The present study investigates how agents and the moral valence of the acts affect moral judgments when two consecutively behaviors are perceived, with each describing morally salient behaviors done by the same or different agent(s). Participants had to rate the likableness/pleasantness of the agents/behaviors. Behavioral results indicated that rating the likableness of the agent was mainly depended on the morally diagnostic character of the agent while rating the pleasantness of the behaviors was mainly depended on the moral valence of the behaviors per se. ERP results showed: 1) larger N1 was found in response to the agent consistently acting immorally, indicating an early detection of social threatening information. 2) Compared with agents who consistently act morally which provided no norm- or expectation-violation information, other conditions induced larger N400, indicating greater cognitive effort was recruited when the present moral information violated the participants’ prior knowledge to the agent. 3) Increased LPP was found in response to the agent consistently acting morally (vs. moral behaviors acted by different agents), representing that people may allocate more attention to positive information during this stage. We suggest that this three-stage scheme is a common model when people encounter consecutive moral events.

本研究调查了当感知到两个连续行为时，代理和行为的道德效价如何影响道德判断，每个行为都描述了相同或不同代理所做的道德显着行为。参与者必须对代理人/行为的可爱度/愉悦度进行评分。行为结果表明，评价代理人的可爱程度主要取决于代理人的道德诊断特征，而评价行为的愉快程度主要取决于行为本身的道德效价。ERP 结果显示：1) 发现较大的 N1 以响应代理一贯的不道德行为，表明对社会威胁信息的早期检测。2) 与始终遵守道德且未提供违反规范或期望违反信息的代理相比，其他条件会导致更大的 N400，表明当当前的道德信息违反了参与者对代理的先验知识时，会招募更大的认知努力。3) 发现 LPP 增加是对代理人一贯的道德行为的反应（与不同代理人的道德行为相比），这表明人们在这个阶段可能会更多地关注积极信息。我们建议当人们遇到连续的道德事件时，这种三阶段方案是一种常见的模型。代表人们在这个阶段可能会更多地关注正面信息。我们建议当人们遇到连续的道德事件时，这种三阶段方案是一种常见的模型。代表人们在这个阶段可能会更多地关注正面信息。我们建议当人们遇到连续的道德事件时，这种三阶段方案是一种常见的模型。