Poster E132, Monday, March 27, 2:30 – 4:30 pm, Pacific Concourse
Anxiety differences in reducing reliance on pre-existing biases by learning from outcome feedback
Cristina G. Wilson1, Paul M. Whitney1, John Hinson1; 1Washington State University
Previous research has examined two information processing biases that are characteristic of people with high trait anxiety: (1) increased likelihood that ambiguous stimuli will be interpreted as affectively negative, and (2) greater attention towards threat-related stimuli. Traditionally, the effects of these biases have been studied independently. The goal of the present study was to evaluate the joint influence of biases on risky decisions and to determine whether high trait anxious individuals could use feedback to overcome their pre-existing biases. We developed an ambiguous risk gambling task similar to the Iowa Gambling Task, but with a framing manipulation on each trial to bias people toward gambling or not gambling. In addition to assessments of advantageous decision making, we tested for acquisition of knowledge about choice outcomes through knowledge probes. Skin conductance response (SCR) was measured during the task to assess affective reactions to choices and outcomes. Both high (n = 78) and low (n = 76) trait anxious individuals learned to make better choices over time, thereby reducing the impact of bias, but low trait anxiety individuals showed greater improvement. Knowledge probes and SCR data showed that the poorer decision making by high trait anxious individuals was not attributable to poorer knowledge or altered affective responsiveness. Rather, anxiety differences in performance reflected differences in how task information was weighted. High trait anxious individuals gave less weight to new knowledge gained from feedback and instead relied more strongly on pre-existing biases to make choices.
Topic Area: THINKING: Decision making