University of Melbourne, Parkville, VIC, Australia.
Melbourne Centre for Behaviour Change, University of Melbourne, Parkville, VIC, Australia.
Sci Rep. 2024 Nov 12;14(1):27647. doi: 10.1038/s41598-024-76139-w.
Misinformation about climate change is a complex societal issue that requires holistic, interdisciplinary solutions at the intersection between technology and psychology. One proposed solution is a "technocognitive" approach, involving the synthesis of psychological and computer science research. Psychological research has identified that interventions that counter misinformation require both fact-based (e.g., factual explanations) and technique-based (e.g., explanations of misleading techniques and logical fallacies) content. However, little progress has been made on documenting and detecting fallacies in climate misinformation. In this study, we apply a previously developed critical thinking methodology for deconstructing climate misinformation in order to develop a dataset mapping examples of climate misinformation to reasoning fallacies. This dataset is used to train a model to detect fallacies in climate misinformation. We evaluate the model's performance using the score, which measures how well the model detects relevant cases while avoiding irrelevant ones. Our study shows scores that are 2.5-3.5 times better than previous works. The fallacies that are easiest to detect include fake experts and anecdotal arguments, while fallacies that require background knowledge, such as oversimplification, misrepresentation, and slothful induction, are relatively more difficult to detect. This research lays the groundwork for development of solutions where automatically detected climate misinformation can be countered with generative technique-based corrections.
气候变化错误信息是一个复杂的社会问题，需要在技术和心理学的交叉点上采取整体的、跨学科的解决方案。一种被提议的解决方案是“技术认知”方法，涉及心理学和计算机科学研究的综合。心理学研究已经确定，纠正错误信息的干预措施既需要基于事实的（例如，事实解释），也需要基于技术的（例如，对误导性技术和逻辑谬误的解释）。然而，在记录和检测气候错误信息中的谬误方面几乎没有取得什么进展。在这项研究中，我们应用了先前开发的用于解构气候错误信息的批判性思维方法，以便开发一个将气候错误信息示例映射到推理谬误的数据集。该数据集用于训练模型来检测气候错误信息中的谬误。我们使用 分数来评估模型的性能，该分数衡量模型在检测相关案例的同时避免不相关案例的能力。我们的研究表明， 分数比以前的工作提高了 2.5-3.5 倍。最容易检测到的谬误包括假专家和轶事论证，而需要背景知识的谬误，如过度简化、歪曲和懒惰归纳，则相对较难检测到。这项研究为开发解决方案奠定了基础，在这些解决方案中，可以使用自动检测到的气候错误信息来对抗基于生成技术的纠正措施。
