Data in Brief - Dataset of Annotated X (Twitter) Misinformation Posts with Human and Large Language Model (LLM) Responses for Cognitive Analysis

We provide a dataset which can help researchers to conduct cognitive analysis on social media posts such as: study how individuals perceive, interpret and react to online posts. Cognitive analysis of LLMs not only analyse user metrics and binary misinformation labels but it also aims to uncover the underlying reasoning patterns, attitudes, and biases that shape social media behaviour. This perspective is crucial for understanding why certain misinformation narratives gain virality, how beliefs evolve, and what influences user trust or disbelief in online content.

我们提供一个数据集，可助力研究人员对社交媒体帖子开展认知分析（cognitive analysis），例如：探究个体如何感知、解读和回应线上帖子。大语言模型（LLM）的认知分析不仅考察用户指标与二元错误信息标签，更旨在揭示塑造社交媒体行为的潜在推理模式、态度及偏见。这一视角对于理解特定错误信息叙事何以获得传播力、信念如何演变，以及哪些因素影响用户对线上内容的信任或怀疑至关重要。