A Dynamic Analysis of Conspiratorial Narratives on Twitter During the Pandemic

Since the breakout of COVID-19 in late 2019, various conspiracy theories have spread widely on social media and other channels, fueling misinformation about the origins of COVID-19 and the motives of those working to combat it. This study analyzes tweets (N = 313,088) collected over a 9-month period in 2020, which mention a set of well-known conspiracy theories about the role of Bill Gates during the pandemic. Using a topic modeling technique (i.e., Biterm Topic Model), this study identified ten salient topics surrounding Bill Gates on Twitter, and we further investigated the interactions between different topics using Granger causality tests. The results demonstrate that emotionally charged conspiratorial narratives are more likely to breed other conspiratorial narratives in the following days. The findings show that each conspiracy theory is not isolated by itself. Instead, they are highly dynamic and interwoven. This study presents new empirical insights into how conspiracy theories spread and interact during crises. Practical and theoretical implications are also discussed.