A Multi-Motivational Approach to Understanding Polarized Discourse

Dataset for A Multi-Motivational Approach to Understanding Polarized Discourse This is the dataset for the published paper "A Multi-Motivational Approach to Understanding Polarized Discourse". The dataset was disclosed per requested by the review commitee and in support of open science. This dataset is for the review process only. Reuse of the dataset of any research or commerical purpose requires permissions from the original authors. Additional Information and materials are available on our GitHub repository.https://github.com/research-repo-open/A-Multi-Motivational-Approach-to-Understanding-Polarized-Discourse.gitAbstract. Discourse in politically polarized environments is often shaped by layered, psychologically complex motivations that challenge traditional analytic methods. While Quantitative Ethnography (QE) enables structured coding and visualization of discourse patterns, such tools may implicitly assume that utterances reflect singular, codable constructs. This study combines three techniques to explore the possibility of multiple, codable constructs. They include ENA discourse coding, expert psychological analysis, and Ordered Network Analysis (ONA). Using an AI-generated social media thread designed to simulate politically charged commentary, the analysis models relationships among motivational constructs. Three licensed mental health experts interpreted the same dataset, generating divergent but plausible psychological narratives. These differences illustrate the interpretive ambiguity inherent in discourse and underscore the value of multiple analytical lenses. By integrating ONA, the study offers a novel means of visualizing temporal sequencing and affective escalation. However, ethical caution is warranted when applying motivational modeling to real individuals, particularly public figures. This work contributes to efforts within the QE community to refine tools for analyzing affectively charged discourse.「用于理解极化话语的多动机研究方法」数据集 本数据集对应已发表论文《用于理解极化话语的多动机研究方法》，系应审稿委员会要求公开，以支持开放科学事业。 本数据集仅用于论文审稿流程。任何基于该数据集的研究复用或商业使用，均需获得原作者的许可。 更多信息与研究材料可通过本团队的GitHub仓库获取：https://github.com/research-repo-open/A-Multi-Motivational-Approach-to-Understanding-Polarized-Discourse.git 摘要：政治极化环境下的话语往往由多层级、心理层面的复杂动机所塑造，这对传统分析方法构成了挑战。定量民族志（Quantitative Ethnography, QE）虽可实现话语模式的结构化编码与可视化，但此类工具通常隐含假设：单条言论仅反映单一可被编码的心理建构。本研究结合三种技术，探索多条可编码心理建构共存的可能性，所采用的技术包括ENA话语编码、专家心理学分析以及有序网络分析（Ordered Network Analysis, ONA）。研究依托一段由人工智能生成、模拟政治争议性评论的社交媒体讨论线程，对各动机建构间的关联进行建模分析。三名持证心理健康专家对同一数据集进行解读，生成了虽存在分歧但均合乎情理的心理学叙事。这些差异凸显了话语解读中固有的歧义性，也印证了多维度分析视角的价值。通过整合有序网络分析技术，本研究提供了一种可视化时序演进与情感升级的全新路径。但在将动机建模应用于真实个体（尤其是公众人物）时，需保持伦理层面的审慎。本研究为定量民族志学界优化情感性话语分析工具的相关工作提供了有益补充。