I quickly checked the impact of the Facebook MisinformationPolicy (announced on March 7, 2019) on the endorsements of vaccine-related posts by Italian misinformation actors, employing interrupted time series analysis (you can read about this technique in the didactic material I have made available online at https://nicolarighetti.github.io/Time-Series-Analysis-With-R/).
There is a small instantaneous drop in likes, a slight inversion in the overall trend, and a persistent decrease in the average likes from before to after the policy implementation. The analysis is just exploratory and has several limitations (e.g., small sample size), but the effect looks clear.
The analysis is directly inspired by the recent paper of Gu et al. It is, in fact, a quick check of their findings in the Italian context: Gu, J., Dor, A., Li, K., Broniatowski, D. A., Hatheway, M., Fritz, L., & Abroms, L. C. (2022). The impact of Facebook’s vaccine misinformation policy on user endorsements of vaccine content: An interrupted time series analysis. Vaccine, 40(14), 2209-2214.
The Facebook accounts spreading misinformation that I employed in the analysis come from the MINE project: Giglietto, F., Farci, M., Marino, G., Mottola, S., Radicioni, T., & Terenzi, M. (2022, January 7). Mapping Nefarious Social Media Actors to Speed-up Covid-19 Fact-checking. https://lnkd.in/eRMzrrvP