DocumentCloud_Epstein_Docs_p01398.png

Case 18-2868, Document 283, 08/09/2019, 2628241, Page431 of 883 the articles, to which the articles containing the statements made against Ms. Giuffre could have been disseminated, assuming these individuals are all unique and have not already read one of the articles. iv. However, I did not include these social media shares in my calculations. Vv. Since news article viewing follows a power law”’ distribution”®, there is no direct linear ratio of number of social media shares to readership. There is published research that does report average of views of an article on a news website and also average social media shares’. In a direct calculation with numbers from this article*’, 23 articles views per social media share, using 33,758 social media shares, this would be 776,434 article views. However, this ratio would vary by website, number of daily unique visitors, type of news article, time for accumulating shares, and possibly other factors. Plus, this number would not account for the people receiving the social media share that viewed the title, post, and snippet but did not click on the share to view the article on the website, thereby undercounting views of the articles. vi. Also, given the topical nature of the underlying news story, one could expect lower social media sharing but higher article viewing, as people will tend to read articles on such topics privately but not share on social media*!. So, I would expect the social media number itself to be an undercount. h. I did not include articles that link to one of the articles containing the statements made against Ms. Giuffre in my calculations of dissemination. Unless the article >? https://en.wikipedia.org/wiki/Power_law 28 See for example, Tatar, A., de Amorim, M. D., Fdida, S., & Antoniadis, P. (2014). A survey on predicting the popularity of web content. Journal of Internet Services and Applications, 5(1), 1. 2° See for example, Castillo, C., El-Haddad, M.., Pfeffer, J., & Stempeck, M. (2014, February). Characterizing the life cycle of online news stories using social media reactions. In Proceedings of the 17th ACM conference on Computer supported cooperative work & social computing (pp. 211- 223). ACM. *° Castillo, C., El-Haddad, M., Pfeffer, J., & Stempeck, M. (2014, February). Characterizing the life cycle of online news stories using social media reactions. In Proceedings of the 17th ACM conference on Computer supported cooperative work & social computing (pp. 211-223). ACM. 3! See for example, Agarwal, D., Chen, B. C., and Wang, X. Multi-faceted ranking of news articles using post-read actions. In Proc. of CIKM, ACM (2012), 694-703. 30