stanford faculty in 1978 and won the American Psychological Association's award for distinguished scientific contribution in 1982 and the MacArthur and Guggenheim fellowships in 1984.
He was elected to the American Academy of Arts and Sciences in 1980 and as a foreign associate of the National Academy of Sciences in 1985. He also was awarded honorary doctorates by the University of Chicago, Yale University, the University of Goteborg (Sweden) and the State University of New York at Buffalo. He served on Stanford's Faculty Senate from 1990 until his death and was a member of the Academic Council's advisory board to the president and provost.
His ties to Israel were also an important part of his life, Barbara Tversky said. He was affiliated with Tel Aviv University as a fellow of the Sackler Institute of Advanced Studies and a visiting professor of economics and psychology, directing student research there as well as at Stanford. "He was a great storyteller and kept audiences on the edges of their chairs in English and Hebrew," she said.
Tversky and Kahneman conducted experiments demonstrating that people don't really know how to routinely compute probabilities, but "they use heuristics, such as 'representativeness,' based on automatic assessments of similarity, something they are good at, and 'availability,' based on assessments of memory retrievability," Barbara Tversky said. In addition, they showed that people are tuned to see patterns in random performance, such as stock market fluctuations, coin tosses or baseball hitting streaks. Their choices and judgments are strongly influenced by the language or context in which a choice is framed.
For example, in one experiment Tversky showed that if someone loses a $10 theater ticket on the way to the theater, he or she is not likely to buy another, because it will seem like paying $20 for a $10 ticket. They are very likely to buy the ticket, however, if they lose $10 in cash on the way to the theater instead.
In his last decade, Tversky turned his attention to the implications of cognitive processes for conflict, Arrow said. The two helped found the Stanford Center on Conflict and Negotiation.
More recently, Tversky and several colleagues have shown that people will try to cheat on their medical exams to convince themselves they are in better health than they are, and that doctors and politicians are subject to cognitive biases in making choices between three or more alternatives. In one study, he challenged the long-standing belief of some doctors and arthritis sufferers that weather affects joint pain.
Besides his wife, Barbara, Tversky is survived by two sons, Oren of San Francisco and Tal of Stanford, and a daughter, Dona of Stanford, and by his sister, Ruth Ariel of Jerusalem.
Memorial services have been scheduled for 3 p.m. Wednesday, June 5, at Hills of Eternity in Colma and a university memorial service will be held sometime in the fall quarter, Markman said. In lieu of flowers, the Tversky family suggests contributions to favorite charities.
Download this release and its related files.
The release is provided in Adobe Acrobat format. Any images shown in the release are provided at publishing quality. Additional images also may be provided. Complete credit and caption information is included.