Brain activation can predict strategies people use to make risky decisions

May 27 2009

Watching people's brains in real time as they handle a set of decision-making problems can reveal how different each person's strategy can be, according to neuroscientists at the Duke University Medical Center.

"People in our study, like the population at large, differed in the strategies they use to make economic decisions," said Scott Huettel, Ph.D., co-director of the Duke Center for Neuroeconomic Studies and senior author of the study published in Neuron online on May 27. "What sort of strategy people tended to use could be predicted, surprisingly, by how their brain responded to rewards: if there were large responses to monetary reward in a brain area called the ventral striatum, that person tended to simplify decision problems to only consider winning or losing."

"Using studies like this to build a better understanding of how our brains represent our decision strategies may someday allow researchers to use someone's personal traits – say, an adolescent with high impulsivity, but ongoing depression – to predict the decisions that he or she will make," Huettel said. "This could lead to many real-world benefits: designing more effective interventions or creating more useful educational material."

Twenty-three participants were scanned using functional magnetic resonance imaging (fMRI) which reveals real-time changes in brain function as they evaluated complex multi-outcome lotteries. The problems involved real monetary gains and losses – something akin to the universe of factors that someone buying a car must consider, for example. When faced with each problem, the participant chose among ways to improve their lottery chances, such as reducing the worst possible loss.

This study was a collaboration between neuroscientists and decision making experts at the Fuqua School of Business at Duke.

"Our goal was to come up with a risky decision task that would both discriminate between alternative models of choice and represent something that often happens when people allocate scarce resources to make a risky choice more attractive," said John W. Payne, Ph.D., the Joseph J. Ruvane, Jr. Professor of Business. "It was also nice that the task is complex enough to relate to 'real-world' decisions but simple enough to be studied using functional MRI."

The study showed that the brain regions classically associated with "rational" processing, notably the lateral prefrontal cortex, were most active when subjects used a simplifying strategy inconsistent with traditional rational-choice models.

"This result suggests that it was the type of computation that the participants were doing at any given time that activates a brain region, not whether the thought process is rational or irrational," said lead author Vinod Venkatraman, from the Department of Psychology and Neuroscience at Duke.

The finding "argues strongly against the commonly held notion that there are 'rational' and 'irrational' parts of the brain,'" Huettel added.

The study also showed that the medial prefrontal region of the brain shapes moment-to-moment changes in the strategies people use to make decisions. "We all make some decisions opposite to our usual tendencies. When we do so, this brain region comes online and alters activation in other choice-related regions," Huettel said. "We previously knew that this part of the brain played an important role in simple sorts of behavioral control, but this research shows that it retains a switching role even during complex decision making."

"It's critical not to think of decision-making as just a competition between some rational and emotional parts of our brains," Huettel said. "Instead, the brain selectively enlists many different regions, each performing different computations, depending on the decision problem and our strategic biases."

Huettel said behavioral economics has not made much progress yet toward understanding individual differences between people. "The decision sciences have provided many good models for how we make choices, but most such models assume that the decision maker is a confident, financially savvy young adult who only attends to self-interest," he said. "People vary dramatically in their cognitive abilities, preferences and emotional states. One important goal for neuroeconomic research is to create flexible and biologically plausible models for decision making. Then we could apply those models in unexpected new directions – for example, knowing that older adults have specific deficits in prefrontal cortex function will allow scientists to predict how they will respond in novel decision situations."