Conspiracy theories are a current and dangerous misinformation problem. Unfortunately, during the Covid-19 pandemic, many conspiracy theories, such as QAnon and 5G’s alleged link to Covid-19, spread and became mainstream. As a result, I decided to conduct a research project discussing conspiracy theories during the Covid-19 pandemic.
I first determined whether to obtain a set of viral tweets or tweets from a Twitter handle known for disseminating misinformation. I chose to use the latter, as it would allow me to identify whether a handle spreads a specific conspiracy theory consistently or many different conspiracy theories over time. I obtained a set of tweets from a twitter handle known for spreading conspiracy theories and misinformation. While I examined both the visual and verbal misinformation, I chose to focus on verbal misinformation for the duration of this summer fellowship. I began evaluating the tweets by generating the unigrams, bigrams, trigrams, and tetragrams for the entire corpus of tweets. Unfortunately, even without stopwords, there aren’t a lot of overlapping trigrams or tetragrams. I have not begun to evaluate the unigrams and bigrams yet.
While I initially began working with a 17-month set of tweets, I quickly realized that a 7-month timeframe would better serve me. This is both due to the subtext of tweets (although these tweets contain valuable information, they often reference or support other statements) and the general scalability. I first selected the timeframe (August 2020-March 2021) based on vaccine rollout as well as tensions within the United States political atmosphere. This left me with about 600 tweets to evaluate. I’ve been hand coding themes found in the tweets. This process has also allowed me to research different references found in the tweets-many of these tweets discuss other countries, such as Ireland and Germany. As a result, I’ve also explored what Covid-19 pandemic public policy was in each of these countries. My future work will focus on identifying which themes are most prominent over time, as well as designing an interactive visualization meant to show the use of different conspiracy theories’ prominence or insignificance over time. I have also begun working on the corresponding paper for the tweets, which I will hopefully submit for publication at the end of the summer.
I’ve currently been using Excel to store and the tweets and will be using Tableau to show the discussion of different conspiracy theories over time. I will also use Weka to conduct a sentiment analysis on the tweet if I get a chance.