Speaker
Description
Optimal decision-making benefits from exploring available options. However, in stable environments exploitation and using the same decision strategy can be more efficient. We tested exploration (gaze pattern) in a base-rate task.
Sixty-two participants had to determine a hypothetical individual’s most likely class membership using two information pieces: base-rates (prevalence) and an individual attribute. The task encompassed congruent, incongruent, and neutral conditions, defined by the alignment of this information. Base rates were either provided before or after attribute information. Responses could be either base-rate congruent (choosing the class favored by the base-rate) or stereotype congruent (choosing the class favored by the attribute). Accuracy, response times, gaze at information pieces (proxy for exploration), and pupil dilation (reflecting cognitive effort) were contrasted across conditions and response types.
Results revealed interindividual differences in response strategies: Stereotype responders primarily gave stereotype congruent responses and seemed to neglect the base-rate information, with no condition effects on response times. Base-rate responders primarily gave base-rate congruent responses and seemed sensitive to base-rate changes, with condition effects on response times and gaze. Deviation from the default response increased response times for both groups. For stereotype responders, base-rate congruent responses were further accompanied by larger pupil dilation, indicating greater cognitive effort. Response time and gaze patterns for base-rate responders in the neutral condition indicated cognitive decoupling even in the absence of conflicting information.
These findings indicate that decision strategies influence exploration (gaze pattern) behaviour.
