This was an 18 month research project that ran between June 2020 and December 2021 and examined the interaction between miscommunications and conspiracy theories in relation to key factors such as anonymity, membership length, peer groups and postage frequency, within the context of the current Covid-19 pandemic and Islamophobia on social media. The project was hosted at Birmingham City University and funded by the UKRI and Economic and Social Research Council (ESRC) under their Covid-19 rapid response call. The project explored irrational beliefs and thoughts that are disseminated on social media, covering important coverage of communications surrounding conspiracy theories online whilst paying attention to the content associated to racist ‘infodemic’ messages. The project also sought to provide insights into the drivers of Covid-19 narratives and consequences in fuelling existing extreme communications and Islamophobic language both online and offline.There is a direct gap in understanding how conspiracy theories and miscommunication on social media sites is being used to create a Covid-19 'infodemic'. This is particularly relevant in the context of Muslim communities as members of the far-right are able to use irrational beliefs and fake news ideology to peddle hate, with such narratives quickly being able to penetrate the mainstream and become normalised. For example, one video shared on the social media site Telegram by the former leader of the English Defence League, Tommy Robinson, alleges to show a group of Muslim men leaving a secret mosque in Birmingham to pray. West Midlands Police debunked this video as being fake as the Mosque had already been closed down. A number of similar examples show the rising tension of fear fake news creates and the implications for such information which risks alienating communities and can have a real significant offline effect where people become more insular. Understanding the drivers of such communication is critical to ensuring a more effective and trustworthy media source where complex information, can be used to aid policy-makers and the wider general public. This study will address this gap through our rich empirical data that can be used to highlight what law enforcement should do when confronting online conspiracy theories and offline attacks. The nature of such information is that it can spread quickly and our project will address the drivers of this and the perpetrators involved which will be significant for social media companies, the police, policy-makers and other key stakeholders. The current climate of conspiracy theories and racist 'infodemic' miscommunication on Covid-19 can have significant consequences when social distancing measures are lifted. Due to the nature of social media, and the range of social media comments and behaviour gathered, this project will be able to focus on national issues as we identify trigger events. The detail provided by the social media comments, including in some cases location (either explicitly in the social media comment, user profile, or comment meta-data), will allow for there to be a focus on certain regions with the UK, or countries as a whole. This may facilitate the tracking of and response to localised issues linked to Covid-19, extreme content and miscommunication.

This research project consisted of four studies that examined language, sentiment, narratives, cases studies and the relationship between the online and offline COVID-19 misinformation theories and Islamophobia across Twitter and YouTube. We examined both language and sentiment used when discussing COVID-19 misinformation and related Islamophobia. Our approach involved two separate stages of analysis. One with a quantitative focus on language and the comparison of language used between different user types. This involved categorising users based on their levels of anonymity, membership length, and postage frequency and then using corpus linguistic analysis to explore differences in language use between users. The second approach involved using the same sample, but instead using sentiment analysis to determine if there were any differences in emotion or sentiment being expressed in these comments quantitatively.