Chatbots have a growing role to play in political discourse, including in political campaigns, voter mobilization ventures, and dissemination of political news, though chatbots in the political domain are relatively understudied. While testing the machine heuristics and political self-concepts frameworks, we carried out a 2 × 2 experiment where both perceived conversational partner (i.e., bot, human) and topic (i.e., political, casual) were manipulated ( N = 126). During the experiment, participants exchanged chat messages with trained research confederates for 30 min. In support of the machine heuristics and political self-concepts frameworks, participants assigned to human partners reported more positive relationships and higher political interest. Through moderation analysis, liking the partner was found to differ between the perceived partner conditions, with perceived political knowledge varying more in the human conditions. Thus, the experimental findings add nuance to interpersonal (i.e., impression management and social identity theory) and human-computer interaction theories (i.e., machine heuristics and Computers Are Social Actors), and have broader implications for online political interactions and for decisionmakers of online political discourse spaces.

中文翻译：

---

人与否？：利用聊天机器人操作来测试机器启发和政治自我概念的实验

聊天机器人在政治话语中发挥着越来越大的作用，包括政治竞选、选民动员活动和政治新闻传播，尽管政治领域的聊天机器人的研究相对较少。在测试机器启发式和政治自我概念框架时，我们进行了 2 × 2 实验，其中感知的对话伙伴（即机器人、人类）和主题（即政治、休闲）都被操纵（N = 126）。实验期间，参与者与训练有素的研究伙伴交换了 30 分钟的聊天消息。为了支持机器启发法和政治自我概念框架，分配给人类伙伴的参与者报告了更积极的关系和更高的政治兴趣。通过适度分析，发现对伴侣的喜欢程度因感知到的伴侣条件而异，感知到的政治知识在人类条件下的差异更大。因此，实验结果为人际（即印象管理和社会认同理论）和人机交互理论（即机器启发法和计算机是社会参与者）增添了细微差别，并对在线政治互动和在线决策者产生了更广泛的影响。政治话语空间。