Posts: 376 +5
The study undertaken by a team led by PhD candidate Thomas Marlow examined about 6.5 million tweets from the time frame surrounding President Trump's decision to exit the Paris Climate Accord. These tweets were categorized and then passed through a tool called "Botometer" to estimate whether they were produced by a human or a bot.
Marlow told The Guardian that he originally had the idea for the study after wondering "why there’s persistent levels of denial about something that the science is more or less settled on."
Everyone knows Twitter has a bot problem, but the prevalence as shown in this study is what had the authors worried. Bots were responsible for 38% of all tweets mentioning "fake science" and 28% of tweets about Exxon. When looking at tweets in support of science and climate activism, the authors found that just 5% of them were from bots.
Auto-generated content by itself isn't necessarily bad unless it reaches and influences many people. Although the authors couldn't definitively identify who was behind the bot accounts or how much influence they had, they did discover that many had tens of thousands of followers.
Accounts like these tend to follow each other and circulate false information in echo chambers. Regardless of topic, researchers have shown that people keep believing and spreading this misinformation due to their perception that there is a valid alternative opinion.
Stephan Lewandowsky, a co-author from the University of Bristol, adds that "the more denialist trolls are out there, the more likely people will think that there is a diversity of opinion and hence will weaken their support for climate science." It's unclear whether these bots have affected politicians into enacting or repealing any policies, but there is a growing concern that they are beginning to influence government officials.