Quantifying political discourse of Latino-American Facebook groups
Project Summary
January 2021 — March 2021
Claudia-Flores Saviaga, Saiph Savage
Minorities have a greater disposition of encountering electoral-misinformation to fill an absence of relevant content.
Compare amounts of political discourse in the 8 weeks leading up to the U.S. election in Latino Facebook groups to determine where vulnerabilities exist.
News media pages published the most politically-relevant posts. Separating by political theme, I found that political groups made the most content associated with racism. Community groups produced the most interest in immigration and economic discourse. Further analysis of community groups revealed that majority of posts were artificially generated, spam content.
Findings
Importance
What are Data Voids?
Data voids are the absence of reliable information that highlight malicious and fictitious information.
Do they exist in social media?
In social media data voids, users take a more passive role in absorbing content. When users see information that they are not invested enough or lack personal experience to fact check, they may be more willing to accept it as true. Krull and Anderson rationalize this phenomenon with the process of explanation.
Coronavirus Data Voids
The most recent example of this phenomenon would be early-stage coronavirus-related content. Misinformation went viral filling in the data void for any information on how to prevent infection.
Latino News Coverage
Latino American social media has been targets of misinformation due to leveraged cultural values and less access to authoritative information. A survey from the Pew Research Center was used to identify themes of interest.
I created a list of proxy words and phrases, those that might be used to show interest in one of the themes above. Economy and economic inequality were merged to simplify categorization.
Facebook groups with an explicit Latino focus were identified and sorted into different categories as they affect the context of political discussion.
A study on Latino Media by the City University of New York was used to find reputable information outlets.
Facebook posts from these groups were collected over an 11 week period between Joe Biden's presidential nomination and 2020 election day. More than 270,000 posts were collected across nearly 1,200 groups.
A statistical analysis found that group categorization impacted political visibility except for foreign policy.
News pages published the most content on average and in total. However, there was heavy focus on coronavirus information leaving room for misinformation in less-discussed topics.
Removing coronavirus-related information and observing the mean posts per category, we can view how political groups have a tendency of capitalizing on racism-related discourse.
The above visualization also highlights the community demand for economic and immigration-related information. But why is immigration, a topic that was surveyed as less popular than the Supreme Court and climate change, have about three times as much conversation than either of those topics? And what political perspectives are being shared in these posts?
I conducted a small qualitative analysis by randomly sampling immigration posts to understand the discussions.
Most of these posts solicited immigration services. While the services advertised could be legitimate, many had low engagement rates and used terminology associated with immigration scams. The lack of Facebook verification features increases the potential of immigration fraud for those needing help. One unique example, was a service targeting Latinos in major midwestern cities, advertising moving to Canada.
In the context of politics, there are two main concerns with this finding. First, Latinos may be overexposed to immigration discourse and be skewed to believe in a problem with undocumented people in America. Secondly, this study shows how ethnic groups can easily be targeted on Facebook.
Below is a time series analysis of immigration-related posts.
I initially believed that since there were no observable trends, there was a coding error. But with the qualitative evaluation shows the batched messages being shared across multiple groups. In some instances, the same exact posts were found nearly 100 times over the eleven-week period.
This highlights an extreme vulnerability for manipulation and misinformation targeting regional ethnic groups on Facebook. If we consider how little politics are mentioned in community groups, it could be easy for these viewpoints to become unchallenged.
References
Bode, L., & Vraga, E. K. (2017). See Something, Say Something: Correction of Global Health Misinfor-mation on Social Media. Health Communication, 33(9), 1131–1140. https://doi.org/10.1080/10410236.2017.1331312
Bordia, P., & Difonzo, N. (2004). Problem Solving in Social Interactions on the Internet: Rumor As Social Cognition. Social Psychology Quarterly, 67(1), 33–49. https://doi.org/10.1177/019027250406700105
Bright, J. (2016). The Social News Gap: How News Reading and News Sharing Diverge. Journal of Com-munication, 66(3), 343-365. https://doi.org/10.1111/jcom.12232
Chen, X., Sin, S. C. J., Theng, Y. L., & Lee, C. S. (2015). Why Students Share Misinformation on Social Me-dia: Motivation, Gender, and Study-level Differences. The Journal of Academic Librarianship, 41(5), 583–592. https://doi.org/10.1016/j.acalib.2015.07.003
CrowdTangle Team (2020). Crowdtangle. Facebook, Menlo Park, California, United States. [1481516]
Department of Justice U.S. Attorney’s Office. (2021) Phony Immigration Attorney Who Filed Hundreds of Fraudulent Asylum Applications Sentenced to More than 20 Years in Federal Prison. https://www.justice.gov/usao-mdfl/pr/phony-immigration-attorney-who-filed-hundreds-fraud-ulent-asylum-applications-sentenced
FTC. Scams Against Immigrants. (n.d.) Federal Trade Commission Consumer Information. https://www.consumer.ftc.gov/articles/0141-scams-against-immigrants
Golebiewski, M., & Boyd, D. (2019) Data Voids: Where Missing Data Can Easily Be Exploited. Data & Society.
Krull, D. S., & Anderson, C. A. (1997). The process of explanation. Current Directions in Psychological Science, 6(1), 1-5.
Laskot, J. (2020, July 24). Immigration Lawyers Should Be On Facebook- Here Are Some Tips On How And Why. LinkedIn. https://www.linkedin.com/pulse/immigration-lawyers-should-facebook-here-some-tips-how-jared-jaskot/
Olteanu, A., Diaz, F., & Kazai, G. (2020). When Are Search Completion Suggestions Problematic?. Proceedings of the ACM on Human-Computer Interaction, 4(CSCW), 1-25.
Smith, R., Cubbon, S., & Wardle, C.(2020). Under the Surface: Covid-19 vaccine narratives, misinformation and data deficits on social media. First Draft. Retrieved from https://firstdraft-news.org/vaccine narratives-full-report-november-2020
Starbird, K., Spiro, E., Edwards, I., Zhou, K., Maddock, J., & Narasimhan, S., (2016). Could This Be True?: I think So! Expressed Uncertainty in Online Rumoring. Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 360-371). DOI: 10.1145/2858036.2858551