HARNESSING SOCIAL MEDIA TO DISPEL XENOPHOBIA: A Policy, Sentiment, and Regression Analysis of News Sources’ Immigration-Related Tweets and Engagement

Abstract

At a time when social media usage and xenophobia are simultaneously at a record, all-time high, this thesis seeks to investigate how social media can be harnessed to dispel xenophobic sentiment by breaking down the online pathways of xenophobic growth and using social media as an educational tool on the immigration experience. The central questions of this thesis are threefold: (1) What are the drivers of xenophobia? (2) How does the media facilitate the growth of xenophobia? (3) What elements of social media can be leveraged to maximize the educational potential of immigrant- related content, with the greater purpose of dispelling xenophobia?

To answer the first two questions, I synthesize the theory and results of prior research with the goal of strengthening our understanding of the intersection of xenophobia and social media, and especially how the latter exacerbates the growth of the former. With that, I found that xenophobia is driven by economic, security, socio-cultural, and psychological drivers. Additionally, the drivers of xenophobia are exacerbated by the most widely used and impactful media sources, which are news media, social media, and news sources on social media. News media can exacerbate the growth of xenophobia through agenda setting, narrative shaping, and framing. Social media can exacerbate the growth of xenophobia through its algorithms, communities, viral content, and lack of sufficient regulation on hate speech and misinformation. And news sources on social media can exacerbate xenophobic growth due to the disparity between how news sources are accessed traditionally and now on social media platforms, and also because of trust and emotional attachment that people have with news sources combined with how social media amplifies the misinformation.

To answer the third question, this thesis relied on a novel data set and a unique quantitative analysis, complemented by qualitative interviews with social media managers of pro-immigrant and anti-xenophobia organizations across America and Europe. I chose to collect tweets between January and April 2021 from mostly American-focused news sources to conduct this data analysis, use engagement as a proxy for educational potential, and focus on tweet sentiment for investigating what elements of social media can be leveraged to maximize the educational potential of immigrant-related content. Being mindful of the current US immigrant relations, I created 3 datasets out of the dataset with all the collected tweets: tweets on immigration, tweets on anti-Asian violence, and tweets on the border crisis. Then, I conducted an exploratory data analysis of the different data sets, a thematic factor analysis, a sentiment analysis, and regression analyses, all in effort to test three hypotheses:

H1: News Sources’ social media posts on immigration lean more towards a negative sentiment, similar to the articles of their traditional newspaper counterparts. H2: The more negative the sentiment of a news source’s social media post on immigration, the higher the engagement of the post. H3: The sentiment intensity of a news source’s social media post has just as strong of a relationship with the post’s engagement as the sentiment direction (positive or negative) of the post.

Based on the quantitative analysis, I found that news sources’ tweets on immigration do not lean more towards the negative side of the sentiment spectrum (nullifying hypothesis 1), there is a statistically significant negative relationship between sentiment and engagement (confirming hypothesis 2), and the relationship between sentiment intensity and the engagement of a news source’s tweet is not as strong as the relationship between sentiment direction and engagement (nullifying hypothesis 3).

Based on the five interviews I conducted with social media directors of pro-immigrant organizations for the qualitative analysis of this thesis, I learned (1) it is imperative to reframe how we talk about immigrant issues through hope-based communication, (2) the most effective form of social media education is publicizing for different audiences in mind, with a focus on people who can actually make change happen and the people who can influence the change, and (3) immigrant voices need to be amplified more.