社会与工程心理学研究室知识库

摘要：决策是广泛而重要的心理加工过程，其中包括了选项评估、行为输出以及结果反馈等三个阶段。个体对选项特征的加工和评估直接影响到其行为选择，因此考察个体如何评估选项特征具有重要价值，是决策研究面临的新课题。本研究招募35名被试参与金钱博弈任务，采用2(低、高金额)*2(风险、模糊决策背景)的被试内设计，要求被试在可能出现正/负收益的冒险性选项和零收益的安全性选项中做选择。目的是探讨两类选项特征对决策行为的影响，包括“选项价值”和“选项收益概率已知性”；同时采用高时间分辨率的事件相关电位( event-related potential, ERP)技术来考察个体对选项特征评估的过程。结果发现：1）在行为层面上，个体在模糊决策背景中的冒险率显著低于风险背景，且个体在高金额选项中的冒险率显著低于低金额选项，说明相较于收益概率已知或低价值选项条件，个体在收益概率未知或者面对高价值选项的得失时更不愿意采取冒险性行为；相较于风险背景，在模糊背景中低金额选项的冒险率显著高于高金额选项，行为选择受到选项价值影响更大，提示在收益概率未知的情况下，个体很可能更加依赖其他选项特征来辅助其做出决策行为。2）在脑电层面上，个体对两个选项特征加工存在分离。低金额选项比高金额选项诱发更大的P1波幅，推断P1波幅可能反映个体对选项价值的心理编码；风险决策背景比模糊决策背景则诱发更负的N2波幅，意味着在收益概率已知条件个体决策时面临更高的认知冲突水平；再则，P1波幅更大的试次中个体更可能进行冒险行为。结论：本文采用金钱博弈范式并结合ERP技术考察个体进行选项评估的过程，P1、N2等早期ERP成分反映了个体对选项特征进行加工的过程，且P1波幅对个体的选择行为有预测作用。本研究为理解人类决策行为中的选项评估过程提供了初步神经电生理学证据。

Abstract：Decision-making is vital to human behavior, which can be divided into multiple stages including option assessment, behavioral output, and feedback processing. Studying how people evaluate option characteristicswould provide important knowledge about human decision-making. The present study investigated whether and how two kinds of option characteristics influence behavior decisions, that is, option value and knowability of the probabilities of possible outcomes. Using event-related potential (ERP), we examined the neural mechanism of evaluating option characteristics. In this study, 35 volunteers participated in a monetary gambling task, in which they were asked to choose between a risky option and a safe one. Behavioral results revealed that participants made more risky decisions in the small magnitude condition than the larger magnitude condition, and in the risk (known probabilistic outcome) condition than the ambiguity (unknown probabilistic outcome) condition. Interestingly, there was also an interaction between two option characteristics: when participants made ambiguous decision-making, more risky options were chosen in the small magnitude than the large magnitude condition, indicating that if the probabilities of possible outcomes are unknown, participants would be more dependent on other available option characteristics to make decisions.

ERP results showed a double dissociation pattern, with the P1 being sensitive to magnitude but insensitive to the knowability of the probabilities, and the N2 showing the opposite pattern. The results suggest that these two fundamental features of options are assessed rapidly and separately in the human brain. Specifically, the P1 amplitude elicited by small magnitude was larger than large magnitude, inferring that P1 reflected the mental processing of magnitude; meanwhile, more negative N2 amplitude was elicited by the risk condition than the ambiguous one, reflecting that cognitive conflict was increased in the former condition. Specifically, the P1 amplitude predicted human decisions. These findings indicate that the early components (P1, N2) reflect the processing of option characteristics, and P1 amplitude can further predict decision behavior. The study provides preliminary temporal electrophysiological evidence of the processing of option evaluation during decision-making.