社会与工程心理学研究室知识库

风险决策的神经经济学研究一直致力于寻找主流决策理论特定属性(如，报酬和概率)的神经基础。这些决策理论假设决策冲突可以通过在概率和报酬之间进行权衡这一补偿性过程解决。尽管已经有行为学研究质疑，人们在进行风险决策时是否遵循这一补偿性规则，但很少有研究从神经经济学角度探寻主流决策理论所遵循的这一补偿性规则假设是否合理。为探究风险决策是否由补偿性规则引导这一问题，我们发展了一种需要补偿性的决策过程参与的决策任务，即基于判断的选择任务(judgment-based choice task)，并将其与偏好选择任务(preferential choice task)进行对比研究。我们的基本假设是:如果人们真实的偏好选择是遵循补偿性规则，那么在执行该任务时关键脑区的神经元活动应与基于判断的选择任务没有区别，否则应有不同;这种不同一方面可能体现在与决策冲突有关的脑区活动模式上(假设1) ，另一方面可能体现在与决策的两个维度相关的脑区的功能整合程度上(假设2)。为验证这一假设，我们采用了事件相关的功能磁共振实验范式，从激活区检测和脑功能连接两种角度，对比了在执行偏好选择任务和基于判断的选择任务时脑功能活动的差异。假设1:采用基于一般线性模型的激活区检测方法，我们发现偏好选择任务和基于判断的选择任务均激活了外侧额口卜、外侧顶叶、脑岛、纹状体、丘脑，均负激活内侧额叶、内侧顶叶、颖叶前部等脑区;但是，与冲突监测和情感活动有关的脑区，如双侧背侧前扣带、脑岛和内侧前额叶皮层，在偏好选择任务中更活跃;没有发现在基于判断的选择任务中更活跃的脑区。进一步功能连接分析发现，背侧前扣带皮层与脑岛的功能连接强度在偏好选择时更强。提示，在偏好选择任务中经历了更多的冲突和相应的情感体验。假设2:根据文献以及在偏好选择任务和基于判断的选择任务中共同激活的脑区，我们选择了与概率有关的脑区和与奖赏幅度有关的脑区作为种子点，采用功能连接的分析方法研究了这些脑区之间功能连接强度在这两个任务之间的差异。从功能连接模式看，在这两个任务中，这些脑区可清晰地分为两个子网络，即由外侧额顶和脑岛等与概率有关脑区组成的子网络，和由纹状体和丘脑等与奖赏幅度有关脑区组成的子网络。从脑区间功能连接强度看，基于判断的选择任务中概率有关脑区和奖赏幅度有关脑区之间的连接强度显著强于偏好选择任务，月_概率有关脑区之间的连接也强于偏好选择任务。提示，在偏好选择时不需要那么强的概率与奖赏幅度信息的整合。这些发现提供了神经影像学证据支持了偏好选择不能简单地由补偿性过程所刻画，相反，偏好选择可能是一种经历了更多冲突和情感活动的非补偿性过程;人们在进行偏好选择时并不一定需要整合概率和奖赏两个维度的信息去做决策。因此，我们的研究结果质疑了假定遵循纯粹的补偿性规则的决策理论是否能恰当地描述个体的风险决策行为。

Neuroeconomics research on risky choice has examined the neural bases of specific characteristics (i.c., payoff and probability) of the mainstream decision theories. Few studies have paid attention to the underlying assumptions of these theories, which assume that the problem of decision conflict can be solved by a compensatory process of trading off probability against payoff. To explore whether risky choice is guided by a compensatory rule or not, we developed a judgment-based choice task that demanded a compensatory decision process and compared it with individuals' preferential choice tasks Using event-related functional magnetic resonance imaging, we examined whether neural activity would differentiate the preferential choice from the judgment-based choice. The results revealed that the regions related to conflict monitoring, emotion, and task difficulty, such as the bilateral dorsal anterior cingulate cortices (dACC), insula and medial prefrontal cortex among other regions were activated more in the preferential choice task. And the functional connectivities between the dACC and the bilateral insula were stronger in the preferential choice task. On the other hand, we found the functional connectivities between the regions associated with probability and those regions associated with reward magnitude were stronger in the judgment-based choice task. Additionally, the connectivities among the regions associated with probability were stronger in the judgmcnt-based choice task. These findings provide neural evidence that preferential choice in risk cannot be simply characterized by a compensatory process. Rather, risky choice might generally be a noncompensatory process suffering from more conflicts experiencing more emotions, but undergoing less complex operations. And -the payoff is not necessarily weighted by the probability when making risky choice. Our results, therefore, question whether the family of existing theories that assume a compensatory rule can provide an adequate description of individual risky decision.