Correlations Between Adolescents Social Media Activity and Well-Being through DLATK Analysis

Authors: Myra Arif, Andrew Schwartz, Selin Akcay

SUNY Campus: Stony Brook University

Presentation Type: Poster

Location: Old Union Hall

Presentation #: 25

Timeslot: Session C 1:45-2:45 PM

Abstract: The relationship between social media usage and well-being has been studied extensively, but the types of social media usage remain underexplored. This study aims to understand why adolescents use social media and how it affects their well-being. Our objective is to determine whether adolescent social media usage strongly correlates with well-being by analyzing factors such as stress, life satisfaction, sadness, relaxation, worry, happiness, and energy levels. Through the Character Lab Research Network, we asked adolescents when social media makes them feel lasting happiness, temporary satisfaction, stress, or the inability to give it up. Natural language processing allowed us to efficiently organize the data, facilitating correlation assessments between social media use and well-being. Using MySQL and the Differential Language Analysis Toolkit (DLATK), we categorized responses to identify significant correlations. Results are displayed through word clouds that group topics based on the significance (p < 0.05) of survey response features. The font size of each word reflects how often adolescents associate these phrases with social media. These visualizations help illustrate key correlations, supporting our hypothesis that specific social media behaviors impact adolescent well-being. Our findings provide insight into how social media influences adolescent emotions. Future work will refine the data further and implement advanced correlation analyses to deepen our understanding of these relationships.