Survey results from the experiment on the effect of emotional recognition on user satisfaction

The files correspond to the pre-test and the actual survey results collected with participants recruited via Prolific. The survey is constructed in Qualtrics. This is a 2x2x2 between-subjects study, with emotion acknowledgment (yes/no), agent type (human vs. chatbot), and emotional context (positive vs. negative) being the conditions. We measured how our conditions affect user satisfaction through 4 mediating variables: trustworthiness, social presence, empathy, and privacy risk.

本数据集文件包含通过Prolific平台招募的受试者所参与的预测试与正式调研的结果数据。本次调研基于Qualtrics平台搭建完成。本研究为2×2×2被试间实验设计，共设置三类实验条件：情绪确认（是/否）、智能体类型（人类 vs 聊天机器人）以及情绪情境（积极 vs 消极）。本研究选取可信度、社会临场感、共情与隐私风险作为中介变量，探究各实验条件对用户满意度的影响机制。