The study, as yet unpublished, looked at millions of tweets in the weeks before and after the US announced its withdrawal from the Paris climate agreement in June 2017. Marlow and colleagues found that about 25% of all climate-related tweets were generated by bots on an average day during the period they studied. Most of these were supportive of the Trump administration’s decision to withdraw from the climate agreement and/or were spreading misinformation about climate science (see the Guardian story for details about the research and results).
DSI was happy to have supported this work through one of our Data Science @ Brown grants, providing the funds to purchase over 17 million tweets from Twitter. We have recently awarded the second round of these grants, and will be accepting applications again in Fall 2020.