Are Algorithmic Biases on Social Media Platforms Increasing Political Divide?

(Warning): Unfortunately, I kept having issues with my mic, so near the end of the video, there was some static. The volume may be loud during that time, so be careful with the volume level.

Reflection

I found this LEAP 3 project to be very interesting but also very difficult. Most of this difficulty came due to the question that I had asked. The research inquiry question that I asked is:

Are Algorithmic Biases on Social Media Platforms Creating a Political Divide?

I wanted to research this question because, from my own personal experience, I have found that on the internet, social media is forming rabbit holes and ideological bubbles where people on one side of the political party never see content from the other side. However, I did not expect this question to be so deep and complicated.

I conducted my research through Google Scholar. Reflecting now, it is ironic that I did not utilize other internet browsers or information collections. While I believe Google Scholar does a very good job of providing unbiased information, there still could be some bias in what it recommends as the top results.

Regardless, I found all of my sources through Google Scholar and read through many articles analyzing data from multiple methods and platforms analyzing algorithmic biases on social media platforms. One of the most commonly looked at platforms by these sources was Facebook. I believe this is because it has a very polarised user base, and it would be more familiar with researchers who are, on average, older in age. To make the video, I used this website: https://online-video-cutter.com/video-editor. This website was very helpful due to my limited experience with video editing. I made the graphic in Google Slides.

I was surprised to learn how complicated this issue was, from the difficulties in collecting data to the mixed results. I was also surprised at how inconclusive much of the data was. Many of the findings could not conclude that social media algorithms had a direct correlation to the political divide. However, it was clear that they were having an effect on the spread of misinformation and hate speech. I believe this is because, at the end of the day, people’s ability to make decisions for themselves, regardless of the news that is shown, is the driving factor. Social media may just be exposing people's differences more due to the global nature of it. Before social media, people would not come into contact with opinions from geographically far away, as often due to being surrounded by people with primarily similar opinions and not having a global network of people to interact with.

One thing that I learned about myself was how my own biases played into my research topic. At the beginning of my research, I had projected my own experience with social media algorithms and had already formulated a conclusion in my head before any research. During the 2016 election, I became a victim of a rabbit hole myself when I started watching and being continuously recommended right-leaning YouTube channels. Now that I have different political views today, I may have put too much weight on the effects of social media algorithms giving me those political views and not considering how my family was heavily right-leaning politically. Seeing how the research was as mixed as it was made me realize my own biases and how I was projecting my personal experiences were affecting my view of the data.

After completing this project, I would like to look more into how much social media is creating and deepening divides vs. how much social media is just exposing previously existing divides. Throughout this project, I considered how people’s political beliefs were formed before and after social media. It made me think about how, before social media, people, especially those living in rural areas, would only have access to people who share their views. I think my research was primarily about the negative effects social media is having. Still, I would also like to look into social media's potential benefits in spreading information to previously ideologically isolated areas.

Annotated Bibliography

Barberá, P. (n.d.). 3 Social Media, Echo Chambers, and Political Polarization. Retrieved April 28, 2024, from https://www.cambridge.org/core/services/aop-cambridge-core/content/view/333A5B4DE1B67EFF7876261118CCFE19/9781108835558c3_34-55.pdf/social_media_echo_chambers_and_political_polarization.pdf#:~:text=URL%3A%20https%3A%2F%2Fwww.cambridge.org%2Fcore%2Fservices%2Faop

This source examines social media's influence on political polarization. It focuses mainly on how social media can create echo chambers. I chose this because the author is a notable figure in political sciences and does an excellent job explaining how echo chambers can create a political divide.

Brown, M., Nagler, J., Bisbee, J., Lai, A., & Tucker, J. (2022, October 13). Echo chambers, rabbit holes, and ideological bias: How YouTube recommends content to real users. Brookings. https://www.brookings.edu/articles/echo-chambers-rabbit-holes-and-ideological-bias-how-youtube-recommends-content-to-real-users/

This article focuses on YouTube’s algorithm and how it might tend to lead users into rabbit holes. I chose this article because it has a good range of data and shows how the results for my research topic vary greatly. In this case, YouTube was not found to be creating rabbit holes.

Garcia, D. (2023). Influence of Facebook algorithms on political polarization tested. Nature. https://doi.org/10.1038/d41586-023-02325-x

This article focuses on Facebook’s algorithm and how it may create a divide. I chose this source because it provided empirical evidence to suggest that algorithmic content recommendations can significantly influence political views.

Gil de Zúñiga, H., Cheng, Z., & González-González, P. (2022). Effects of the news finds me perception on algorithmic news attitudes and social media political homophily. Journal of Communication. https://doi.org/10.1093/joc/jqac025

This article discusses social media users' “News finds me” perception. This is the belief that social media will do a good enough job showing the user without seeking other sources. These sources found a link between people who believe in the perception and political division.

Gran, A.-B., Booth, P., & Bucher, T. (2020). To be or not to be algorithm aware: a question of a new digital divide? Information, Communication & Society, 24(12), 1779–1796. https://doi.org/10.1080/1369118x.2020.1736124

This paper examines how users' knowledge of how algorithms work affects their perception of news generated for algorithmic recommendations. The author argues that digital media literacy can shape users' experiences when consuming digital content.

Greene, C. (2019). DSpace Angular Universal. Dr.lib.iastate.edu. https://dr.lib.iastate.edu/handle/20.500.12876/31870

This data-heavy article examines how the data can be analyzed to show a correlation between political divide and algorithmic biases. They found that biased news reporting has increased political polarization in the United States.

Jingnan, H. (2023, July 23). New study shows just how Facebook’s algorithm shapes conservative and liberal bubbles. Opb. https://www.opb.org/article/2023/07/27/new-study-shows-just-how-facebook-s-algorithm-shapes-conservative-and-liberal-bubbles/

This article looks at Facebook’s algorithms and how they create ideological bubbles. I chose this article because it helps to show how different platforms create political bubbles of different beliefs. The article showed how the difference in content shown to users belief to be left of right-leaning varied greatly.

Kulshrestha, J., Eslami, M., Messias, J., Zafar, M. B., Ghosh, S., Gummadi, K. P., & Karahalios, K. (2017). Quantifying Search Bias. Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work and Social Computing. https://doi.org/10.1145/2998181.2998321

While not completely about social media, this article looked at how search results can be biased and affect a user’s perception of the political landscape. I chose this because it shows how different parts of the internet could be creating a divide.

Prasetya, H. A., & Murata, T. (2020). A model of opinion and propagation structure polarization in social media. Computational Social Networks, 7(1). https://doi.org/10.1186/s40649-019-0076-z

This article examines how differing opinions are propagated through social media. I chose it because it examined how information spreads on social media.

Sîrbu, A., Pedreschi, D., Giannotti, F., & Kertész, J. (2019). Algorithmic bias amplifies opinion fragmentation and polarization: A bounded confidence model. PLOS ONE, 14(3), e0213246. https://doi.org/10.1371/journal.pone.0213246

This article looked into how algorithmic bias can be found to amplify the political divide. I chose this article because it shows how social media algorithms do not just create division but also amply pre-existing division.