Assessing the informational credibility of conspiracy theories: online discussion about the Nord Stream damage

ASLIB JOURNAL OF INFORMATION MANAGEMENT(2023)

引用 0|浏览0
暂无评分
摘要
PurposeTo elaborate the picture of credibility assessment by examining how participants of online discussion evaluate the informational credibility of conspiracy theories.Design/methodology/approachDescriptive quantitative analysis and qualitative content analysis of 2,663 posts submitted to seven Reddit threads discussing a conspiracy operation, that is, the damage of the Nord Stream gas pipelines in September 2022. It was examined how the participants of online discussion assess the credibility of information constitutive of conspiracy theories speculating about (1) suspected actors responsible for the damage, (2) their motives and (3) the ways in which the damage was made. The credibility assessments focussed on diverse sources offering information about the above three factors.FindingsThe participants assessed the credibility of information by drawing on four main criteria: plausibility of arguments, honesty in argumentation, similarity to one's beliefs and provision of evidence. Most assessments were negative and indicated doubt about the informational believability of conspiracy theories about the damage. Of the information sources referred to in the discussion, the posts submitted by fellow participants, television programmes and statements provided by governmental organizations were judged most critically, due to implausible argumentation and advocacy of biased views.Research limitations/implicationsAs the study focuses on a sample of posts dealing with conspiracy theories about a particular event, the findings cannot be generalized to concern the informational credibility conspiracy narratives.Originality/valueThe study pioneers by providing an in-depth analysis of the nature of credibility assessments by focussing on information constitutive of conspiracy theories.
更多
查看译文
关键词
conspiracy theories,informational credibility,online discussion
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要