Researchers at the University of Tsukuba introduced the “dynamic prospect theory,” merging prospect theory and reinforcement learning to explain how past outcomes affect decision-making under uncertainty. Their study revealed that individuals, including monkeys, who encountered unexpectedly large rewards subsequently perceived higher chances of winning in future lotteries, despite objective probabilities being unchanged. This change in probability perception was independent of any shifts in how rewards were valued, suggesting that prior experiences can irrationally distort future assessments of risk.
In an experiment involving 70 participants selecting between two lotteries, the results indicated a consistent pattern: following unexpectedly high outcomes, individuals behaved as if their chances of success had increased. Assistant Professor Hiroshi Yamada noted that this tendency persists even when probabilities are clearly outlined and not reliant on experiential learning. Similar behavior was observed in macaque monkeys, highlighting the parallels between human and primate decision-making processes.
These findings underscore the influence of previous outcomes on perceived probabilities and provide insights into the neural mechanisms underlying reward perception and risk assessment. Understanding these dynamics may enhance our grasp of decision-making in uncertain circumstances, both in humans and monkeys. The research received support from various funding agencies and was published in Science Advances.
