Department of Psychology, Michigan State University.
J Exp Psychol Gen. 2014 Feb;143(1):142-62. doi: 10.1037/a0031106. Epub 2012 Dec 31.
Beyond simply a decision process, repeated risky decisions also require a number of cognitive processes including learning, search and exploration, and attention. In this article, we examine how multiple response pathways develop over repeated risky decisions. Using the Balloon Analogue Risk Task (BART) as a case study, we show that 2 different response pathways emerge over the course of the task. The assessment pathway is a slower, more controlled pathway where participants deliberate over taking a risk. The 2nd pathway is a faster, more automatic process where no deliberation occurs. Results imply the slower assessment pathway is taken as choice conflict increases and that the faster automatic response is a learned response. Based on these results, we modify an existing formal cognitive model of decision making during the BART to account for these dual response pathways. The slower more deliberative response process is modeled with a sequential sampling process where evidence is accumulated to a threshold, while the other response is given automatically. We show that adolescents with conduct disorder and substance use disorder symptoms not only evaluate risks differently during the BART but also differ in the rate at which they develop the more automatic response. More broadly, our results suggest cognitive models of judgment decision making need to transition from treating observed decisions as the result of a single response pathway to the result of multiple response pathways that change and develop over time.
除了决策过程本身，重复的风险决策还需要一系列认知过程，包括学习、搜索和探索以及注意力。在本文中，我们研究了重复的风险决策如何形成多种反应途径。我们以气球模拟风险任务（BART）为例，表明在任务过程中会出现两种不同的反应途径。评估途径是一种较慢、更受控制的途径，参与者会深思熟虑是否冒险。第二种途径是更快、更自动的过程，不会发生深思熟虑。结果表明，随着选择冲突的增加，人们会采取较慢的评估途径，而较快的自动反应则是一种习得的反应。基于这些结果，我们修改了现有的 BART 决策认知模型，以解释这两种反应途径。较慢、更慎重的反应过程通过顺序采样过程进行建模，其中证据累积到阈值，而另一种反应则自动给出。我们发现，有品行障碍和物质使用障碍症状的青少年不仅在 BART 中对风险的评估方式不同，而且在形成更自动反应的速度上也存在差异。更广泛地说，我们的结果表明，判断决策的认知模型需要从将观察到的决策视为单一反应途径的结果转变为随时间变化和发展的多种反应途径的结果。
