A Tale of Two Cities: COVID-19 and the Emotional Well-Being of Student-Athletes Using Natural Language Processing

Abstract

Read online

Student-athletes at the Division I institutions face a slew of challenges and stressors that can have negative impacts in eliciting different emotional responses during the COVID-19 pandemic. We employed machine-learning-based natural language processing techniques to analyze the user-generated content posted on Twitter of Atlantic Coast Conference (ACC) student-athletes to study changes in their sentiment as it relates to the COVID-19 crisis, major societal events, and policy decisions. Our analysis found that positive sentiment slightly outweighed negative sentiment overall, but that there was a noticeable uptick in negative sentiment in May and June 2020 in conjunction with the Black Lives Matter protests. The most commonly expressed emotions by these athletes were joy, trust, anticipation, and fear, suggesting that they used social media as an outlet to share primarily optimistic sentiments, while still publicly expressing strong negative sentiments like fear and trepidation about the pandemic and other important contemporary events. Athletic administrators, ACC coaches, support staff, and other professionals can use findings like these to guide sound, evidence-based decision-making and to better track and promote the emotional wellness of student-athletes.

Keywords

• machine learning
• natural language processing
• student-athletes
• emotional well-being
• sentiment analysis
• COVID-19