Important research from @BKCHarvard that highlights the risks of over-reliance on studies evaluating the role of bots and the limitation of automated bot detection.

https://cyber.harvard.edu/news/2020-03/false-positive-problem-automatic-bot-detection-social-science-research

This research highlights that the public policy debate in recent years may have been skewed by a desire for potentially misleading metrics that were based on limited, public data analysis, amplified by some media coverage and commercially-motivated actors.

The desire for simple metrics is understandable, but by its nature misinformation and disinformation is a human problem that goes far beyond how many times an account Tweets. The landscape has changed significantly since the emergence of bot research and analysis.

Our head of Site Integrity, @yoyoel has outlined his views on the analysis in more detail in a thread that is worth reading: https://twitter.com/yoyoel/status/1245443311428497408