Bishop and fellow researchers used decision-making tasks, behavioral and physiological measurements and computational models to gauge the probabilistic decision-making skills of 31 young and middle-aged adults whose baseline anxiety levels ranged from low to extreme. Probabilistic decision-making requires using logic and probability to handle uncertain situations, drawing conclusions from past events to determine the best choice.
“An important skill in everyday decision-making is the ability to judge whether an unexpected bad outcome is a chance event or something likely to reoccur if the action that led to the outcome is repeated,” Bishop said.
The researchers’ measures also included eye-tracking to detect pupil dilation, an indicator that the brain has released norepinephrine, which helps send signals to multiple brain regions to increase alertness and readiness to act.
Participants were asked to play a computerized “two-armed bandit-style” game in which they repeatedly chose between two shapes, one of which, if selected, would deliver a mild to moderate electrical shock.
To avoid getting shocked, participants needed to keep track of the shape that most frequently delivered electrical jolts. During one part of the game, the shock-delivering shape did not change for a long stretch of time. However, during another part of the game, it changed more frequently. Highly anxious people had more trouble than their less anxious counterparts adjusting to this and thus avoiding shocks.
“Their choices indicated they were worse at figuring out whether they were in a stable or erratic environment and using this to make the best choices possible,” Bishop said.
Also weaker in highly anxious participants was their pupil response to receiving a shock (or not) during the erratic phase of the game. Typically, our pupils dilate when we take in new information, and this dilation increases in volatile environments. Smaller pupils suggested a failure to process the rapidly changing information that was more prevalent during the erratic phase of the game.
“Our findings help explain why anxious individuals may find decision-making under uncertainty hard as they struggle to pick up on clues as to whether they are in a stable or changing situation,” Bishop said.
In addition to Bishop, co-authors and researchers on the study were Timothy Behrens, Michael Browning, Gerard Jocham and Jill O’Reilly at the University of Oxford in England. The study was funded by grants from the European Research Council and the National Institutes of Health.